UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Applications of U-decay series isotopes to studying the meridional overturning circulation and particle… Luo, Yiming 2013

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2013_spring_luo_yiming.pdf [ 5.16MB ]
Metadata
JSON: 24-1.0073792.json
JSON-LD: 24-1.0073792-ld.json
RDF/XML (Pretty): 24-1.0073792-rdf.xml
RDF/JSON: 24-1.0073792-rdf.json
Turtle: 24-1.0073792-turtle.txt
N-Triples: 24-1.0073792-rdf-ntriples.txt
Original Record: 24-1.0073792-source.json
Full Text
24-1.0073792-fulltext.txt
Citation
24-1.0073792.ris

Full Text

  Applications of U-decay series isotopes to studying the meridional overturning circulation and particle dynamics in the ocean by Yiming Luo A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in The Faculty of Graduate Studies (Oceanography) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) April 2013 © Yiming Luo 2013  ii  Abstract Two important questions in the fields of paleoceanography and marine biogeochemistry (the reconstruction of past changes in the strength and geometry of the ocean’s overturning circulation and the quantification of particle flux to the seafloor) are addressed using three isotopes from the U-decay series ( 234 Th, 230 Th and 231 Pa). Two-dimensional scavenging models of the Atlantic and Pacific Ocean were tuned to reproduce the 230 Th and 231 Pa seawater activity profiles measured in these oceans and used to establish the distribution of sediment 231 Pa/ 230 Th generated by simple meridional overturning circulation cells. The results indicate that circulation is the main factor controlling the distribution of sediment 231 Pa/ 230 Th in the Atlantic and confirm the use of this proxy as a paleocirculation tracer. In the Pacific, both circulation and boundary scavenging are important in determining the distribution of sediment 231 Pa/ 230 Th. Thorium-234 scavenging and moored sediment traps yield similar particle flux estimates in Saanich Inlet, on the coast of British Columbia. This study highlights the possibility of estimating the flux of organic carbon in coastal waters by simply measuring 234 Th and POC on particles, which would provide a simple and rapid method for large scale monitoring. Measurements of 234 Th and 230 Th dissolved in seawater and adsorbed on three different size classes of particle were used to estimate particle flux in the epipelagic and mesopelagic zone of the ocean at station Papa. The results suggest that a significant fraction of the carbon flux can be associated with very large, rapidly-sinking particles with very low Th activities, and unaccounted for in Th-based flux estimates.  iii  Preface Contributions of Authors 1) A version of chapter 2 was published as: Luo, Y., Francois, R., and Allen, S. E.: Sediment 231 Pa/ 230 Th as a recorder of the rate of the Atlantic meridional overturning circulation: insights from a 2-D model, Ocean Sci., 6, 381-400, doi: 10.5194/os-6-381-2010, 2010. The dissolved and particulate 231 Pa and 230 Th data were collected by Dr. Roger Francois and Dr. Mike Bacon. I developed the model with the help from Dr. Susan Allen and I tuned the model with my supervisor Dr. Roger Francois. I wrote the manuscript with Dr. Roger Francois and Dr. Susan Allen. Some of the results from chapter 2 were also used in Gherardi, J.-M., Y. Luo, R. Francois, J. F. McManus, S.E. Allen, and L. Labeyrie 2010. Response to Comment by S. Peacock on “Glacial-interglacial circulation changes inferred from 231 Pa/ 230 Th sedimentary record in the North Atlantic region”. Paleoceanography, Vol. 25, PA2207, 5 PP., 2010 doi: 10.1029/2009PA001867. and  iv  Lippold, J., J.-M. Gherardi, and Y. Luo (2011), Testing the 231 Pa/ 230 Th paleocirculation proxy: A data versus 2D model comparison, Geophys. Res. Lett., 38, L20603, doi: 10.1029/2011GL049282 2) A version of chapter 3 has been published in Nature Geosciences as: Lippold J., Luo Y., Francois R., Allen S.E., Gherardi J., Pichat S., Hickey B. and Schulz H.: Strength and geometry of the glacial Atlantic Meridional Overturning Circulation, Nature Geosci 5 (2012) 813-816. J.L., Y.L. and R.F. developed the concept and designed the study. J.L., J.G., S. P. and B.H. performed 231 Pa/ 230 Th measurements. J.L., Y.L. and R.F. performed opal measurements. Y.L., R.F. and S.A. developed and applied the model. S.P., J.G and H.S. provided sample material and age models. J.L., Y.L. and R.F. wrote the manuscript. 3) A version of chapter 4 will be submitted as: Luo Y., Francois R. and Allen S.E. The influence of deep water circulation on the distribution of 231 Pa and 230 Th in the water column and sediments of the Pacific Ocean. The dissolved and particulate 231 Pa and 230 Th data were collected by Dr. Roger Francois. I developed and tuned the model with Dr. Roger Francois and Dr. Susan Allen. The manuscript is being written in collaboration with Dr. Roger Francois and Dr. Susan Allen.   v  Table of Contents Abstract……………………………………………………………………………. ii Preface……………………………………………………………………………..iii Table of Contents…………………………………………………………………v List of Tables…………………………………………………………………...……xiv List of Figures……………………………………………………………………….xvi Acknowledgements……………………………………………………………….xxvi Dedication………………………………………………………………………xxvii 1 Introduction……………………………………………………………… .……..1 1.1 Ocean’s role in climate change……………………………………………………….1 1.2 U-series isotopes………………………………………………………………………4 1.3 Uranium budget in the ocean……………………………………………..…………7 1.4 Thorium and protactinium in the ocean…………..…………………………………9 1.4.1 Particle flux from the euphotic zone – (234Th:238U)…………………………..11 1.4.2 230Th normalization to reconstruct particle flux to the seafloor……………….12 1.4.3 Sediment (231Paxs/ 230 Thxs) to monitor past changes in the meridional overturning circulation of the ocean…………………………………………………………...14 1.5 Thesis objectives………………………………………………..……………………17  vi  2 Sediment 231Pa/230Th as a recorder of the rate of the Atlantic meridional overturning circulation: insights from a 2-D Model.........................................20 2.1 Introduction……………………………………………………………… ..……20 2.2 Model descriptions…………………………………………………………… ...21 2.2.1 Formulation……………………………………………..……………….…21 2.2.2 Overturning circulation…………………………………………..………..……..26 2.2.3 Parameterization……………………………………………………………. ..29 2.3 Dissolved 230Th and 231Pa water column profiles: Data-Model comparison...40 2.4 Fractionation factor: Data-Model comparison………………………………47 2.5 230Th and 231Pa distribution in the control run……………………………..…50 2.5.1 Dissolved 230Th………………………………………………………… ..…51 2.5.2 Particulate 230Th…………………………………………………… ...………52 2.5.3 Dissolved 231Pa……………………………………….……………………..55 2.5.4 Particulate 231Pa……….………………………………………………………56 2.5.5 Dissolved 231Pa/230Th……………………………………………………….58 2.5.6 Particulate 231Pa/230Th……………………………………………………….61 2.6 Sediment 231Pa/230Th: Data-Model comparison……………………………….63 2.7 Discussion………………………………………………………………………..66 2.7.1 The effect of AMOC on sediment 231Pa/230Th…………………………...……....66 2.7.1.1 Vertical variations in sediment 231Pa/230Th induced by the AMOC………….66 2.7.1.2 Horizontal variations in sediment 231Pa/230Th induced by the AMOC……....68  vii  2.7.1.3 Changes in sediment 231Pa/230Th resulting from changes in the rate of the AMOC………………………………………………………………………..71 2.7.1.4 Changes in sediment 231Pa/230Th resulting from changes in the geometry of the AMOC………………………………………………………………… .…….71 2.7.1.5 Possible sampling strategy to constrain past changes in AMOC from sediment 231 Pa/ 230Th………………………………………………………………… ...72 2.7.2 The effect of AABW on sediment 231Pa/230Th………………………………73 2.7.3 The effect of particle composition on sediment 231Pa/230Th……………….74 2.8 Conclusions………………………………………………………………… ..77 3 Strength and geometry of the Glacial Atlantic Meridional Overturning Circulation………………………………………………………………………80 3.1 Introduction……………………………………………………….……………..80 3.2 Materials and Methods…………………………………………...……………83 3.2.1 Core locations and chronology………………….…………..……………..83 3.2.2 Analytical methods ………………………………….………….……………84 3.2.2.1 231Pa/230Th data…………………………………………………….……….……85 3.2.2.2 Biogenic silica……………………………………………………..………..……85 3.2.3 Modeling……………………………………………………………………..85 3.3 Results and discussion………………………………………………………….90 3.3.1 The influence of biogenic silica on the distribution of sediment 231Pa/230Th in Atlantic sediments……………………………………………………………90  viii  3.3.2 The influence of AMOC on the distribution of sediment 231Pa/230Th in Atlantic sediments…………………………………………………………………… .93 3.3.2.1 Holocene……………………………………………………………………..93 3.3.2.2 LGM………………………………………………………………………… .96 3.3.3 Sensitivity test on the LGM 2-D scavenging model…………………………98 3.3.3.1 Sensitivity to sinking rate and fractionation factor…………………..……..98 3.3.3.2 Sensitivity to the strength of the overturning circulation cells (GNAIW; AABW)………………………………………………………………………100 3.3.3.3 Sensitivity to the geometry of the overturning circulation cells (GNAIW; AABW)……………………………………………………………………102 3.3.4 Can the Holocene circulation scheme explain the LGM observations by varying the scavenging parameters?………………………………………………… ..104 3.4 Conclusions……………………………………………………………………107 4 The influence of deep water circulation on the distribution of 231Pa and 230Th in the water column and sediments of the Pacific Ocean…………………….……109 4.1 Introduction……………………………………………………………………...109 4.2 Model descriptions…………………………………………………..…….……112 4.2.1 Overturning circulation……………………………………………………..112 4.2.2 Formulation of the two-dimensional scavenging model in the Pacific Ocean…………………………………………………………………… ..114 4.2.3 Parameterization of the scavenging model…………………………..……..116  ix  4.2.4 Estimating removal by “boundary scavenging” in the 2 -D scavenging model…………..……………………..…….…..…….…..………………....119 4.3 Results and discussion…..…….……..…….…..………………………………....119 4.3.1 Dissolved 230Th and 231Pa water column profiles…………………...………119 4.3.1.1 Data-model comparison………………...……………..…………...………119 4.3.1.2 The influence of boundary scavenging on the curvature of the 231Pa seawater profiles in the North Pacific……...…………..……...……………...………130 4.3.2 231Pa and 230Th sections generated by the Pacific 2-D scavenging model….131 4.3.2.1 Dissolved 230Th...…………..……...……………………………………..…132 4.3.2.2 Particulate 230Th..……...……………………………………………………133 4.3.2.3 Dissolved 231Pa……………………………………….…………….………135 4.3.2.4 Particulate 231Pa………………….…………….….……….……….………137 4.3.2.5 Dissolved Pa/Th….…….…………………….……………………………138 4.3.2.6 Particulate and sedimentary Pa/Th………………..………...………………139 4.3.3 Comparison between sediment Pa/Th measured in the Pacific and model predictions……………..………......……………..………...……………….143 4.3.4 Is the decreasing trend in Pa/Th with depth at mid-depths a result of the PMOC?.……………….……………………………..……………………….151 4.3.5 Sensitivity of sediment Pa/Th in low productivity regions to changes in the rate of PMOC…………….……………………………………………….………….152  x  4.3.6 Relative sensitivity of sediment Pa/Th in low productivity regions to changes in PMOC and “boundary scavenging” – What is the most promising approach to constrain PMOC from sediment Pa/Th? …………………………...……157 4.4 Conclusions………………………………………………………………………159 5 A comparison of POC fluxes recorded by sediment trap and 234Th:238U disequilibrium in a coastal region (Saanich Inlet, British Columbia)……...161 5.1 Introduction…………………………………………………………………….…161 5.2 Measuring the sinking flux of carbon with sediment traps………………...163 5.3 Measuring the sinking flux of carbon using 234Th/238U disequilibrium in surface waters……………………………………………………………………...…..164 5.4 Materials and methods…. …. …. ………………………………………………..168 5.4.1 Study Site………………………………………………………………… ..168 5.4.2 Sample collection, preparation and analyses………………………………170 5.4.2.1 Hydrography………………………………………………………..……..…170 5.4.2.2 Sediment traps………………………………………………………………171 5.4.2.3 234Th method…………………………………………………………………172 5.4.2.3.1 Total 234Th: ……………………………………………………………..173 5.4.2.3.2 Particulate 234Th: …………………………………...…………………..174 5.5 The U-salinity relationship in Saanich Inlet waters…………………………175 5.6 Results and discussion……………………………………………….…………179 5.6.1 Salinity, temperature, density and O2……………………………………….179 5.6.2 Sediment traps…………………………………………………….…………182  xi  5.6.2.1 Sample mass and concentrations……………………………………………182 5.6.2.2 Fluxes from the sediment traps……………………………………………..185 5.6.3 234Th deficits…………………………………………………………….……..192 5.6.3.1 Total 234Th profiles………………………………………………………….193 5.6.3.2 Particulate 234Th…………………………………………….……………….198 5.6.3.3 Dissolved 234Th………………………………………………………….……..204 5.6.3.4 POC/234Th ratio……………………………………………………………..205 5.6.3.5 Comparison between LVP samples and 25mm TQ samples.……………….209 5.6.3.6 Fluxes derived from 234Th:238U disequilibria………………………………210 5.6.3.6.1 234Th fluxes…………………………………………………………….210 5.6.3.6.2 POC fluxes……………………………………………………………218 5.6.4 Comparing POC flux measured with sediment traps and 234Th deficit…223 5.6.5 Can POC fluxes in the coastal region be estimated by only measuring particulate 234 Th and POC? ………………………………………………………. .226 5.7 Conclusions and future perspectives…………………………………………232 6 Particle fluxes and dynamics in the northeast Pacific Ocean from paired water column measurements of Th-230 and Th-234 activity……………….235 6.1 Introduction………………………………………………………………..………235 6.2 Materials and methods………………………………………….………..…...…238 6.2.1 Sample collection and preparation……………………………………...……….238 6.2.1.1 234Th samples……………………………………………………..…………238 6.2.1.2 POC………………………………………..…………………………………242  xii  6.2.1.3 230Th samples……………………………………………………..…………242 6.2.1.4 232Th, P, Al and Ca samples………………………..……………………..…244 6.2.2 Sample analysis……………………………………………………………..244 6.2.2.1 234Th measurements…………………………………………………………244 6.2.2.2 POC measurements……………………………………………………….…245 6.2.2.3 230Th measurements…………………………………………………………246 6.2.2.4 232Th, P, Al and Ca measurements………………………………..………246 6.3 Results……………………………………………………………………………249 6.3.1 234Th…………………………………………………………………………..249 6.3.2 POC……………………………………………………………………….….253 6.3.3 POC/234Th.…………………………………………………………….……..255 6.3.4 230Th…………………………………………………………………………..256 6.3.5 POC/230Th…….. ……………………………..……………………………..260 6.3.6 P and Ca……………………………………………………………..………261 6.3.7 Al………………………………………………………………………..……262 6.4 Discussions………………………………………………………………..…..…264 6.4.1 Th-234 fluxes ………………………………………………………………264 6.4.2 Fluxes of POC, Ca and Al estimated from the 234 Th: 238 U deficit………267 6.4.2.1 POC fluxes……………………………………………………….…….……267 6.4.2.2 Ca fluxes……………………………………………………………………..270 6.4.2.3 Al fluxes……………………………………………………………………..271 6.4.3 230Thxs fluxes……………………………………………………………..…..276  xiii  6.4.4 Fluxes of POC, Ca and Al estimated by normalization to 230Thxs……..277 6.4.4.1 POC fluxes…………………………...………………………………….…..278 6.4.4.2 Ca fluxes…………………………………………………………………..….280 6.4.4.3 Al fluxes…………………………………………………………………..….282 6.4.5 Particle dynamics…………………………………………...…………….…283 6.4.5.1 Model A……………………………………………………………………...284 6.4.5.2 Model B………………………………………………………………….…..288 6.4.5.3 Model C………………………………………………………………………..294 6.5 Conclusions………………………………………………………….……..……303 7 Conclusions and perspectives…………………………………..…………...306 7.1 Summary of major findings and contributions…………………………...............306 7.2 Future research perspectives………………………. …………………...................311 Bibliography……………………………………………………………..………….314 Appendix A…………………………………………………………………………333 Appendix B…………………………………………………………………………346 Appendix C………………………………………………………………….….…..382 Appendix D………………………………………………………………….……...409 Appendix E…………………………………………………………………………432    xiv  List of Tables Table 1.1 Known inputs and outputs of U to the ocean……..……………………………..8 Table 1.2 Summary of major U, Th and Pa isotope data in world’s ocean………………10 Table 2.1 List of abbreviations and values for the model parameters…………………29 Table 2.2 230 Th and 231 Pa activities in sea water..…………………………………….….32 Table 2.3 “Equilibrium” Fractionation Factors……………..……………………….……39 Table 2.4 Holocene Pa/Th in 5 North Atlantic cores…………………………..…….…64 Table 3.1 List of abbreviations and values for the scavenging parameters……………..87 Table 3.2 Latitudinal variations of the equilibrium Fractionation Factors………….…89 Table 4.1 List of abbreviations and values for model parameters………………………117 Table 4.2 “Equilibrium” Fractionation Factors for the Pacific Model…………….……119 Table 4.3 230 Th and 231 Pa activities in sea water………………………….……………121 Table 5.1 Saanich Sampling schedule…………………………………………………171 Table 5.2 Dissolved U concentrations in Saanich Inlet……..…………………………177 Table 5.3 The comparison of fluxes…...……………..…………………………..……..190 Table 6.1 List of samples……………………………………………………………239 Table 6.2 List of Thorium isotope data……………..………………………………247 Table 6.3 POC and POC/ 234 Th on particulate samples…………..…………………248 Table 6.4 Data for particulate P, Al and Ca………………..…………..……….…252 Table 6.5 POC and POC/ 230 Thxs on particulate samples………….……..…………259 Table 6.6 Th-234 fluxes and POC fluxes based on 234 Th………..………………266  xv  Table 6.7 Biogenic Ca/Th and Al/Th ratios for both 234 Th and 230 Thxs……...…………271 Table 6.8 Biogenic Ca and Al fluxes calculated by 234 Th and 230 Thxs methods……...…274 Table 6.9 POC fluxes calculated by 230 Thxs normalization……………………………275                     xvi  List of Figures Figure 1.1 Uranium decay series………………..…………………………………………5 Figure 1.2 The principle of 234 Th: 238 U method…..……………………………………….13 Figure 2.1 The scavenging model…………………………………………………….......25 Figure 2.2 Velocity vector plot………………………………………………………........27 Figure 2.3 Station locations for the water column profile..………………………..….....31 Figure 2.4 Dissolved 230 Th and 231 Pa at 60° - 70°N...…………..……………………......41 Figure 2.5 Dissolved 230 Th…………………………………………………………….....45 Figure 2.6 Dissolved 231 Pa…………………………………………………………….....46 Figure 2.7 Concentration profiles of dissolved 230 Th and 231 Pa in the Southern Ocean…47 Figure 2.8 Distribution of fractionation factors……………………………………….....49 Figure 2.9 Dissolved 230 Th section generated by the model………………………….....51 Figure 2.10 Particulate 230 Th section generated by the model……………………….....52 Figure 2.11 Total 230 Th concentration measured at Station KNR06-3……………….....53 Figure 2.12 Fraction of total 230 Th associated with particles generated by the model.....54 Figure 2.13 Dissolved 231 Pa section generated by the model………………………….....55 Figure 2.14 Particulate 231 Pa section generated by the model…………..…………….....57 Figure 2.15 Fraction of total 231 Pa associated with particles generated by the model….57 Figure 2.16 Total 230 Th concentration measured in the southwestern Atlantic……….....58 Figure 2.17 Dissolved 231 Pa/ 230 Th section generated by the model………………….....59 Figure 2.18 Dissolved 231 Pa/ 230 Th profiles in the North and Equatorial Atlantic…….60  xvii  Figure 2.19 Particulate 231 Pa/ 230 Th section generated by the model………………….....61 Figure 2.20 Sediment 231 Pa/ 230 Th section generated by the model………………….....62 Figure 2.21 Sediment 231 Pa/ 230 Th generated with an opal belt…………………..….....65 Figure 2.22 Lateral velocity profile…………………………………………….…….....69 Figure 2.23 Contrasting lateral velocity profiles between the control run…...…….....70 Figure 2.24 Sediment 231 Pa/ 230 Th field generated without AABW………….……..74 Figure 2.25 The influence of Southern Ocean fractionation factor on the sediment 231 Pa/ 230 Th in Atlantic sediments…………………………………………………….....76 Figure 3.1 Core locations…………………………………………………………….....84 Figure 3.2 Holocene overturning scheme………………………………………………86 Figure 3.3 LGM overturning scheme……………………………………………………87 Figure 3.4 Overturning scheme used to test the very shallow………………………….90 Figure 3.5 Correlation between sediment 231 Pa/ 230 Th and 230 Th-normalized opal flux...91 Figure 3.6 Correlation between sediment 231 Pa/ 230 Th and opal concentration…………93 Figure 3.7 231 Pa/ 230 Th versus water depth……………….………………………….....94 Figure 3.8 Correlation between sediment 231 Pa/ 230 Th and model outputs……………95 Figure 3.9 Variations in the linear correlation between the sediment 231 Pa/ 230 Th database and model output…………………………………………………………………….....99 Figure 3.10 Variations in the mean square weighed deviation (mswd) between the sediment 231 Pa/ 230 Th database and model output………………………………….…99 Figure 3.11 Fit between observations and model outputs generated with the optimal LGM model geometry with varying GNAIW and AABW strengths……………………….....100  xviii  Figure 3.12 Fit between observations and model outputs generated with a very narrow GNAIW as a function of GNAIW and AABW strengths……………………..….....102 Figure 3.13 Sediment 231 Pa/ 230 Th superimposed to the sediment 231 Pa/ 230 Th section…105 Figure 3.14 Linear correlation coefficient and mean square weighed deviations obtained when correlating the sediment database with sediment 231 Pa/ 230 Th generated by the Holocene circulation scheme with varying particle sinking rates and fractionation factors……………………………………………………………………… .........106 Figure 4.1 Velocity vector plot………………………………………………………...113 Figure 4.2 Scavenging and transport model in each model grid box……………............116 Figure 4.3 Station locations………………………………………….………………......121 Figure 4.4 231 Pa and 230 Th profiles measured in the Southern Ocean……………..128 Figure 4.5 Dissolved 230 Th and 231 Pa at equatorial Pacific……….…….…………...129 Figure 4.6 Dissolved 230 Th and 231 Pa at station Aloha………………………………...130 Figure 4.7 Changes in the curvature of the dissolved 231 Pa profile generated at 21°N with 26Sv PMOC and varying effective rate constant for removal to the margins……….....131 Figure 4.8 Dissolved 230 Th section generated by the model…………………………...132 Figure 4.9 Difference between the dissolved 230 Th concentration generated by the 2-D model and the concentration predicted in the absence of circulation………………...133 Figure 4.10 Particulate 230 Th section generated by the model……………..…………...134 Figure 4.11 Fraction of total 230 Th associated with particles generated by the model…134 Figure 4.12 Dissolved 231 Pa section generated by the model…………………………...136  xix  Figure 4.13 Difference between the dissolved 231 Pa concentration generated by the 2-D model and the concentration predicted in the absence of circulation……………....136 Figure 4.14 Particulate 231 Pa section generated by the model………………………….137 Figure 4.15 Fraction of total 231 Pa associated with particles generated by the model…138 Figure 4.16 Dissolved 231 Pa/ 230 Th section generated by the model…………………...138 Figure 4.17 Dissolved 231 Pa/ 230 Th below 1000m……………………………………....140 Figure 4.18 Particulate 231 Pa/ 230 Th section generated by the model…………………...141 Figure 4.19 Sedimentary 231 Pa/ 230 Th section generated by the model………………....142 Figure 4.20 Sediment 231 Pa/ 230 Th generated by the 2D model at 21°S and 21°N……142 Figure 4.21 Distribution of core-top Pa/Th data……………………………..………...144 Figure 4.22 Sediment Pa/Th measured in the Pacific vs. sediment Pa/Th generated by the 2D scavenging model…………………………………………………………….….....145 Figure 4.23 Pa/Th in core tops versus depth……………………………………….....147 Figure 4.24 (a) 230 Th-normalized opal flux against sediment Pa/Th in the equatorial Pacific; (b) Difference between Pa/Th measured and estimated from 230 Th-normalized fluxes as a function of depth and the linear regression…………………………...….....148 Figure 4.25 Vertical sediment Pa/Th profiles generated between 35°N and 45°S by the 2D scavenging model in the absence of PMOC and with varying boundary scavenging strength…………………………………………………………………………….....151 Figure 4.26 Variations in sediment Pa/Th as a function of latitude in low productivity regions predicted by the 2D scavenging model with varying PMOC rates…………154  xx  Figure 4.27 Changes in the sediment Pa/Th vertical gradient between deep and intermediated depth as a function of latitude and PMOC strength………………155 Figure 4.28 Difference in sediment Pa/Th generated by the 2D scavenging model with PMOC = 26Sv and 13Sv…………………………………………….…………….....156 Figure 4.29 Variations in sediment Pa/Th as a function of latitude in low productivity regions predicted by the 2D scavenging model with varying boundary scavenging….158 Figure 4.30 Changes in the difference in sediment Pa/Th between 3000 and 4750m as a function of latitude for a fixed PMOC (26Sv) and varying effective removal rate constant to the margins……………………………………………………………………….....159 Figure 5.1 Map of Saanich inlet and coastal southwestern British Columbia………….169 Figure 5.2 Time-series hydrographic data from surface to 200m during the 2 year period………………………………………………………………………………...181 Figure 5.3 Seasonal changes in Opal% and OM% in the material collected by sediment traps at 3 depths in Saanich Inlet…………………….………………………………...184 Figure 5.4 (a) Seasonal variations in the OC/N ratio of sediment trap material collected at the three depths. (b) OC versus N for all the sediment -trap samples……186 Figure 5.5 Mass fluxes recorded by sediment traps deployed at three different depths over the 2-year time series………………………………………………………… ...187 Figure 5.6 Opal fluxes recorded by sediment traps deployed at three different depths over the 2-year time series……………………………………………………………...188 Figure 5.7 POC fluxes recorded by sediment traps deployed at three different depths over the 2-year time series…………………………………………………………………...189  xxi  Figure 5.8 Total 234 Th activities measured……………………………………………...194 Figure 5.9 Average total 234 Th measured at a given depth……………………………...195 Figure 5.10 Total 234 Th during (a) late winter, (b) early spring, (c) late summer or early fall…………………………………………………………………………………...197 Figure 5.11 Monthly composite time-series of total 234 Th……………………………197 Figure 5.12 Average fraction of particulate 234 Th (% of total 234 Th) as a function of water depth………………………………………………………………………………....198 Figure 5.13 Particulate 234 Th activities over an annual cycle…………………………199 Figure 5.14 Average particulate 234 Th measured at a given depth……………………199 Figure 5.15 Particulate 234 Th during (a) late winter, (b) early spring, (c) late summer or early fall………………………………………………………………………………...200 Figure 5.16 Monthly composite time-series of particulate 234 Th………………………..201 Figure 5.17 Average fraction of dissolved 234 Th as a function of water depth………….201 Figure 5.18 Dissolved 234 Th activities over an annual cycle…………………………...202 Figure 5.19 Average dissolved 234 Th profile…………………………………………….202 Figure 5.20 Dissolved 234 Th during (a) late winter, (b) early spring, (c) late summer or early fall………………………………………………………………………...............203 Figure 5.21 Monthly composite time-series of dissolved 234 Th………………………204 Figure 5.22 POC/ 234 Th on particulate samples over an annual cycle………………..206 Figure 5.23 Average of all the POC/ 234 Th measured in particles collected at a given depth………………………………………………………………………………..206  xxii  Figure 5.24 POC/ 234 Th during (a) late winter, (b) early spring, (c) late summer or early fall………………………………………………………………………................207 Figure 5.25 Monthly composite time-series of POC/ 234 Th……………………………208 Figure 5.26 Comparison between POC, particulate 234 Th and POC/ 234 Th obtained by collecting particles with Large Volume in-situ Pumps (LVP) and small volume filtration………………………………………………………………………................209 Figure 5.27 234 Th fluxes calculated from estimates of 238 U seawater concentration obtained from equation 5.6 (blue line) and 5.10 (red line)……………………………212 Figure 5.28 Average calculated 234 Th flux profile…………………………………213 Figure 5.29 (a) 234 Th fluxes in late winter and during the spring bloom; (b) 234 Th fluxes in late summer and early fall………………………………………………………………214 Figure 5.30 Th-234 fluxes calculated with the steady state model vs the non-steady state model……………………………………………………………………… ...........216 Figure 5.31 Ratio of fluxes estimated with the non-steady state model to those estimated with the steady state model from 238 U seawater concentration estimates obtained from equation 5.5………………………………………………………………………..........217 Figure 5.32 Average POC fluxes derived at a given depth with the steady state model from 238 U seawater concentration estimates obtained from equation 5.5………………219 Figure 5.33 (a) POC fluxes in late winter and during the spring bloom, (b) POC fluxes in late summer or early fall………………………………………………………………..220  xxiii  Figure 5.34 Contour plot of the seasonal and depth variation in POC fluxes derived with the steady state model from 238 U seawater concentration estimates obtained from equation 5.5………………………………………………………………………..................221 Figure 5.35 Average POC fluxes derived from 234 Th deficit assuming steady state from 238 U seawater concentration estimates obtained from equation 5.6 and average POC fluxes measured by sediment traps…………………………………………………………..222 Figure 5.36 Two year time series records of POC fluxes measured with sediment traps and corresponding fluxes derived from 234 Th deficits…………………………………223 Figure 5.37 (a) Ratios of organic carbon fluxes obtained with equation 5.12 and 5.5; (b) ratios of organic carbon fluxes obtained with equation 5.13 and 5.5……………228 Figure 5.38 Comparison of organic carbon fluxe time series obtained with sediment traps, equation 5.5 (blue dots) and equation 5.13 (black dots) from 238 U seawater concentration estimates obtained from equation 5.6 for all cases……………………………………230 Figure 5.39 (a) POC fluxes estimated from equation 5.5 and 238 U seawater concentration estimated from equation 5.6 versus POC fluxes from the sediment traps; (b) POC fluxes estimated from equation 5.13 and 238 U seawater concentration estimated from equation 5.6 versus POC fluxes from the sediment traps………………………………………231 Figure 6.1 Model (A) used to describe thorium cycling in seawater…………………237 Figure 6.2 Total 234 Th, particulate 234 Th and 238 U profiles at OSP……………………250 Figure 6.3 Profiles of fine particulate 234 Th obtained by small volume filtration on 25mm TQ filtration, by LVPs on Supor filters, and large (>53 μm) particulate 234Th collected with LVPs on Nylon mesh…………………………………………………………….251  xxiv  Figure 6.4 Profiles of 234 Th in large particles and “extra-large” particles……………251 Figure 6.5 POC profiles and POC/ 234 Th for the particles collected on 25mm TQ filters, the fine particles collected by LVP on Supor filters, and Nitex mesh………………...254 Figure 6.6 Redfield ratio derived from the 25mm TQ and LVP Supor data by assuming the POC/ 234 Th are the same on those two different types of fine particles……………255 Figure 6.7 Profiles of dissolved 230 Th, 230 Th on fine particles and large particles…257 Figure 6.8 The fraction of total 230 Th associated with fine particles………………257 Figure 6.9 Profiles of 230 Th on fine, large and XL particles…………………………258 Figure 6.10 POC/ 230 Th for particles collected by LVP on Supor filters and Nitex mesh.......................................................................... .............................258 Figure 6.11 P, Ca, and Al concentrations associated with fine, large and XL particles………………………………………………………………………................263 Figure 6.12 Th-234 fluxes based on 234 Th: 238 U deficit………………………………….265 Figure 6.13 POC fluxes derived from the 234 Th flux and POC/ 234 Th ratio on the 25mm TQ filters.……………………………………………………………………… .............268 Figure 6.14 POC fluxes derived from the POC/ 234 Th ratio on all three samples………270 Figure 6.15 Biogenic Ca fluxes derived from the biogenic Ca/ 234 Th ratio on LVP-Supor and mesh samples………………………………………………………………………271 Figure 6.16 Al fluxes derived from the Al/ 234 Th ratio on LVP samples………………272 Figure 6.17 POC fluxes derived by 230 Th normalization on fine particles compared with the POC fluxes estimated by other means……………………………………………278  xxv  Figure 6.18 Biogenic Ca fluxes derived by 230 Thxs normalization compared to those derived by 234 Th: 238 U and sediment traps………………………………………………282 Figure 6.19 Al fluxes derived by 230 Thxs normalization compared to those derived by 234 Th: 238 U…………………………………………………………………………..283 Figure 6.20 Output from model A with K1 = 0.5/y and K-1 = 1.6/y; B1 = 3/y and B-1 = 150/y; and S = 150m/day………………………………………………………...286 Figure 6.21 Model B…………………………………………………………………..288 Figure 6.22 The optimal run of model B………………………………………………290 Figure 6.23 Model B for POC dynamics………………………………………………291 Figure 6.24 POC concentrations generated by Model B………………………………293 Figure 6.25 Model C used to describe the complex thorium cycling in seawater……294 Figure 6.26 Model C used to describe the complex POC dynamics in seawater………295 Figure 6.27 Th concentrations produced by model C…………………………………296 Figure 6.28 POC concentrations produced by model C………………………………297 Figure 6.29 POC fluxes calculated by [POC]*S from Model C………………………298 Figure 6.30 POC/ 230 Th ratios on fine particles produced by model C compared to that measured on LVP-Supor samples………………………………………………………301 Figure 6.31 POC fluxes based on 230 Th normalization from model C output compared to the POC fluxes results by other means…………………………………………………301 Figure 6.32 POC fluxes based on 230Th normalization compared to the ‘real’ POC fluxes calculated by [POC]*S from model output and the sediment trap data………………302   xxvi  Acknowledgements I would like to sincerely thank my advisor, Dr. Roger Francois, for his support and guidance on my research. His contagious energy and passion on science have shown me how to be a great scientist. He always enlightens me with his fresh ideas when I got stuck with my research. I feel so lucky to have Roger as my supervisor not only because of his support on my research but also because of his encouragement and help on my life when I felt down. Special thanks are given to my committee members: Dr. Susan Allen, Dr. Lisa Miller, Dr. Kristin Orians and Dr. Evgeny Pakhomov for their valuable and constructive comments, suggestions and proofreading of my thesis. In particular, Susan’s help with the set-up of my 2D scavenging model and Lisa’s help with the beta counting of my 234Th samples are very crucial to my Ph.D. thesis. I also greatly appreciate the help from all the past and present lab members of Roger Francois’s research group in the department of earth and ocean sciences. I want to thank Maureen Soon, Samuel Jaccard, Kristina Brown, Bart De Baere, Drew Snauffer, Genna Patton and Aram Goodwin for their generous help during my study here. I would like to thank all my collaborators Joerg Lippold, Jeanne Gherardi and Sylvain Pichat for their priceless input to all our published and unpublished research fruits. I also would like to thank the Journal of Ocean Science and Nature Geosciences for giving me the permission to reproduce the figures in this thesis. Lastly, but most importantly, I would like to express my deep and sincere gratitude to all my family members for their support for the success of the Ph.D. research work.  xxvii  Dedication To my parents & my wife Yu Ling           1   Chapter 1 Introduction  The ocean impacts global climate in several important ways: its biogeochemistry and overturning circulation strongly influence the atmospheric CO2 level and greenhouse warming (Sigman et al., 2010), while its surface and overturning circulation contribute significantly to the redistribution of solar heat on the surface of the planet (Ganachaud and Wunsch, 2000). In this thesis, the importance of these fundamental processes for climate evolution is further explored by applying U-decay series isotope systematics, which provides some of the most important geochemical tools to estimate rates of processes occurring in the modern and past oceans (Cochran, 1992; Henderson and Anderson, 2003). This introduction describes the link between ocean processes and climate and some of U-series isotopes applications to study these processes, focusing on thorium (Th) and protactinium (Pa) isotopes.  1.1 Ocean’s role in climate change The ocean strongly affects the evolution of Earth’s climate on decadal to multi-millennial timescales. This stems from the fact that the ocean is a very large reservoir of heat  2  (Rahmstorf, 2002) and carbon (Sigman et al., 2010), capable of responding to forcing on these timescales. In particular, it is believed that the ocean plaid a key role in the climatic variability associated with the waxing and waning of the ice ages during the Quaternary (Rahmstorf, 2002; Sigman et al., 2010). While it is well established that the timing of Quaternary climate variability is controlled by periodic changes in the eccentricity of Earth’s orbit, the tilt of its axis of rotation, and the precession of the equinoxes, the amplitude of glacial-interglacial climate changes appears to be controlled to a large extent by physical, chemical and biological processes taking place in the ocean. Abrupt climate changes that punctuated the ice ages and deglaciations have also been attributed to the response of the ocean to addition of freshwater from collapsing ice sheets (Schmittner et al., 2002; Clark et al., 2002). Decadal climate oscillations, such as the El Nino Southern Oscillation and Pacific Decadal Oscillation, are also strongly influenced by ocean processes (McPhaden, 1999) and the ocean also plays a key role in mitigating global warming from anthropogenic CO2 emissions by absorbing a significant fraction of the emitted CO2 (Sabine and Tanhua, 2010) and by storing some of the heat induced by greenhouse warming (Meehl et al., 2011). Among the ocean processes that play a critical role in climate regulation, the Ocean’s meridional overturning circulation (MOC), also known as the thermohaline circulation, is particularly important for redistributing solar heat on the surface of the planet, and particularly, in delivering heat from low to high latitudes in the North Atlantic (Rahmstorf, 2002). This heat transport has a profound impact on the regional climate of land masses in the northern hemisphere and the response of the MOC to global warming is one of the  3  main uncertainties in our prediction of the future evolution of climate. Heat and moisture transport by the MOC is also believed to be fundamentally important in understanding the waxing and waning of the northern continental ice sheets during the Quaternary and the recurrence of abrupt, millennial-scale climatic events in the past. Moreover, changes in the overturning circulation are also believed to have significant impacts on the concentration of CO2 in the atmosphere because it determines the rate at which deep waters with high metabolic CO2 content return to the surface. Carbon uptake by the ocean occurs through a range of mechanisms that are often referred to as 'carbon pumps'. We distinguish the “Solubility Pump”, which is controlled by temperature, the “Biological Pump”, which is controlled by the balance between the export flux of organic carbon from surface to deep water and the return to the surface of CO2 produced by the decomposition of organic matter in the deep sea by the meridional overturning circulation, and the “Carbonate Pump”, controlled by the formation and dissolution of calcium carbonate, which dictates the alkalinity of seawater. Reference is also sometime made to the “Continental Shelf Pump” to emphasize the particular role of highly productive shelf waters and marginal seas (Lee et al., 2011). Because the dissolved inorganic carbon content of the ocean is sixty times larger than the CO2 content of the atmosphere, ocean circulation, biological productivity, and the chemical and physical properties of seawater largely dictate the level of atmospheric CO2 and its variability through time on time scales greater than the mixing time of water in the ocean (~ 1000 years). Likewise, increasing sea surface temperature and pH as a result of anthropogenic CO2 emissions will likely produce changes in marine ecosystems that could significantly  4  affect the ocean’s biological, carbonate and solubility pumps in a way which would hamper CO2 uptake by the ocean and exacerbate the climatic impact of future emissions. While it is clear that oceanic processes largely control the level of atmospheric CO2 on a range of timescales, the details of the physical, biological and chemical mechanisms involved and their complex interactions are still poorly understood. Such an understanding is, however, crucial to make more reliable predictions regarding the evolution of climate in the coming centuries and guide society in taking appropriate actions. Two distinct aspects of this overarching question are addressed in this thesis. In Chapter 2, 3 and 4, I contribute to the development of a new tool to constrain past changes in the meridional overturning circulation on glacial-interglacial timescales, while in chapter 5 and 6, I further develop an approach to better assess the export of marine particles from surface to deep water. To tackle these questions, I make use of isotopes from the uranium decay series, which provide a unique tool box to quantify the rates of various oceanic processes, and in particular the rate of the overturning circulation and the sinking flux of particles, as described in the following sections of this chapter.  1.2 U-series isotopes Uranium has two long-lived isotopes initiating a decay chain (Fig. 1.1): 238 U (half-life = 4.47 10 9  y) and 235 U (half-life = 7.04 10 8  y). They were produced billions of years ago in supernovae by nucleosynthesis, but because of their very long half-lives, they still persist to this day. When left undisturbed, all the daughters in the series (Fig. 1.1) have the same  5  decay rate (or the same “activity”). The decay series is then said to be in “secular equilibrium”. In this situation, each radioactive isotope of the series decays at a rate which is equal to the rate at which it is produced (which is the rate at which its parent is decaying). Ultimately, the decay rates of all the daughters are dictated by the decay rate of the initial parent ( 238 U or 235 U), which decreases very slowly with time.  Figure 1.1: Uranium decay series  6  The usefulness of these isotopes in environmental science stems from the fact that processes on Earth’s surface often separate daughters from parents, thereby perturbing secular equilibrium (Bourdon et al., 2003). This arises because the successive decay series daughters consist of different elements with very different chemical properties (Fig. 1.1). For instance, while uranium and radium are relatively soluble in water, thorium, protactinium, polonium and lead have very low solubility and are readily removed from aqueous solutions by adsorption on surfaces of particles or sediments. On the other hand, radon is a gas which diffuses into the atmosphere from the rocks or water where it is produced from the decay of radium. When an environmental process (e.g., particle scavenging, air-sea exchange, etc.) separates a daughter from its parent, the activity of the daughter isotope (i.e., the rate at which it decays) decreases and becomes lower than the activity of its parent isotope. Assuming steady-state, we can write a simple mass balance equation stating that the rate at which the daughter is produced (i.e., the rate at which the parent decays or the activity of the parent; parent Nparent) must be equal to the rate at which the daughter decays (i.e., the activity of the daughter; daughter Ndaughter) plus the rate at which the daughter is removed by the environmental process (Rdaughter). Thus: Rdaughter (atoms.min -1 ) = parent Nparent - daughter Ndaughter where  is the decay constant, N is the number of atoms and R is the rate of removal or addition of the daughter isotope by the environmental process of interest  7  Measuring the difference in activity between parents and daughters in environmental samples thus provides a means of quantifying the rate of removal (or addition) of the daughter from (to) the sample. In turn, determining the rate of removal of the daughter by a given process under different environmental conditions provides quantitative information on the relative intensity of this process under these environmental conditions and adds insight into the mechanisms involved.  1.3 Uranium budget in the ocean Uranium is supplied to the ocean by rivers at an estimated rate of 1.1*10 10 g/year (±35%) (Cochran, 1992; Palmer and Edmond, 1993). Additional sources from wind-blown dust and ground water discharge are less important and difficult to estimate (Dunk et al., 2002; Henderson and Anderson, 2003). In most of the ocean where oxidizing conditions prevail, uranium is soluble by forming soluble uranyl-carbonate species in sea water. Uranium is removed from the ocean mainly by burial in anoxic or suboxic sediments after reduction to its insoluble tetravalent state in pore waters, or by the alteration of ocean basalts. Estimates of the major terms of the U budget are given in Table 1.1 (Dunk et al., 2002). Within relatively large uncertainties (e.g., the total removal flux is only known within a factor of 2, Henderson and Anderson, 2003), the oceanic U budget appears to be balanced, and the oceanic residence time of U has been estimated at 320-560 kyr (Dunk et al., 2002). With such a long residence time, U is a conservative element with a nearly constant  8  concentration in seawater (3.3 ppb). Estimating the U sedimentary sink during glacial periods suggests that seawater U concentration did not change significantly during the 100 kyr glacial-interglacial cycle (Rosenthal et al., 1995). Table 1.1: Known inputs and outputs of U to the ocean. Some other fluxes such as groundwater input and input or removal in estuaries are poorly known and are not included in this table (Henderson and Anderson, 2003).    9  1.4 Thorium and protactinium in the ocean Thorium is found at the +4 oxidation state and forms a highly insoluble neutral hydroxide (Th(OH)4 0 ) in seawater (Langmuir and Herman, 1980). As a result, it has a strong tendency to adhere to particle surfaces (Cochran, 1992). Thorium isotopes produced by uranium decay are thus removed rapidly from the water column by particle scavenging. Among the Th isotopes commonly used in oceanographic studies, 232 Th is supplied from continental sources by rivers and aeolian dust and initiates its own decay series, while 230 Th and 234 Th are added to seawater by decay of U isotopes ( 234 U and 238 U, respectively). Since uranium is conservative, these thorium isotopes are thus produced uniformly over the entire water column, which is a key advantage for using these isotopes to quantify particle flux and ocean circulation. Because of its long half-life (75.69 ± 0.23 ka; Cheng et al., 2000), 230 Th is mostly removed from seawater by scavenging. Its high affinity for particle surfaces results in short residence times in seawater, ranging from a few months in surface waters to a few decades in deep waters (Table 1.2). In contrast, removal of the short lived 234 Th (half-life: 24.1 days) from seawater is mostly by radioactive decay over most of the water column, but in surface waters, both scavenging and decay become important. Thorium-234 is thus in secular equilibrium with its parent isotope 234 U over most of the water column, except at the surface where rapid scavenging results in a residence time similar to 234Th’s half-life. Thorium-234 is best suited for studying faster processes occurring in the upper ocean (e.g., export flux; Buesseler et al., 2006 and references therein) and surface sediments (bioturbation; Aller and DeMaster, 1984), while the longer half-life of 230 Th provides constraints on the rate of slower processes occurring  10  deeper in the water column (e.g., scavenging, Bacon and Anderson, 1982; circulation, Marchal et al, 2000) and in sediments (e.g., sedimentation; Bacon, 1984). Table 1.2: Summary of major U, Th and Pa isotope data in world’s ocean. Nuclide Concentration in the ocean Residence time Half life Mode of removal (g/g) (dpm/T) (years) (years) 238 U 3.3*10 -9  2.5*10 3  4*10 5  4.46*10 9  Burial & Basalt 235 U 2.4*10 -11  3.5*10 3  4*10 5  2.34*10 7  Burial & Basalt 234 U 2*10 -13  2.9*10 3  1.9*10 5  2.45*10 5  Burial, Basalt, Decay 234 Th (1.9-4.8)*10 -20  (~1 – 2.5)*10 3  ~0.04 – 0.1 6.6*10 -2  Decay & Scavenging 230 Th (0.1-3.3)*10 -17  (0.5-15)*10 -1  0.1 - 55 7.57*10 4  Scavenging 231 Pa (0.1-5.7)*10 -18  (0.1-6)*10 -1  1 - 240 3.25*10 4  Scavenging  Protactinium-231 is also produced uniformly in seawater by the decay of 235 U, and it has a relatively long half-life (32.71 ± 0.11 ka; Robert et al., 1969). As for 230 Th, it is mostly removed from the water column by scavenging, but 231 Pa is not as particle reactive as 230 Th. As with Th 4+ , Pa forms insoluble Pa(V) hydroxide complexes in aqueous condition. Its generally lower particle reactivity may stem from the strong complexing tendency of fluoride ions for Pa(V), which produces a range of negatively charged ions (Pal’shin et al., 1970). When the surface available for adsorption is silica, however, this difference disappears. In the ocean, silica is an important biomineral produced mostly by diatoms in the more productive regions of the ocean, such as upwelling regions and the southern Ocean. Although this observation has been confirmed in field (Walter et al., 1997) and  11  laboratory (Gueguen and Guo, 2002) studies, its chemical underpinning is not yet fully understood, but may be due to colloidal conditions at the silica-water interface, which reduces the hydrolysis of the Pa(V) (Roberts, 2008). Nonetheless, outside the oceanic regions dominated by diatoms (mainly the southern ocean and the North Pacific), the contrast between the solubility of Pa and Th forms the basis for several oceanographic and paleoceanographic applications. Below, I review briefly some of the main applications of Th and Pa isotopes in marine studies.  1.4.1 Sediment (231Paxs/ 230 Thxs) to monitor past changes in the meridional overturning circulation of the ocean. Uranium-234 decay produces 2.67 × 10 -2  dpm/m 3 /y of 230 Th while 235 U decay produces 2.46 × 10 -3  dpm/m 3 /y of 231 Pa resulting in a constant production rate ratio of ( 231 Pa/ 230 Th) = 0.092 (dpm/dpm). However, the 231 Pa/ 230 Th ratio measured in marine sediments mostly deviates from this ratio. This is because the two isotopes have different residence times in the water column (Table 1.2). Protactinium-231 is usually less particle-reactive, and because of the resulting longer residence time in the water column, it is more effectively transported laterally by advection or diffusion than 230 Th. As a result, sediment 231 Pa/ 230 Th is higher than the production rate ratio in regions that receive laterally transported 231 Pa, while sediments underlying regions exporting 231 Pa have ratios < 0.092 (e.g., Francois, 2007).  12  The lateral transport of 231 Pa can be driven by turbulent mixing between areas of contrasting scavenging intensity, the latter depending on particle flux and composition. Scavenging intensity increases with particle flux and the concentration of opal in the settling particles. The latter reflects the greater affinity of 231 Pa for silica than for other substrates (Chase et al., 2002). As a result, there is a net 231 Pa transport from the low productivity central gyre regions of the ocean towards the margins, a process called “boundary scavenging” (Bacon, 1988), which has been used to explain the sediment 231 Pa/ 230 Th distribution pattern in the Pacific (Yang et al., 1986; Anderson et al., 1990) and to estimate variations in opal flux in the past (Anderson et al., 2009). Alternatively, lateral transport of 231 Pa can be driven by deep water circulation and particularly the meridional overturning circulation of the ocean. This type of lateral transport forms the basis for using sediment 231 Pa/ 230 Th to reconstruct past changes in the Atlantic Meridional Overturning Circulation [AMOC] (Yu et al., 1996; Marchal et al., 2000; McManus et al., 2004; Gherardi et al., 2005, 2009; Hall et al., 2006; Negre et al., 2010) and this is the application which is further developed in this thesis.  1.4.2 Particle flux from the euphotic zone – (234Th:238U) Particle export from the euphotic zone is a key aspect of the biological pump which sequesters carbon to the deep sea, and it has become a focus in the research of many oceanographers interested in the fate of anthropogenic CO2 added to the atmosphere.  13  Measuring the disequilibrium between the 234 Th and 238 U in surface waters has become a tool of choice to estimate this process.  Figure 1.2: The principle of 234 Th: 238 U method (Weinstein et al., 2005). M is the element of interest in the particles that scavenged the Th in the water column. CM is the concentration of this element in the particles and A p Th is the activity of 234 Th scavenged by the same particles. PTh is the flux of 234 Th derived from the 234 Th: 238 U deficit in upper water column. PM is the flux of M. The 234 Th activity in seawater results from the balance between production from 238 U decay, removal by radioactive decay and scavenging on sinking particles, and transport by advection and diffusion. The temporal change in total 234 Th activity in the mixed layer is expressed by: әTh/әt = λAU- λATh-P+V (1.1) D ep th  Activity  14  where AU and ATh are dissolved 238 U and total 234 Th activities integrated over the depth of the mixed layer (dpm/m 2), respectively, λ is the decay constant of 234Th (=0.02876/day), P is the net removal flux of 234 Th by scavenging (dpm/m 2 /d), and V is the sum of the advective and diffusive fluxes. Advection and diffusion can often be neglected, particularly when the water column is stratified (Savoye et al, 2006). Therefore, under these conditions, by assuming steady state conditions, equation 1.1 can be simplified as: P =λAU- λATh (1.2) Since 238 U is conservative in the ocean, its activity in the water column can be derived from salinity. The net 234 Th flux at the base of a layer (P) equals the 234 Th deficit integrated over that layer multiplied by the decay constant of 234 Th. Measuring profiles of total 234 Th activity in the upper water column and the ratio of organic carbon concentration to 234 Th activity in sinking particles provides an estimate of the export flux of organic carbon (Fig. 1.2). The application of 234 Th: 238 U method expanded during the JGOFS program (Buesseler et al., 1992) and has been increasingly used since (Waples et al., 2006).  1.4.3 230Th normalization to reconstruct particle flux to the seafloor Th-230 is produced in the water column by decay of 234 U. Its removal by scavenging is best described by reversible adsorption on the surface of settling particles (Bacon and  15  Anderson, 1982; Nozaki et al., 1987), which predicts a gradual increase in dissolved and particulate 230 Th concentration or activity with depth. This prediction has now been confirmed by multiple measurements in the world ocean (e.g., Francois, 2007). Because of its very high particle reactivity, the residence time of 230 Th in the water column is short (at most 40 years in deep waters), and considering its very long half-life, almost all of the removal is due to scavenging. As a result of its short residence time, 230 Th is effectively removed by scavenging into the underlying sediment before it can be extensively transported laterally by advection or diffusion. This means that the scavenged flux of 230 Th to the seafloor is always nearly equal to its production rate in the overlying water, which is a simple linear function of water depth (Bacon, 1984): PTh =Z ( 234U) λ230 (1.3) where PTh is the production rate of 230 Th in the oceanic water column (dpm/m 2 /y), Z is the water depth (m), ( 234 U) is the activity of 234 U in seawater (= 2750 dpm/m 3) and λ230 is the decay constant of 230 Th (y -1 ). If removal by scavenging were instantaneous, the vertical flux of 230 Th would be exactly equal to the production rate: FTh=PTh. Of course, this is only an approximation, since the removal is not instantaneous. Nonetheless, it is fast enough that FTh rarely deviates from PTh by more than 30% (Henderson et al., 1999; Yu et al., 2001). This forms the basis for an important application in paleoceanography in which 230 Th is used as a constant flux  16  tracer to normalize the flux of other sedimentary constituents accumulating on the seafloor (Bacon, 1984; Francois et al., 2004). Accurately estimating past variations in the flux of biogenic and lithogenic particles reaching the seafloor is central to better understanding the response of the ocean to past environmental conditions and assess its role in the evolution of these environmental conditions. Sedimentation rate on the seafloor is however controlled not only by the vertical rain rate of particles from the overlying surface, which is generally the quantity of interest when assessing past environmental conditions, but also on complex interactions between bottom topography and ocean currents that redistribute sediments on the seafloor after their initial deposition. Because post- or syn-depositional transport by bottom currents redistribute equally 230 Th and all other sediment constituents (except in cases of significant size fractionation), normalizing the sediment constituents to 230 Th by assuming that 230 Th flux to the seafloor is equal to its rate of production in the overlying water can therefore correct for lateral redistribution by bottom currents. This method was first proposed by Bacon (1984) and further developed in the following two decades (Francois et al., 2004). It has now become a standard tool in paleoceanography to assess past changes in the flux of particles reaching the seafloor and preserved in sediments.     17  1.5 Thesis objectives Applications of U-decay series isotopes in oceanographic and paleoceanographic studies have multiplied in recent decades, but key questions remain unresolved. For instance, the relative importance of particle composition and ocean circulation in controlling sediment 231 Pa/ 230 Th is still being debated (e.g., Keigwin and Boyle, 2008; Gherardi et al., 2009; Peacock et al., 2009), as well as the validity of 230 Th-normalization to reconstruct particle flux from the sedimentary record (Francois et al., 2008; Lyle et al., 2008). The exact interpretation of sediment 231 Pa/ 230 Th in terms of changes in deep water circulation has also been questioned (Thomas et al., 2006; Negre et al., 2010). Likewise, while calculating the 234 Th export flux from the 234 Th/ 238 U disequilibrium in surface waters is straightforward, estimating the associated flux of carbon is much more problematic (Rutgers van der Leoff et al., 2006; Buesseler et al., 2006), stemming from uncertainties in the dynamics and composition of the particles removing 234 Th from surface waters. Although 234 Th provides a useful tool to estimate export flux from surface water, it does not provide information on the fluxes of particles in deeper water, where most of the carbon remineralization occurs. Yet, the depth of organic matter remineralization is the key to establishing the efficiency of export production in sequestering carbon to the deep sea. These are the questions and problems that this thesis attempts to address. It is subdivided into 5 chapters following this introduction and concludes with a brief “Conclusions and Perspectives” section. The five core chapters are:  18  Chapter 2: Sediment 231 Pa/ 230 Th as a recorder of the rate of the Atlantic Meridional Overturning Circulation: insights from a 2-D model In this chapter, I develop a 2D scavenging model to establish the meridional and vertical distribution patterns of sediment 231 Pa/ 230 Th generated by simple Atlantic meridional overturning cells. Chapter 3: Strength and geometry of the Glacial Atlantic Meridional Overturning Circulation In this Chapter, I report results from a collaboration with colleagues who measured 231 Pa/ 230 Th in sediment samples from the Holocene and last glacial sections of multiple cores from the Atlantic to constrain the rate and geometry of the Atlantic Meridional Overturning Circulation during the last ice age. My contribution was in using my model to better interpret the data generated by my co-authors. Chapter 4: The influence of deep water circulation on the distribution of 231 Pa and 230 Th in the water column and sediments of the Pacific Ocean In this chapter, I expand the 2D scavenging model to include a representation of the Pacific Ocean to explore the potential of sediment 231 Pa/ 230 Th in constraining past changes in the overturning circulation in this ocean. Chapter 5: Comparison of POC fluxes measured with sediment trap and 234 Th: 238 U disequilibrium in a coastal setting (Saanich Inlet, British Columbia)  19  In this chapter, I report a time-series study (Feb/2009 to Feb/2011) comparing fluxes measured by 234 Th scavenging and sediment traps in Saanich Inlet, British Columbia. Chapter 6: Particle fluxes and dynamics in the northeast Pacific Ocean studied by paired measurements of thorium isotope activities In this chapter, I explore the possibility of combining measurements of 234 Th and 230 Th dissolved in seawater and adsorbed on three different size classes of particle to estimate particle flux in the mesopelagic zone of the ocean to assess the efficacy of the biological pump.          20   Chapter 2 Sediment 231Pa/230Th as a recorder of the rate of the Atlantic Meridional Overturning Circulation: insights from a 2-D model  2.1 Introduction Ocean circulation plays an important role in climate control by transferring solar heat from low to high latitudes (Ganachaud and Wunsch, 2000). In particular, rapid changes in the strength and geometry of the Atlantic Meridional Overturning Circulation (AMOC) have been invoked to explain the abrupt variations in climate that have punctuated the last ice age and deglaciation (Schmittner et al., 2002; Clark et al., 2002). However, documenting the link between changes in climate and ocean circulation still remains a major challenge in paleoclimatology (Lynch-Stieglitz et al., 2007). Past changes in circulation were first inferred from the sedimentary records of nutrient proxies (Boyle and Keigwin, 1987). While these tracers provide important information on changes in the geometry of the overturning circulation, they do not constrain changes in the rate of overturning (Legrand and Wunsch, 1995). To address this problem, several kinematic tracers of ocean circulation  21  are being investigated (Lynch-Stieglitz et al., 2007). The 231 Pa/ 230 Th ratio of Atlantic sediments is one of these tracers. This proxy has recently been used to investigate past changes in the rate of the AMOC from the last glacial maximum to present (McManus et al., 2004; Hall et al., 2006; Gherardi et al., 2005; 2009). Because both 231 Pa and 230 Th have uniform production rates (from the decay of dissolved uranium) and 231 Pa has a longer residence time than 230 Th in the water column, the AMOC exports 231 Pa more effectively from the Atlantic into the Southern Ocean (Yu et al., 1996; Francois, 2007). The modern rate of overturning results in the mean residence time of deep water in the Atlantic roughly equivalent to the mean residence time of 231 Pa in the water column (~ 200 years), so that nearly half of the 231 Pa produced in Atlantic water is exported to the southern ocean with the water in which it formed. On the other hand, with its much shorter residence time (~ 30 years), nearly all of the 230 Th produced in this water is removed into the sediments of the Atlantic and little is exported to the Southern Ocean. As a result, the 231 Pa/ 230 Th ratio in Atlantic sediments is, on average, about half the production rate ratio (0.092 dpm/dpm) of these two isotopes in the water column. Faster rates of overturning export a larger fraction of 231 Pa and further decrease 231 Pa/ 230 Th, while slower rates of overturning increase this ratio (Marchal et al., 2000; Siddal et al., 2007). Application of this simple principle is, however, complicated by two factors. First, sedimentary 231 Pa/ 230 Th is not only controlled by the rate of overturning, but also by the removal rate of the two isotopes from the water column by particle scavenging. Scavenging rates, which are controlled by the flux and composition of settling particles  22  (Bacon, 1988; Walter et al., 1997; Chase et al., 2002; 2003), dictate the residence time of 231 Pa in seawater and the extent to which it can be exported from the Atlantic by the AMOC (Yu et al., 1996). On the other hand, 230 Th has a residence time sufficiently short to severely limit its redistribution by circulation and mixing, even when the rate of overturning is fast (Francois et al., 2004). It is possible to assess the impact of changes in particle scavenging by analyzing the composition of the sediment, which informs us on changes in particle flux and composition at the site of study and their possible overprint on sediment 231 Pa/ 230 Th at this location (Gherardi et al., 2009), at least to the extent that we can take into account the effect of diagenesis. However, the extent to which scavenging can also affect sediment 231 Pa/ 230Th further “downstream” in the overturning circulation cell still needs to be investigated. The second point of contention is the extent to which sediment 231 Pa/ 230 Th integrates circulation rates over the overlying water column. In a recent study using a 1-D scavenging model, Thomas et al. (2006) have argued that sedimentary 231 Pa/ 230 Th may only record overturning occurring in about 1000m of water overlying the analyzed sediment and shallower overturning cannot be recorded in deep sediments. Several studies have investigated the distribution of 231 Pa and 230 Th in the ocean using three dimensional circulation models based on simplified dynamics (Henderson et al., 1999; Siddall et al., 2005; 2007) or the primitive equations (Dutay et al., 2009). In this study, we take a very different approach and develop a simple 2-D scavenging model to establish the patterns of 231 Pa/ 230 Th distribution that can be generated by an ascribed overturning circulation. The results provide possible explanations for some of the existing  23  field observations in the water column and sediments and a baseline for further evaluating the influence of the other factors that affect the distribution of 231 Pa/ 230 Th in the real ocean. They also suggest sampling strategies to maximize the information on paleocirculation that could be obtained from a very limited sediment database.  2.2 Model descriptions Water column profiles of dissolved and particulate 230 Th and 231 Pa concentration indicate that these two isotopes are removed from seawater by reversible scavenging (Bacon and Anderson, 1982; Nozaki et al., 1987). We use the same formalism to describe scavenging imbedded in a 2-D circulation scheme to investigate how the concentration of 231 Pa and 230 Th in the water column and sediments can potentially be affected by changes in circulation and scavenging rate. 2.2.1 Formulation The scavenging model used for both 231 Pa and 230 Th is shown in figure 2.1. Dissolved 230 Th and 231 Pa concentrations ([X]d; where X represents 230 Th or 231 Pa) are controlled by the production rates of the respective nuclides (PX; dpm.m -3 .y -1 ), their adsorption (K1 X ) and desorption (K-1 X ) rate constants (y -1 ), and the transport rates imposed by the circulation scheme (V; m.y -1 ), while 230 Th and 231 Pa particulate concentrations ([X]p) are controlled by the adsorption/desorption rate constants, transport rates and the sinking rates  24  (S; m.y -1 ) of the particles that scavenge the two nuclides from the water column. At steady-state, we can write: Px - K1x[X]d + K-1x[X]p + VΔ[X]d = 0        (2.1) K1x[X]d - K-1x[X]p + VΔ[X]p + dFlux/dZ = 0         (2.2) dFlux/dZ = S ([X]p (i+1)  - [X]p (i) )          (2.3) Where X represents 230 Th or 231 Pa, Z is water depth (m), i is the vertical index, and Δ is an “upwind” difference divided by the grid spacing (Press et al., 1992). The model uses a uniform grid with a horizontal grid spacing of 2.5 degrees latitude and a vertical grid spacing of 250 m.  25   Figure 2.1: The scavenging model consists of a meridional section (from 70°N to 70°S) evenly distributed into 56*20 grids (20 layers evenly distributed over 5000m depth and 56 columns evenly distributed over the Meridional section; 2.5 latitude per column). In each box, X represents 230 Th or 231 Pa. Xd’s = dissolved concentration (dpm/m 3 ). Xp’s = particulate concentration (dpm/m 3 ). Px = production rate from U decay (dpm/m 3 /y). S = sinking rates of particles (m/y). V = transport rates (m/y). The MATLAB code of the model is reported in Appendix A. These equations are used to calculate the concentration of dissolved and particulate 230 Th and 231 Pa as a function of depth and latitude. Using the upwind scheme with a horizontal velocity u = 5.3 10 -3  m/s and a horizontal grid spacing x = 278 103 m, the inherent  26  mixing in our model (Kdiff) is ~ 800 m 2 /s (Kdiff = u x/2; based on equivalence of the upwind scheme applied to an advective-reactive equation and an analytic diffusive-advective-reactive equation; e.g. Press et al., 1992). This is in the upper range of the along-isopycnal tracer diffusivities reported for the southern ocean (100-800 m 2 /s; Zika et al., 2009). Initial tests indicate that using a smaller grid size to decrease the model’s diffusivity does not result in significant differences in the model results. 2.2.2 Overturning Circulation The 2D meridional overturning circulation scheme (control run) used in this study is ascribed within a meridional section in the Atlantic Ocean (constant depth of 5000m from 70°N to 70°S) and based on the meridional overturning transports for the North Atlantic reported by Talley (2003). It consists of two meridional overturning cells flowing in opposite directions (Fig. 2.2). The Atlantic Meridional Overturning Circulation (AMOC) is initiated by the formation of 20.5 Sv of North Atlantic Deep Water (NADW) (Friedrichs and Hall, 1993; Macdonald, 1998; Talley et al., 2003) resulting from water flowing north in the upper 1500m of the water column and sinking between 60°N and 70°N. This latitudinal range coincides roughly to the latitudes where deep water forms in the Labrador and Nordic seas. The site of deep water formation (60°N ~ 70°N) is represented by one homogenized region between 250 to 4250 m depth to represent rapid deep water convection.  27   Figure 2.2: (a) Velocity vector plot. Size of the arrows is proportional to the transport rates used in the model. (b) Overturning fluxes in the model at the equator. Water from this homogeneous region is then transported horizontally to the south at different rates (Fig. 2.2b). The depth distribution of lateral transport was chosen so that the model generates dissolved 230 Th and 231 Pa profiles consistent with observations (see below). At 10°N, the NADW flow increases to 22.5 Sv with the addition of 2 Sv from the Antarctic Bottom Water (AABW) between 10°N and 35°N. Two Sverdrups of AABW are added further south, resulting in a total flow of 24.5 Sv of NADW, which is close to the NADW strength (23±3 Sv) estimated from the World Ocean Circulation Experiment (WOCE) data (Ganachaud and Wunsch, 2000). NADW starts to gradually upwell at 42.5°S towards a mixing zone (i.e. one homogeneous region) located above 1000m between 67.5°S and 57.5°S. Water from this mixing zone feeds surface and intermediate water forming the shallow return limb of the AMOC. 5 5 55 5 5 5 5 5 1 0 1010 1 0 1 0 10 10 1 0 1 5 1515 1 5 15 15 15 0 0 0 0 202020 20 20 20 -5 -5 -5 -5 -5 Latitude D e p th Atlantic Flux and Velocity Vector   60S 50S 40S 30S 20S 10S 0 10N 20N 30N 40N 50N 60N 4750 4250 3750 3250 2750 2250 1750 1250 750 250 -5 0 5 10 15 20 -4 -2 0 2 4 6 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 Flux(Sv) D e p th (m ) Flux at Equator  28  The second overturning cell is initiated by 8 Sv of AABW, originating from the same mixing cell, flowing into the southernmost region (67.5° - 70° S) and sinking directly to 3500 m (Fig. 2.2). Four Sverdrups are transported northward between 3500m and 4500m depth and entrained in the upwelling NADW south of 40°S. The remaining 4 Sv are transported northward below 4500m. This northward flow is gradually attenuated by entrainment to the NADW and disappears at 37.5°N, which is roughly consistent with hydrographic observations (Sloyan and Rintoul, 2001). In this study, we do not specifically represent the Antarctic Intermediate Water (AAIW). Although the rate of AAIW formation may affect the 231 Pa/ 230 Th of sediment deposited at intermediate depths in the South Atlantic, preliminary model runs indicate that this water mass has little or no effect on the 231 Pa/ 230 Th of deep sea sediments. 2.2.3 Parameterization Among all the parameters needed to constrain the model shown in Figure 2.1, the production rates for 230 Th and 231 Pa are best known since they are essentially constant and only depend on the well-established concentration of 234 U and 235 U in seawater (Delanghe et al., 2002; Robinson et al., 2004). The other parameters (Table 2.1) are associated with greater variability and uncertainties. Sinking rate (S): Most estimates of the average sinking rate of fine particles (S) obtained from water column profiles of particulate 230 Th (e.g. Krishnaswami et al., 1981; Rutgers van der Loeff and Berger, 1993; Scholten et al., 1995; Moran et al., 2001) range between  29  400-800 m.y -1 . Since there are no clear indications of systematic variability in this parameter, we chose a uniform and intermediate value of 500 m.y -1 (Table 2.1) K1 Th  and K-1 Th : The adsorption (K1) and desorption (K-1) rate constants for 230 Th have been estimated using a reversible scavenging model (Bacon and Anderson, 1982; Nozaki et al., 1987; Clegg and Whitfield, 1991; Clegg et al., 1991) and mostly range from 0.2 to 0.8 y -1  for K1 Th  and 1 to 3 y -1  for K-1 Th . We chose values within this range (Table 2.1) which generate dissolved 230 Th profiles broadly consistent with water column profiles measured at several locations in the Atlantic and in the Southern Ocean (Fig. 2.3; Table 2.2). K1 Th  is lower in the Southern Ocean than in the Atlantic, consistent with the data of Chase et al. (2002). We also used higher K1 Th  in the upper 500m to reflect the increase in K1 Th  with particle concentrations (Bacon and Anderson, 1982). Table 2.1: List of abbreviations and values for the model parameters. Variables Symbol Control run Units 231 Pa production rate PPa 0.00246 dpm/(m3*yr) 230 Th production rate PTh 0.0267 dpm/(m3*yr) Particle sinking rate  S 500 m/yr 230 Th adsorption rate (70°N-50°S)  0-250m  K1 Th  1.0 1/yr  250-500m  K1 Th  0.75 1/yr           > 500m  K1 Th  0.5 1/yr 230 Th adsorption rate (50°S-70°S)           0-250m K1 Th  0.6 1/yr           250-500m K1 Th  0.45 1/yr           > 500m K1 Th  0.3 1/yr 230 Th desorption rate (70°N-70°S)  30  Variables Symbol Control run Units  All depths K-1 Th  1.6 1/yr 231 Pa adsorption rate (70°N-42.5°S)  0-250m  K1 Pa  0.08 1/yr  250-500m  K1 Pa  0.06 1/yr           > 500m  K1 Pa  0.04 1/yr 231 Pa adsorption rate (42.5°-45°S)  0-250m  K1 Pa  0.2 1/yr  250-500m  K1 Pa  0.15 1/yr           > 500m  K1 Pa  0.1 1/yr 231 Pa adsorption rate (45°-47.5°S)  0-250m  K1 Pa  0.3 1/yr  250-500m  K1 Pa  0.225 1/yr           > 500m  K1 Pa  0.15 1/yr 231 Pa adsorption rate (47.5°S-70°S)  0-250m  K1 Pa  0.44 1/yr  250-500m  K1 Pa  0.33 1/yr           > 500m  K1 Pa  0.22 1/yr 231 Pa desorption rate (70°N-70°S)  All depths K-1 Pa  1 1/yr  K1 Pa  and K-1 Pa : The adsorption and desorption rate constants for 231 Pa are even less constrained and we selected their values so as to obtain dissolved 231 Pa concentration profiles (Table 2.2) and fractionation factors (Table 2.3) that are also broadly consistent with observations in the field.  31   Figure 2.3: Station locations for the water column profiles used to constrain the parameters in the model. Nordic Seas: N-A, N-B, N-C (Moran et al., 1995; 1997; 2002); Western Atlantic: W-A to W-F  32  (Table 2.2); Eastern Atlantic: E-A to E-F (Table 2.2); Southern Ocean: S (Rutgers van der Loeff and Berger, 1993). The fractionation factor is defined as (Anderson et al., 1983): F = ([ 231 Pa]d/[ 230 Th]d)/([ 231 Pa]p/[ 230 Th]p)          (2.4) F has been directly measured in the Atlantic and southern ocean (Walter et al., 1997; Moran et al., 2001; Chase et al., 2002). Particle composition affects the fractionation factor (F) due to the stronger affinity of opal for 231 Pa. In carbonate dominated regions, F is much higher than in opal dominated regions, where F is close to 1. We have adjusted the adsorption and desorption rate constants with latitude (Table 2.1) to produce systematic variations in the “equilibrium” fractionation factor which broadly reflect the field observations (Moran et al., 2002; Walter et al., 1997; Table 2.3). Table 2.2: 230 Th and 231 Pa activities in sea water (dpm/1000kg). W-A: Station EN407-3 (39°28’N; 68°22’W) depth Diss. 230 Th Diss. 231 Pa m dpm/1000kg (± 95% CI) 250 -  - 0.062 ± 0.004 500 0.140 ± 0.005 0.110 ± 0.005 751 0.187 ± 0.007 0.147 ± 0.006 1001 0.204 ± 0.008 0.134 ± 0.006 1250 0.161 ± 0.006 0.092 ± 0.005 1501 0.265 ± 0.009 0.148 ± 0.009 1800 0.299 ± 0.008 0.166 ± 0.007 2200 0.338 ± 0.011 0.176 ± 0.007  33  depth Diss. 230 Th depth Diss. 231 Pa depth m dpm/1000kg (± 95% CI) m dpm/1000kg (± 95% CI) m dpm/1000kg (± 95% CI) m 2500 0.353 ± 0.011 0.181 ± 0.007 2750 0.329 ± 0.019 0.165 ± 0.011 2980 0.295 ± 0.009 0.140 ± 0.005  W-B: Station EN407-4 (38°36’N; 68°53’W) depth Diss. 230 Th Diss. 231 Pa m dpm/1000kg (± 95% CI) 50 0.044 ± 0.002 0.030 ± 0.004 200 0.090 ± 0.004 0.064 ± 0.005 400 0.116 ± 0.005 0.113 ± 0.007 600 0.238 ± 0.008 0.168 ± 0.009 800 0.235 ± 0.009 0.177 ± 0.006 1000 0.242 ± 0.009 0.152 ± 0.008 1200 0.206 ± 0.008 0.140 ± 0.006 1400 0.247 ± 0.008 0.147 ± 0.006 1600 0.280 ± 0.012 0.170 ± 0.005 1800 0.299 ± 0.010 0.155 ± 0.006 2000 0.303 ± 0.011 0.156 ± 0.007 2200 0.340 ± 0.011 0.193 ± 0.008 2400 0.335 ± 0.014 0.187 ± 0.009 2600 0.337 ± 0.014 0.182 ± 0.007 2800 0.321 ± 0.016 0.165 ± 0.007 3000 0.344 ± 0.015 0.156 ± 0.006 3200 0.276 ± 0.011 0.144 ± 0.007 3400 0.227 ± 0.013 0.125 ± 0.005 3470 0.196 ± 0.010 0.129 ± 0.006     34  W-C: Station KNR07-4 (01°34’N; 23°38’W) depth Diss. 230 Th Diss. 231 Pa m dpm/1000kg (± 95% CI) 50 0.054 ± 0.003 0.038 ± 0.007 400 0.130 ± 0.003 0.122 ± 0.009 800 0.218 ± 0.004 0.253 ± 0.017 1100 0.284 ± 0.005 0.310 ± 0.021 1500 0.372 ± 0.008 0.314 ± 0.013 1800 0.430 ± 0.009 0.290 ± 0.011 2100 0.461 ± 0.007 0.326 ± 0.031 2400 0.489 ± 0.007 0.323 ± 0.019 2700 0.490 ± 0.007 0.348 ± 0.020 3000 0.477 ± 0.007 0.314 ± 0.015 3400 0.502 ± 0.006 0.304 ± 0.017 3800 0.571 ± 0.008 0.264 ± 0.015  W-D: Station KNR07-3 (01°12’S; 25°29’W) depth Diss. 230 Th Diss. 231 Pa m dpm/1000kg (± 95% CI) 50 0.063 ± 0.003 0.051 ± 0.004 400 0.140 ± 0.006 0.128 ± 0.009 800 0.223 ± 0.005 0.234 ± 0.014 1000 0.274 ± 0.004 0.292 ± 0.012 1300 0.325 ± 0.007 0.291 ± 0.016 1500 0.377 ± 0.008 0.275 ± 0.013 2000 0.446 ± 0.009 0.268 ± 0.017 2500 0.449 ± 0.006 0.300 ± 0.016 3000 0.425 ± 0.006 0.283 ± 0.012 3500 0.470 ± 0.006 0.236 ± 0.012 4000 0.572 ± 0.014 0.214 ± 0.012 4500 0.800 ± 0.009 0.263 ± 0.013   35  W-E: Station KNR07-2 (03°44’S; 27°58’W) depth Diss.  230 Th Diss.  231 Pa m dpm/1000kg (± 95% CI) 50 0.051 ± 0.003 0.052 ± 0.011 300 0.174 ± 0.006 0.104 ± 0.010 900 0.271 ± 0.007 0.271 ± 0.012 1100 0.295 ± 0.008 0.308 ± 0.017 1600 0.431 ± 0.010 0.292 ± 0.017 2100 0.490 ± 0.012 0.278 ± 0.014 2600 0.532 ± 0.011 0.321 ± 0.015 3100 0.598 ± 0.019 0.291 ± 0.016 3600 0.594 ± 0.011 0.227 ± 0.014 4000 0.690 ± 0.011 0.229 ± 0.019 4400 0.848 ± 0.015 0.266 ± 0.014 5000 0.845 ± 0.018 0.262 ± 0.010  W-F: Station KNR07-1 (07°10’S; 31°15’W) depth Diss.  230 Th Diss.  231 Pa m dpm/1000kg (± 95% CI) 50 0.062 ± 0.003 0.036 ± 0.006 450 0.172 ± 0.008 0.200 ± 0.016 900 0.304 ± 0.010 0.330 ± 0.018 1350 0.389 ± 0.007 0.282 ± 0.015 1756 0.485 ± 0.009 0.306 ± 0.020 2250 0.492 ± 0.014 0.279 ± 0.011 3150 0.567 ± 0.017 0.278 ± 0.013 3556 0.568 ± 0.010 0.238 ± 0.010 4000 0.734 ± 0.011 0.262 ± 0.012 4456 0.908 ± 0.013 0.267 ± 0.016 5000 0.813 ± 0.013 0.254 ± 0.012    36  E-A: Station EN328-9 (45°32’N; 21°24’W) depth Diss.  230 Th Diss.  231 Pa m dpm/1000kg (± 95% CI) 50 0.030 ± 0.001 0.066 ± 0.004 400 0.139 ± 0.003 0.097 ± 0.007 800 0.194 ± 0.003 0.146 ± 0.008 1000 0.230 ± 0.004 0.168 ± 0.008 1500 0.250 ± 0.004 0.163 ± 0.009 2000 0.312 ± 0.005 0.169 ± 0.010 2500 0.282 ± 0.004 0.176 ± 0.011 3000 0.244 ± 0.009 0.198 ± 0.012 3500 0.272 ± 0.005 0.269 ± 0.014 3827 0.332 ± 0.005 0.300 ± 0.014  E-B: Station EN328-7 (31°00’N; 31°02’W) depth Diss.  230 Th Diss.  231 Pa m dpm/1000kg (± 95% CI) 50 0.058 ± 0.002 0.069 ± 0.007 400 0.151 ± 0.003 0.078 ± 0.005 800 0.193 ± 0.004 0.147 ± 0.006 1000 0.279 ± 0.006 0.205 ± 0.007 1500 0.361 ± 0.007 0.237 ± 0.011 2000 0.429 ± 0.007 0.274 ± 0.014 2500 0.504 ± 0.008 0.283 ± 0.012 3000 0.651 ± 0.010 0.323 ± 0.011 3500 0.795 ± 0.014 0.364 ± 0.014 4000 0.775 ± 0.012 0.365 ± 0.013 4375 0.767 ± 0.017 0.343 ± 0.010     37  E-C: Station EN328-4 (22°00’N; 36°31’W) depth Diss.  230 Th Diss.  231 Pa m dpm/1000kg (± 95% CI) 50 0.062 ± 0.002 0.058 ± 0.005 400 0.174 ± 0.003 0.096 ± 0.006 800 0.231 ± 0.004 0.190 ± 0.008 1000 0.275 ± 0.005 0.240 ± 0.011 1300 0.401 ± 0.007 0.289 ± 0.011 1500 0.453 ± 0.008 0.323 ± 0.013 2000 0.689 ± 0.013 0.414 ± 0.012 2500 0.814 ± 0.011 0.449 ± 0.014 3000 0.902 ± 0.012 0.423 ± 0.013 4000 0.983 ± 0.012 0.368 ± 0.013 4997 0.900 ± 0.009 0.292 ± 0.014 5506 0.843 ± 0.010 0.280 ± 0.010  E-D: Station KNR07-9 (12°56’N; 23°21’W) depth Diss.  230 Th Diss.  231 Pa m dpm/1000kg (± 95% CI) 50 0.061 ± 0.003 0.067 ± 0.011 450 0.142 ± 0.004 0.126 ± 0.017 900 0.233 ± 0.005 -  - 1300 0.334 ± 0.006 0.349 ± 0.031 1700 0.461 ± 0.008 0.429 ± 0.030 2100 0.561 ± 0.009 0.408 ± 0.028 2500 0.545 ± 0.008 0.415 ± 0.027 3000 0.679 ± 0.011 0.437 ± 0.026 3500 0.770 ± 0.009 0.446 ± 0.038 4000 0.699 ± 0.008 0.332 ± 0.013 4500 0.657 ± 0.012 0.260 ± 0.013 4700 0.590 ± 0.008 0.239 ± 0.010   38  E-E: Station KNR07-6 (10°04’N; 23°14’W) depth Diss.  230 Th Diss.  231 Pa m dpm/1000kg (± 95% CI) 50 0.088 ± 0.002 0.077 ± 0.007 450 0.164 ± 0.003 0.132 ± 0.014 900 0.228 ± 0.005 0.227 ± 0.013 1300 0.350 ± 0.008 0.324 ± 0.026 1700 0.466 ± 0.007 0.356 ± 0.021 2100 0.541 ± 0.008 0.352 ± 0.031 2500 0.620 ± 0.010 0.431 ± 0.028 3000 -  - 0.384 ± 0.027 3500 0.731 ± 0.010 0.366 ± 0.025 4000 0.660 ± 0.009 0.302 ± 0.016 4500 0.707 ± 0.011 0.266 ± 0.012 5000 -  - 0.240 ± 0.015  E-F: Station KNR07-5 (07°50’N; 24°37’W) depth Diss.  230 Th Diss.  231 Pa m dpm/1000kg (± 95% CI) 50 0.056 ± 0.002 0.057 ± 0.006 450 0.150 ± 0.004 0.086 ± 0.007 900 0.254 ± 0.006 0.274 ± 0.013 1300 0.369 ± 0.008 0.310 ± 0.020 1700 0.459 ± 0.008 0.298 ± 0.018 2100 0.554 ± 0.010 0.390 ± 0.020 2500 0.586 ± 0.013 0.350 ± 0.019 3000 0.671 ± 0.013 0.358 ± 0.019 3500 0.665 ± 0.011 0.359 ± 0.023 4000 0.640 ± 0.010 0.300 ± 0.023 4500 0.690 ± 0.010 0.273 ± 0.019 4700 0.702 ± 0.015 0.248 ± 0.017   39  Station KNR06-3 (29°32’S; 43°20’W) depth Total  230 Th Total  231 Pa m dpm/1000kg (± 95% CI) 12 0.096 ± 0.003 0.045 ± 0.006 401 0.197 ± 0.004 0.070 ± 0.008 797 0.316 ± 0.005 0.140 ± 0.008 1202 0.494 ± 0.006 0.260 ± 0.012 1600 0.572 ± 0.008 0.353 ± 0.016 1998 -  - 0.307 ± 0.015 2200 0.690 ± 0.009 0.311 ± 0.014 2400 0.692 ± 0.009 0.344 ± 0.020 2800 0.746 ± 0.010 0.376 ± 0.017 3197 -  - 0.308 ± 0.012 3598 0.961 ± 0.011 0.331 ± 0.014 3944 1.412 ± 0.014 0.326 ± 0.015  Table 2.3: “Equilibrium” Fractionation Factors. Latitude “Equilibrium Fractionation Factor” 70°N – 42.5°S 7.8 42.5°S-45°S 3.1 45°-47.5°S 2.1 47.5°S-50°S 1.4 50°S-70°S 0.9  The “equilibrium” fractionation factor is the fractionation factor that would be measured if particles were in equilibrium with surrounding seawater. In this case [X]p/[X]d = K1 X /K-1 X  and F = (K-1 Pa  K1 Th )/(K1 Pa K-1 Th ). As we will discuss below, however, F measured in the field is also affected by particle sinking rates and circulation. The “equilibrium”  40  fractionation factors used in our control run are set at 7.8 in all waters situated north of 42.5°S. Further south, they decrease gradually to reach a minimum of 0.9 south of 50°S. In order to calculate the transport rates (V; m.y -1 ) needed to obtain the desired water transport fluxes (Sv), we fixed the width of the Atlantic basin in our model at 3000 km.  2.3 Dissolved 230 Th and 231 Pa water column profiles: Data-Model comparison We used water column data (dissolved 230 Th and 231 Pa profiles; fractionation factors) to constrain the circulation and scavenging parameters in our model. Dissolved 230 Th and 231 Pa profiles from the North and Equatorial Atlantic (Table 2.2; Fig. 2.3) were measured following the ICP-MS isotope dilution method described by Choi et al. (2000). Samples were collected in 1998 (KNORR 159-7), 1999 (ENDEAVOR 328), and 2005 (ENDEAVOR 407). In this section, we present the fit between field data and those generated by our control run and discuss the processes that generate them. Simple scavenging models using constant S, K1 and K-1 and neglecting circulation predict a linear increase in dissolved and particulate 230 Th and 231 Pa concentrations versus depth (Bacon and Anderson, 1982; Bacon et al., 1985; Nozaki et al., 1987): [X]p = [PX / S] Z        (2.5) [X]d = [PX / K1] + [(K-1 PX)/(K1 S)] Z   (2.6) Where Z is depth.  41  However, most 230 Th and 231 Pa seawater profiles measured in the ocean display significant deviations from linearity because the effect of circulation can rarely be neglected.  The dissolved 230 Th and 231 Pa concentration profiles obtained with our model using the parameters listed in Table 2.1 also deviate often from linearity and are broadly consistent with observations.  Figure 2.4: Dissolved 230 Th and 231 Pa obtained with the control run at 60° - 70°N and measured in the Norwegian Sea (IOC93-13: Moran et al., 1995) and the Labrador Sea in 1993 (IOC93-2; Moran et al., 1997) and 1999 (Labrador: Moran et al., 2002). The model reproduces reasonably well the water column profiles measured in the Labrador and Norwegian Sea (Fig. 2.4). Shallow waters entering the Nordic Seas to 0 1000 2000 3000 4000 0.0 0.2 0.4 0.6 0.8 1.0 D ep th  ( m ) Diss. 230Th: dpm/1000kg Control run: 60°- 70°N IOC93-2: 54°30'N; 48°28'W Labrador: 58°11'N; 50°52'W IOC93-13: 64°48'N; 6°12'W No circulation 0 1000 2000 3000 4000 0.0 0.1 0.2 0.3 0.4 D ep th  ( m ) Diss. 231Pa: dpm/1000kg No circulation Labrador: 58°11'N; 50°52'W Control run: 60°-70°N  42  produce deep water have low 230 Th and 231 Pa concentrations and deep winter convective mixing results in low and nearly constant concentration profiles. Concentrations are higher at shallow depths and lower in deep waters than predicted by the scavenging model in the absence of vertical mixing. The fit of the modeled 230 Th is best with the profiles measured in the Labrador Sea in 1993 (Moran et al., 1997) and in the Norwegian Sea (Moran et al., 1995). The 230 Th concentrations measured in the Labrador Sea in 1999 are significantly higher and have been attributed to a temporary cessation of deep water convection in the Labrador Sea during that period (Moran et al., 2002). The build-up of 231 Pa resulting from the same effect is expected to be much smaller (the response time depends the residence time and is longer for 231 Pa; see below), and we find a reasonable fit between the model and the 1999 Labrador Sea measurements of dissolved 231 Pa (although the model generates somewhat higher concentrations than observed). The concentration deficit in deep waters generated in the Nordic and Labrador Seas spreads southward with the North Atlantic Deep Water. During transit to the Southern Ocean, the newly formed deep water is continuously subjected to the particle rain that originates from surface waters and which scavenges the 230 Th and 231 Pa continuously produced in the water column. When particles reach the depth of the newly formed deep water where the 230 Th and 231 Pa seawater concentrations are below steady-state concentrations dictated by scavenging, desorption from particles is enhanced and a fraction of the 230 Th and 231 Pa scavenged at shallower depths is released to the deep waters instead of being removed into the underlying sediments. Thus, the 230 Th and 231 Pa concentrations in newly formed deep waters gradually increase during transit to the  43  southern ocean until the concentration at steady state with respect to scavenging is regained, at which point the water column profiles have relaxed back to linearity (Francois, 2007). The gradual relaxation of the profiles to linearity has been described for each isopycnal by adding a lateral transport term to the 1-D scavenging model, as first proposed by Rutgers van der Loeff and Berger (1993): X]t/t = PX – S (K[X]t)/Z +  ( i [X]t – [X]t)/w = 0    (2.7) Where i [X]t and [X]t are total 230 Th or 231 Pa concentration measured at two locations on the same isopycnal with i [X]t the concentration in the upstream source region and w is the “transit time” of water between these two sites. In deep waters, K (= [X]p/[X]t) is nearly constant. Integrating the above equation thus gives:  [X]t  (PX w + i [X]t) (1 – e -Z/wSK )             (2.8) In the absence of circulation or mixing, and assuming a constant K, the reversible scavenging model predicts that ss [X]t  PX Z / SK, where ss [X]t is the total concentration of 230 Th or 231 Pa in seawater at steady state with respect to scavenging. Z/SK is thus the residence time with respect to addition by uranium decay and removal by scavenging when the profile has regained linearity (i.e. when it has regained steady state with respect scavenging), defined as ss = ss [X]t / PX. Therefore, ss = Z / SK and:  [X]t  (PX w + i [X]t) (1 – e -ss/w )               (2.9)  44  Equation (2.9) predicts that the radioisotope profiles relax back to linearity more slowly with increasing ss and therefore water depth. Profile linearity is thus regained closer to the source at shallower depths, and 230 Th regains linearity faster than 231 Pa because of its shorter ss. The shapes of the 230 Th profiles measured in the Atlantic are in agreement with this simple conceptual model and are also reproduced in the control run (Figs. 2.5 and 2.6). The seawater data show clearly the gradual southward relaxation of the profiles towards linearity. Linearity is regained faster for 230 Th and at shallower depths. We also note that the profiles from the western Atlantic display a greater deficit farther south, reflecting the stronger ventilation of the western Atlantic basins. The profiles obtained from the model (Figs. 2.5c and 2.6c) show similar trends with dissolved 230 Th and 231 Pa concentrations close to those observed in the ocean.  0 1000 2000 3000 4000 5000 0.0 0.2 0.4 0.6 0.8 1.0 1.2 D ep th  ( m ) 230Th - dpm/1000kg West EN407-3: 39°28'N; 68°22'W EN407-4: 38°36'N; 68°53'W KNR07-3: 1°12'S; 25°29'W KNR07-4: 1°34'N; 23°38'W KNR07-2: 3°44'S; 27°58'W KNR07-1: 7°10'S; 31°15'W No circulation 0 1000 2000 3000 4000 5000 0.0 0.2 0.4 0.6 0.8 1.0 1.2 D ep th  ( m ) 230Th - dpm/1000kg East EN328-9: 45°32'N; 21°24'W EN328-7: 31°00'N; 31°02'W EN328-4: 22°00'N; 36°31'W KNR07-9:12°56'N; 23°21'W KNR07-6:10°04'N; 23°14'W KNR07-5:7°50'N; 24°37'W No circulation  45   Figure 2.5: Dissolved 230 Th measured (a) in the western Atlantic, (b) in the eastern Atlantic, and (c) produced with the control run  0 1000 2000 3000 4000 5000 0.0 0.2 0.4 0.6 0.8 1.0 1.2 D ep th  ( m ) 230Th - dpm/1000kg Model 56N 51N 44N 34N 6S No circulation 0 1000 2000 3000 4000 5000 0.0 0.2 0.4 0.6 0.8 D ep th  ( m ) 231Pa - dpm/1000kg West EN407-3: 39°28'N; 68°22'W EN407-4: 38°36'N; 68°53'W KNR07-3: 1°12'S; 25°29'W KNR07-4: 1°34'N; 23°38'W KNR07-2: 3°44'S; 27°58'W KNR07-1: 7°10'S; 31°15'W No circulation 0 1000 2000 3000 4000 5000 0.0 0.2 0.4 0.6 0.8 1.0 D ep th  ( m ) 231Pa - dpm/1000kg East EN328-9: 45°32'N; 21°24'W EN328-7: 31°00'N; 31°02'W EN328-4: 22°00'N; 36°31'W KNR07-9:12°56'N; 23°21'W KNR07-6:10°04'N; 23°14'W KNR07-5:7°50'N; 24°37'W No circulation  46   Figure 2.6: Dissolved 231 Pa measured (a) in the western Atlantic, (b) in the eastern Atlantic and (c) produced with the control run. Further south, where the deep waters start to upwell, their relatively high 230 Th concentrations exceed the concentrations predicted by the scavenging model in absence of circulation, resulting in convex dissolved profiles (Francois, 2007). This is again clearly seen in measured seawater profiles (Rutgers van der Loeff and Berger, 1993) and model results (Fig. 2.7a). This trend is less apparent for 231 Pa (Fig. 2.7b) because of its slower response time, preventing the 231 Pa profiles from regaining linearity before reaching the Southern Ocean. 0 1000 2000 3000 4000 5000 0.0 0.2 0.4 0.6 0.8 D ep th  ( m ) 231Pa - dpm/1000kg Model 56N 51N 44N 34N 6S No circulation  47   Figure 2.7: Concentration profiles of dissolved 230 Th and 231 Pa measured (Rutgers van der Loeff and Berger, 1993) and modeled (control run) in the southern ocean  2.4 Fractionation factors: Data-Model comparison Fractionation factors (F) are most often obtained by measuring dissolved and particulate 231 Pa and 230 Th concentrations in the same seawater sample and applying equation 4 (e.g. Anderson et al., 1983; Walter et al., 1997; Moran et al., 2002). These measured values are generally viewed as being mostly controlled by particle composition, with opal having a much lower F than the other major constituents of marine particles (Chase et al., 2002; Guo et al., 2002; Geibert and Usbeck, 2004). Our model reflects the generally accepted view that F is much lower in the opal dominated Southern Ocean than in the 0 1000 2000 3000 4000 5000 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 230 Th - dpm/1000kg m ANT VIII/3 1785 (55° 06'S; 27°51'W) 64S-Model No circulation 0 1000 2000 3000 4000 5000 0.0 0.2 0.4 0.6 0.8 231 Pa - dpm/1000kg m ANT VIII/3 1785 (55° 06'S; 27°51'W) 64S-Model No circulation  48  carbonate-dominated Atlantic Ocean and we chose adsorption and desorption rate constants to generate “equilibrium” fractionation factors broadly consistent with field observations (Table 2.3). The fractionation factor generated by the model using equation (4) indicates, however, that F is also significantly affected by the sinking rate of particles and ocean circulation (Fig. 2.8). This is because the chemical equilibrium between particles and seawater cannot be reached when particles sink through vertical dissolved 230 Th and 231 Pa concentration gradients. In the absence of circulation, we can rearrange equations 2.4 – 2.6 to show that:  K1 Th  (S + K-1 Pa  Z) F = -----------------------               (2.10)  K1 Pa  (S + K-1 Th  Z) If the two nuclides have identical desorption rate constants, F would be independent of sinking rates in the absence of circulation. However, if K-1 Pa  < K-1 Th  and K1 Pa  < K1 Th  (Table 2.1) then F calculated with sinking particles rises well above equilibrium values and gradually decreases towards the equilibrium value with depth (Fig. 2.8a). When particles sink through the Atlantic Meridional Overturning cell, the fractionation factors estimated from equation 4 drop below the equilibrium fractionation factor within the core of the NADW (Fig. 2.8b). The fractionation factors measured in the field are therefore not directly comparable to those derived from equilibrium absorption experiments conducted in the laboratory (e.g. Geibert and Usbeck, 2004; Guo et al., 2002).  49    Figure 2.8: Distribution of fractionation factors: (a) obtained in our model with sinking particles but without circulation; (b) obtained with sinking particles in the control run with AMOC (NADW: 21.5Sv;  50  AABW: 8Sv). Note that sinking rates and circulation significantly affect the fractionation factor defined as ( 231 Pad/ 230 Thd)/( 231 Pap/ 230 Thp) (see text for explanation). In contrast to the F generated by our model, the fractionation factors measured by Moran et al. (2002) and Scholten et al. (2008) in the equatorial and southern Atlantic increase with depth down to ~ 1500 m and stay roughly constant or decrease further down. Also, the natural variability in F is much larger than the range observed in our model. The reason for this discrepancy could be depth variation in particle composition, a factor that is not taken into account in our model. Scholten et al. (2008) invoke a drop in the opal content of particles to explain the trend. However, the few available data on the composition of suspended particles (in the Sargasso Sea, Sherrell and Boyle, 1992; and in the North Pacific, Sherrell et al., 1998) do not show a clear trend with depth (except for one profile taken in spring 1991 in the North Pacific). Clearly, more data are needed before adding this variable in any model and this discrepancy must be left unresolved for now.  2.5 230 Th and 231 Pa distribution in the control run: Since our control run is broadly consistent with the limited water column data that are available, we can discuss the general distribution of 230 Th and 231 Pa generated by the model with some level of confidence.   51  2.5.1 Dissolved 230 Th: The model clearly generates the downward penetration of low dissolved 230 Th by deep convection in the high northern Atlantic (Fig. 2.9). The horizontal isolines between 20°N and 30°S indicate however that the vertical dissolved 230 Th profiles quickly regain linearity, as is observed in the field data. South of 30°S, dissolved 230 Th concentrations start to increase at all depths as a result of deep water upwelling (Fig. 2.2). The increase in dissolved 230 Th concentration is enhanced south of 50°S by the lower adsorption rate constants imposed in this region to reflect the dominance of biogenic silica (Table 2.1), while the formation of AABW results in dissolved 230 Th maxima at intermediate depths, similar to observations (Fig. 2.7a).  Figure 2.9: Dissolved 230 Th section generated by the model  52  In surface water, dissolved 230 Th concentration is significantly higher in the Southern Ocean, as has been noted in field data (Rutgers van der Loeff and Berger, 1993; Walter et al., 2001; Chase et al., 2003). 2.5.2 Particulate 230 Th: The pattern of distribution of particulate 230 Th concentration (Fig. 2.10) is similar to that of dissolved 230 Th. There is a conspicuous maximum in particulate 230 Th just north of the southern opal belt, which is a result of the increase in dissolved 230 Th concentration resulting from deep water upwelling. The sharp drop in particulate 230 Th further south is a direct consequence of the lower K1 Th  in the southern ocean.  Figure 2.10: Particulate 230 Th section generated by the model  53  Total 230 Th concentration profiles measured in the western (Table 2.2; Fig. 2.11) and eastern (Scholten et al., 2008) South Atlantic display a near bottom maximum similar to that generated by the model. This result may help explain the presence of a near-bottom maximum in total 230 Th when there is no clear evidence for the presence of a nepheloid layer (Scholten et al., 2008). However, our model generates these near-bottom maxima further south than observed, suggesting that a better representation of the AMOC in our model may require that the shoaling of the deep limb of the overturning cell starts further to the north.  Figure 2.11: Total 230 Th concentration measured in the southwestern Atlantic at Station KNR06-3 (Table 2.2) showing a near-bottom maximum similar to that generated in the South Atlantic by our model.  54  The fraction of total 230 Th in particulate form generated by the model ([ 230 Th]p/ [ 230 Th]t) ranges from 0.18 to 0.22 at low latitudes (Fig. 2.12), which is also conforming to field observations (Bacon and Anderson, 1982; Moran et al., 2002). In the Southern Ocean below 1000m, the model produces somewhat lower fractions in particulate form (0.16-0.18), reflecting the lower affinity of biogenic silica for Th (Table 2.1). Somewhat higher fractions are generated in the upper water column of the Southern Ocean (0.22) and in the Nordic Sea (0.28) reflecting deep convection and longer residence time of particles in these waters.  Figure 2.12: Fraction of total 230 Th associated with particles generated by the model  55  2.5.3 Dissolved 231 Pa: As for dissolved 230 Th, the model produces a clear downward penetration of low dissolved 231 Pa by deep convection at high northern latitudes (Fig. 2.13). However, following expectations and observations, the minimum associated with the core of the NADW propagates much further south, reaching the southern ocean. South of 30°S, dissolved 231 Pa concentrations start to increase as a result of deep water upwelling but the effect is not as pronounced as for 230 Th because of the higher adsorption rate constants imposed in the southern ocean to reflect the dominance of biogenic silica (Table 2.1).  Figure 2.13: Dissolved 231 Pa section generated by the model  56  Surface water dissolved 231 Pa concentrations are significantly higher in the southern ocean, but the effect is less pronounced than for 230 Th because of the higher scavenging rate of 231 Pa. 2.5.4 Particulate 231 Pa: The most prominent feature in the distribution of particulate 231 Pa concentration is the concentration maximum in the southern ocean (Fig. 2.14), resulting from the higher K1 Pa  used in this region. The fraction of particulate 231 Pa generated by the model north of 45°S remains uniform between 0.04 and 0.05 (Fig. 2.15), in general agreement with observations (Moran et al., 2002), while the higher values generated in the southern ocean (0.16 – 0.20) are consistent with some of the extreme values reported by Rutgers van der Loeff and Berger (1993). Profiles of total 231 Pa generated in the south Atlantic in the model are intermediate between measurements made in the western and eastern side of the basin (Fig. 2.16). The lower concentrations measured in the western Atlantic than in the eastern Atlantic suggest that the Deep Western Boundary Current rather than boundary scavenging plays a major role in controlling the distribution of 231 Pa in the water column of this region.  57   Figure 2.14: Particulate 231 Pa section generated by the model  Figure 2.15: Fraction of total 231 Pa associated with particles generated by the model  58   Figure 2.16: Total 230 Th concentration measured in the southwestern Atlantic at Station KNR06-3 (Table 2.2) and in the eastern South Atlantic (Scholten et al., 2008) compared to the model results at 40 ºS. 2.5.5 Dissolved 231 Pa/ 230 Th: Modeled dissolved 231 Pa/ 230 Th ratios systematically decrease with water depth in the North and Equatorial Atlantic, while this trend is less pronounced in the South Atlantic (Fig. 2.17). Data from nine of the North and equatorial Atlantic stations presented in Table 2.2 reflect this trend with a clear decrease in dissolved 231 Pa/ 230 Th with depth below 500m (Fig. 2.18). In shallower water, dissolved 231 Pa/ 230 Th is more variable. This may be a  59  result of the short residence times of 230 Th and 231 Pa at these shallow depths and their limited lateral transport. Shallow dissolved 231 Pa/ 230 Th are likely to be more affected by local changes in particle composition. The lack of a clear trend with depth below 1000 m generated by the model in the South Atlantic is consistent with the observations of Scholten et al. (2008).  Figure 2.17: Dissolved 231 Pa/ 230 Th section generated by the model The model also predicts that the highest ratios would be found in the surface water of the South Atlantic (Fig. 2.17). Walter et al. (2001) report an increasing trend in surface water dissolved 231 Pa/ 230 Th from 0.5 to 2.0 between 65°S and 40°S (their figure 2.4c). However, water column profiles from the South Atlantic available to date (Moran et al., 2002) fail to  60  document the predicted large ratios in surface water. High ratios are generated in our model because surface waters from the southern ocean with relatively high dissolved 230 Th and 231 Pa concentrations are advected north. Since 230 Th is more quickly removed by scavenging, dissolved 231 Pa/ 230 Th initially increases to eventually decrease farther north as the scavenging of 231Pa “catches up” with that of 230Th. Evidently, the complexity of surface water movement in the South Atlantic cannot be fully captured in our simple 2D model and these very high surface values may be artifacts of our simplified circulation. This question needs to be further explored with three dimensional models.  Figure 2.18: Dissolved 231Pa/ 230 Th profiles measured at 9 stations in the North and Equatorial Atlantic (Table 2.2).  61  2.5.6 Particulate 231 Pa/ 230 Th: The distribution of particulate 231 Pa/ 230 Th generated by the model is shown in Figure 2.19. We can take these values as representing the 231 Pa/ 230 Th that sediments would have if they were deposited at a given depth and latitude.  Figure 2.19: Particulate 231 Pa/ 230 Th section generated by the model However, as mentioned when discussing the fractionation factors, settling particles in our model are not in chemical equilibrium with surrounding waters. When particles reach the seafloor, they could possibly come into equilibrium with bottom waters. Whether they do or not depends on how long they are in contact with bottom waters before burial as a  62  result of sedimentation and bioturbation. With the rate constants used in our model, it would take 1-3 years (depending on initial conditions) for surface sediments to be within 95% of their equilibrium value with bottom waters. We can calculate sediment 231 Pa/ 230 Th at equilibrium with bottom waters using [X]p/[X]d = K1 X /K-1 X  for 230 Th and 231 Pa (Fig. 2.20).  Figure 2.20: Sediment 231 Pa/ 230 Th section generated by the model assuming that sediment reach equilibrium with bottom water (see text for explanations). Partial equilibration would result in sediment 231 Pa/ 230 Th intermediate between values reported in Figure 2.19 and 2.20. The difference is relatively small in deep water but significantly larger in shallower waters. This is consistent with the observation of  63  Scholten et al. (2008) who remarked that, at shallow depths, 231 Pa/ 230 Th in suspended particles are significantly lower than 231 Pa/ 230 Th in surface sediments (their figure 5) and suggest that surface sediments do reach equilibrium with bottom water.  2.6 Sediment 231 Pa/ 230 Th: Data-Model comparison In this section, we compare the distribution of sediment 231 Pa/ 230 Th generated by the model with 231 Pa/ 230 Th measured in Atlantic sediments as a test for further validation. The distribution of particulate (Fig. 2.19) and sediment (Fig. 2.20) 231 Pa/ 230 Th generated by the model is clearly controlled both by circulation and particle composition. We find the lowest values near the base of the two overturning cells just downstream of the sites of deep water formation and the highest values in the Southern Ocean. The low values are clearly generated by the overturning circulation cells, while the high values in the southern ocean are a direct consequence of particle composition. Sediment 231 Pa/ 230 Th generally decreases with depth, a pattern dictated by trends in dissolved 231 Pa/ 230 Th which is generated by the overturning circulation. A similar decreasing trend from ~ 0.13 at ~ 1000m to ~ 0.04 at ~ 5000m has been reported by Scholten et al. (2008) for surface sediments in the South Atlantic. Holocene 231 Pa/ 230 Th from the five North Atlantic cores discussed by Gherardi et al. (2009) also show a similar trend, with values approaching the production rate ratios for the two shallower cores and lower values for the three deeper cores (Table 2.4).   64  Table 2.4: Holocene Pa/Th in 5 North Atlantic cores (Gherardi et al., 2009) Core Position Water Depth (m) Holocene 231 Pa/ 230 Th DAPC2 58°58’N, 09°36’W 1709 0.093 ± 0.001 MD95-2037 37°05’N, 32°01’W 2150 0.093 ± 0.004 SU81-18 37°46’N, 10°11’W 3135 0.064 ± 0.005 SU90-44 50°01’N, 17°06’W 4279 0.052 ± 0.004 OCE326-GGC5 33°42’N, 57°35’W 4550 0.054 ± 0.004  Values reported for core tops from the Nordic Seas range from 0.07 to 0.09 (Yu et al., 1996). Our model generates these values with an equilibrum fractionation factor of 7.8, somewhat higher than the fractionation factors measured in the Labrador Sea (3-7; Moran et al., 2002). Significantly higher sediment 231 Pa/ 230 Th have been reported, however, just south of Iceland and the Denmark Strait (0.10-0.15; Yu et al., 1996; Anderson, pers. comm.) but they are generally found in sediments deposited between 1500m and 2000m water depth and seems confined to a relatively small area where Leinen et al. (1986) report opal concentration (carbonate-free wt %) of up to 20%. With the fractionation factors reported in Table 2.3 and Figure 2.8, our model generates sediment 231 Pa/ 230 Th below the production rate ratio at this depth range just south of the site of deep water formation (Fig. 2.20). The model generates the high values reported in this region only if we lower the equilibrium fractionation factor to 3.9 (Fig. 2.21).  65    Figure 2.21: (a) Sediment 231 Pa/ 230 Th generated with an opal belt just south of the site of deep water formation. (b) Differences in the sediment 231 Pa/ 230 Th field generated in with and without the northern opal belt.  66  2.7 Discussion 2.7.1 The effect of AMOC on sediment 231 Pa/ 230 Th In the absence of any circulation, the model generates a field of constant sediment 231 Pa/ 230 Th equal to the production rate ratio (0.092). In this case, changes in the fractionation factor (Table 2.3) produce changes in the dissolved 230 Th and 231 Pa fields but not in the particulate fields. The distribution of particulate and sediment 231 Pa/ 230 Th reported in Figure 2.19 and 2.20 should thus provide information on the ocean overturning circulation. The model results clearly indicate, however, that the relationship between sediment 231 Pa/ 230 Th at any given site and the overturning circulation is very complex, as was also noted by Siddall et al. (2007). Sediment 231 Pa/ 230 Th depends not only on the rate of the overturning and particle scavenging, but also on the detailed geometry of the overturning cell and the distance between the coring site and the site of deep water formation. Sediment 231 Pa/ 230 Th reaches a minimum at a depth dictated by the geometry of the overturning cell and at a latitude dictated by the position of the site of deep water formation and the strength of the overturning circulation (Fig. 2.22). Clearly, it is impossible to constrain the history of changes in the AMOC from the evolution of 231 Pa/ 230 Th at one site, as was attempted by McManus et al. (2004). 2.7.1.1 Vertical variations in sediment 231 Pa/ 230 Th induced by the AMOC The use of sediment 231 Pa/ 230 Th to reconstruct past changes in the AMOC relies on the longer residence time of 231 Pa in the water column. While the short residence time of  67  230 Th severely limits the extent to which it can be laterally transported after its production by uranium decay, the longer residence time of 231 Pa results in its extensive redistribution by ocean circulation. According to equation (2.9), the 231 Pa profiles relax back to linearity at a rate that decreases as ss (the residence time with respect to addition by uranium decay and removal by scavenging in the absence of circulation or mixing) increases.  ss is proportional to water depth. Therefore, if the rate of lateral volume transport were the same at all depths, the fraction of the 231 Pa production that is laterally transported with the water would increase with depth. This effect contributes to the general decrease with depth in dissolved (Fig. 2.17) and particulate (Fig. 2.19) 231 Pa/ 230 Th generated by the model and measured in sediments (Table 2.4). Very little 231 Pa can be laterally exported by circulation at shallow depths but an increasing fraction can be exported with increasing depth. Sediment 231 Pa/ 230 Th integrates the lateral export of 231 Pa over the entire overlying water column. The integration in terms of lateral volume transport, however, is not linear but weighted by ss. At similar rates, shallow overturning cells lower sediment 231 Pa/ 230 Th at the base of the cells less than deeper overturning cells. The relationship between changes in sediment 231 Pa/ 230 Th with depth and changes in lateral volume transport with depth is therefore complex and difficult to intuit. In our control run, sediment 231 Pa/ 230 Th reaches its lowest value at the depth where we find the highest rate of lateral volume transport (Figs. 2.22a, b), but, this is not always the case. For instance, if we use the zonally integrated overturning rates recently derived from the ECCO consortium dataset  68  (Wunsch and Heimbach, 2006), the lowest sediment 231 Pa/ 230 Th is reached 1000m below the depth of maximum lateral volume transport (Figs. 2.23a,b). 2.7.1.2 Horizontal variations in sediment 231 Pa/ 230 Th induced by the AMOC Sediment 231 Pa/ 230 Th also changes systematically with latitude or distance from the site of deep water formation. Latitudinal changes in sediment 231 Pa/ 230 Th at the depth where the minimum ratio is reached documents an initial decrease with distance from the site of deep water formation, followed by an increase (Fig. 2.22c). Dissolved 230 Th and 231 Pa concentrations are low throughout the water column at the site of deep water formation (Fig. 2.4). Because of its shorter ss, 230 Th concentration increases faster to reach its steady-state concentration with respect to scavenging (equation 2.9), thereby gradually decreasing dissolved, particulate and sediment 231 Pa/ 230 Th. Once dissolved 230 Th has reached its maximum value, the slower increase in dissolved 231 Pa results in a slow increase in 231 Pa/ 230 Th further downstream.  69   Figure 2.22: (a) Lateral velocity profile in the control run between 60ºN and 35ºN. (b) Vertical sediment 231 Pa/ 230 Th bathymetric profiles generated by the model at different rates of overturning and at the latitudes where lowest sediment 231 Pa/ 230 Th is found. (c) Latitudinal sediment 231 Pa/ 230 Th profiles for different rates of overturning at the depth where the lowest sediment 231 Pa/ 230 Th is found (3625m) (red symbols represent the lowest depth of maximum lateral velocity (a), minimum 231 Pa/ 230 Th (i.e. the depth for the latitudinal profiles) (b) and latitude of minimum 231 Pa/ 230 Th (i.e. the latitudes for the vertical profiles) (c))  70   Figure 2.23: (a) Contrasting lateral velocity profiles between the control run, the overturning profile from the ECCO consortium (14Sv, Wunsch and Heimbach, 2006) and an arbitrary shallower overturning cell (20.5 Sv). (b) Vertical sediment 231 Pa/ 230 Th generated by three overturning profiles at the latitude where the lowest 231 Pa/ 230 Th is found. (c) Latitudinal sediment 231 Pa/ 230 Th profiles generated by three overturning cells at the depth where the lowest sediment 231 Pa/ 230 Th is found (red symbols represent the lowest depth of maximum lateral velocity or the base of the shallow overturning cell (a), minimum 231 Pa/ 230 Th (i.e. the depth for the latitudinal profiles) (b) and latitude of minimum 231 Pa/ 230 Th (i.e. the latitudes for the vertical profiles) (c)).  71  2.7.1.3 Changes in sediment 231 Pa/ 230 Th resulting from changes in the rate of the AMOC Increasing the rate of overturning in the control run without changing the geometry of the overturning cell has several effects on the distribution of Atlantic sediment 231 Pa/ 230 Th: (1) It pushes the zone of minimum 231 Pa/ 230 Th farther away from the site of deep water formation (Fig. 2.22c); (2) The latitudinal minimum in sediment 231 Pa/ 230 Th does not decrease, but instead increases (Figs. 2.22b, c) (3) Sediment 231 Pa/ 230 Th also increases at the site of deep water formation and directly south of it (Fig. 2.22c); (4) The vertical gradient of sediment 231 Pa/ 230 Th at the latitude corresponding to the minimum sediment 231 Pa/ 230 Th increases (Fig. 2.22b); (5) the largest decrease in sediment 231 Pa/ 230 Th downstream of the deep water formation zone is found in the Southern and equatorial region (Fig. 2.22c). Even without changing the geometry of the overturning cell and particle scavenging, the same value of sediment 231 Pa/ 230 Th can be generated at one site by different rates of overturning. For instance, the same value of 0.052 is produced at latitude 36.25°N at 3635m with overturning rates of 10.25 Sv and 30.75 Sv (Fig. 2.22c). This observation reinforces the fact that sediment 231 Pa/ 230 Th at one site cannot uniquely constrain the rate of the AMOC. 2.7.1.4 Changes in sediment 231 Pa/ 230 Th resulting from changes in the geometry of the AMOC We find systematic changes in the distribution of sediment 231 Pa/ 230 Th when we impose a shallower overturning cell without changing the rate of overturning: (1) The depth of  72  minimum sediment 231 Pa/ 230 Th tends to shoal (Figs. 2.23a, b), although that might not be always the case; (2) The latitudinal gradient at the depth of minimum sediment 231 Pa/ 230 Th decreases (higher sediment 231 Pa/ 230 Th in the North Atlantic and lower sediment 231 Pa/ 230 Th in the South Atlantic, Fig. 2.23c) because 231 Pa has a shorter ss in shallower water and is less effectively exported horizontally; (3) Sediment 231 Pa/ 230 Th increases rapidly with depth below the base of the overturning cell (Fig. 2.23b), largely corroborating the finding of Thomas et al. (2006) that the sediment 231 Pa/ 230 Th signal generated by a shallow overturning circulation is, if not totally absent, at least strongly attenuated in sediments deposited more than 1000m below the base of the overturning cell. 2.7.1.5 Possible sampling strategy to constrain past changes in AMOC from sediment 231 Pa/ 230 Th These results suggest a possible sampling strategy to constrain past changes in the rate and geometry of the AMOC. A series of bathymetric profiles down the eastern and western slope of the North Atlantic, the Mid Ocean Ridge, or the flanks of seamounts, with due attention to possible changes in sediment composition, could document the vertical and horizontal sediment 231 Pa/ 230 Th gradients and the depth of minimum sediment 231 Pa/ 230 Th for different time slices. The shape of the vertical profiles would inform us on the geometry of the meridional overturning cells, while the gradients (horizontal and vertical) would provide constraints on the rate of the overturning. Figure 2.22 and Figure 2.23 also suggest that sediment 231 Pa/ 230 Th at the site of deep water formation may be  73  sensitive to the rate and depth of the AMOC. Whether these simple systematic trends can be reproduced in more complex circulation models, however, still needs to be verified. 2.7.2 The effect of AABW on sediment 231 Pa/ 230 Th Figure 2.20 clearly indicates that the overturning cell initiated in the Southern Ocean by the formation of AABW significantly contributes to lowering sediment 231 Pa/ 230 Th in the South Atlantic. If we eliminate the formation of AABW, sediment 231 Pa/ 230 Th in the South Atlantic significantly increases (Fig. 2.24). The process whereby AABW is producing these low sediment 231 Pa/ 230 Th is the same as for the northern overturning cell but the effect is found at greater depth and is less pronounced because of the smaller flow of water involved and the higher initial dissolved 230 Th and 231 Pa in the water that generates AABW. The low sediment 231 Pa/ 230 Th (< 0.05) in the deep Southeast Atlantic (Scholten et al., 2008) are consistent with the importance of AABW in generating low 231 Pa/ 230 Th in the South Atlantic and suggest that sedimentary records in this region, if unaffected by changes in opal flux, could generate important constraints on variations in the rate of formation of this important water mass (Negre et al., 2010).   74   Figure 2.24: Sediment 231 Pa/ 230 Th field generated in the control run without the formation of the AABW. 2.7.3 The effect of particle composition on sediment 231 Pa/ 230 Th As already indicated above, in the presence of circulation and/or mixing, localized changes in particle composition and fractionation factors produce dramatic but localized changes in sediment 231 Pa/ 230 Th (Fig. 2.21). Such changes can be taken into account by analyzing the opal content of the sediment from which the 231 Pa/ 230 Th record is obtained (Gherardi et al., 2009) with, however, one important caveat. Opal is undersaturated  75  throughout the ocean and much of it dissolves before burial. Below a certain threshold in opal flux and sediment mass accumulation rates, opal is not preserved in sediments but the 231 Pa/ 230 Th generated by the presence of opal in sinking particles could persist. We could further address this question by using a diagenetic model (e.g. Khalil et al., 2007) to estimate the opal concentration in sinking particles reaching the seafloor from sediment mass accumulation rates and use this information to estimate the range of possible fractionation factors to be applied at this site using the sediment trap data compilation of Chase et al. (2003). However, distinguishing between the importance of changes in circulation and opal flux will eventually be best addressed by generating a database large enough to obtain a near-synoptic view of the spatial distribution of sediment 231 Pa/ 230 Th for each time slice of interest, since the distribution generated by the overturning circulation is clearly distinct from the distribution generated by the distribution of opal productivity in the ocean. While Figure 2.21 clearly demonstrates the potential impact of localized variations in fractionation factors, it also shows, and maybe more importantly, that such changes in the North Atlantic have little impact on the 231 Pa/ 230 Th deposited downstream (231Pa/230Th < 0.002; Fig. 2.21b).  76   Figure 2.25: The influence of Southern Ocean fractionation factor on the sediment 231 Pa/ 230 Th in Atlantic sediments. (a) Lateral velocity field used to conduct the experiments. (b) Vertical sediment 231 Pa/ 230 Th gradients generated by the control run (1*FF) and when the Southern Ocean equilibrium fractionation factor (FF=0.9) is doubled (2*FF) or halved (0.5*FF). (c) Sediment 231 Pa/ 230 Th produced in the North Atlantic and the Southern Ocean under these three scenarios. This is however not the case when we change the fractionation factor in the Southern Ocean (Fig. 2.25). Doubling the equilibrium fractionation factor in the Southern ocean from 0.9 to 1.8 not only decreases sediment 231 Pa/ 230 Th in the Southern Ocean from ~ 0.3  77  to ~ 0.2 but also uniformly increases sediment 231 Pa/ 230 Th along the latitudinal transect of the Atlantic by nearly ~ 0.01. Reducing the fractionation factor increases the Southern Ocean 231 Pa sink and decreases 231 Pa/ 230 Th in the Atlantic. However, the slope of the latitudinal gradient of sediment 231 Pa/ 230 Th in the Atlantic is not significantly affected and could still be used to constrain the rate of the overturning. Nonetheless, accurately assessing the extent of the southern ocean 231 Pa sink will be important to evaluate the rate of AMOC.  2.8 Conclusions We have developed a simple 2D scavenging model to address some of the questions that have been raised concerning the use of sediment 231 Pa/ 230 Th as a paleocirculation tracer (Keigwin and Boyle, 2008; Scholten et al., 2008; Lippold et al., 2009). Although our circulation model is clearly too simple to capture all the complexity of ocean circulation, it reproduces many of the features observed in the distribution of dissolved 230 Th and 231 Pa and sediment 231 Pa/ 230 Th and provides a tool to start assessing the relative importance of circulation and particle scavenging in controlling the distribution pattern of sediment 231 Pa/ 230 Th in the Atlantic. The circulation scheme imposed in our model broadly reflects the flow of the main deep Atlantic water masses (NADW, AABW). The detailed geometry of the two overturning cells and the parameters of the imbedded scavenging model have been tuned to reproduce  78  the broad features of the distribution of dissolved 230 Th and 231 Pa and fractionation factors measured in the water column to date. The model produces a general decrease in dissolved, particulate and sediment 231 Pa/ 230 Th with depth, which is consistent with field observations (Fig. 2.18; Scholten et al., 2008; Gherardi et al., 2009). It also produces patterns in the distribution of sediment 231 Pa/ 230 Th which could be used to distinguish the circulation signal from the effect of particle scavenging. The model output also suggests sampling strategies to optimize the information in past circulation that could be derived from sediment 231 Pa/ 230 Th. The most robust circulation signals generated by the model are the vertical and horizontal sediment 231 Pa/ 230 Th gradients, which changes systematically with the rate and geometry of the AMOC (Figs. 2.22, 2.23). However, we still need to establish whether these diagnostic trends can also be produced with more complex 3D circulation models. We have used our 2D model to test the extent to which changes in fractionation factor can obliterate the patterns of sediment 231 Pa/ 230 Th generated by the overturning circulation. While it is clear that changes in particle composition in the North Atlantic can change sediment 231 Pa/ 230 Th locally, our model indicates that the 231 Pa/ 230 Th pattern generated by circulation further downstream is not significantly affected. This may be different for the Southern Ocean, which is the main sink for 231 Pa in our model. Changing the fractionation factor in the Southern Ocean offsets 231 Pa/ 230 Th but has little impact on the gradients below 1500m, and the information on the rate and geometry of the overturning circulation is still preserved.  79  Our 2-D model largely corroborates the results from the 1-D model of Thomas et al. (2006) and indicates that the sediment 231 Pa/ 230 Th signal is rapidly attenuated in sediment deposited below the base of the overturning cell. Finally, low sediment 231 Pa/ 230 Th in the South Atlantic (Scholten et al., 2008) appears to be due to the formation of AABW, which suggest that the 231 Pa/ 230 Th sedimentary record in this region, just north of the zone influenced by biogenic silica, could be used to constrain past changes in the rate of formation of this water mass (Negre et al., 2010).            80   Chapter 3 Reconstruction of the strength and geometry of the Glacial Atlantic Meridional Overturning Circulation using sediment 231Pa/230Th  3.1 Introduction The Atlantic Meridional Overturning Circulation (AMOC) has a major influence on Earth’s climate due to its role in the large scale redistribution of heat, CO2 and nutrients (Rahmstorf, 2002; Sigman et al., 2010). The Cd/Ca and δ13C records in benthic foraminifera indicate that the nutrient depleted North Atlantic Deep Water (NADW) was displaced upwards during the Last Glacial Maximum (LGM) by nutrient rich waters from the Southern Ocean. This shallower glacial NADW is called the Glacial North Atlantic Intermediate Water (GNAIW). These results are now well established (Boyle and Keigwin, 1987; Curry and Oppo, 2005; Marchitto and Broecker, 2006), but while they document changes in the geometry of the overturning circulation, they do not constrain its rate, and there is no consensus on the strength of the AMOC during the LGM (Kitoh et al., 2001; Lynch-Stieglitz et al., 2007; Burke et al., 2011; Hesse et al., 2011).  81  The ratio 231 Paex,0/ 230 Thex,0 (activity ratio of unsupported 231 Pa and 230 Th in sediments decay-corrected to the time of deposition; 231 Pa/ 230 Th hereafter), could potentially provide estimates of past rates of AMOC (Yu et al, 1996; Marchal et al., 2000). Pa-321 and Th-230 are produced in seawater from the radioactive decay of dissolved uranium and rapidly removed to the underlying sediment by adsorption on settling particles. Because 231 Pa is removed at a slower rate than 230 Th, a larger fraction of the 231 Pa produced in the Atlantic is exported to the Southern ocean by the AMOC. The average 231 Pa/ 230 Th of modern Atlantic sediments is thus lower than the production rate ratio of the two isotopes (0.092) and varies with the NADW formation rate (chapter 2). Applying this approach to several Atlantic cores revealed variations in sediment 231 Pa/ 230 Th during the last deglaciation, which was interpreted as reflecting large changes in the rate and geometry of the AMOC (McManus et al., 2004; Gherardi et al., 2005; 2009). Sediment 231 Pa/ 230 Th, however, is also affected by the preferential scavenging of 231 Pa by biogenic silica, which could lead to erroneous interpretation of changes in circulation (Keigwin and Boyle, 2008; Lippold et al., 2009). The role of biogenic opal in controlling 231 Pa/ 230 Th is evident in the Southern Ocean where settling particles have high opal concentrations, and periods of higher 231 Pa/ 230 Th at several locations in the Atlantic have been attributed to increased abundance of diatoms or opal (Hall et al., 2006; Keigwin et al., 2007; Bradtmiller et al., 2007; Lippold et al., 2009; Anderson et al., 2009). Recent interpretation of 231 Pa/ 230 Th in terms of paleocirculation considered this possible overprint by ascertaining that changes in 231 Pa/ 230 Th did not coincide with changes in the sediment content of biogenic silica (Gherardi et al., 2005; 2009; Gihou et al., 2010). However, this  82  argument is weakened by the fact that opal preserved in sediment is only a fraction of the settling particle flux that originally scavenged 231 Pa from the water column. Dissolution during early diagenesis could remove opal while leaving the scavenged 231 Pa in the sediment. Clearly, both the effect of circulation and opal flux must be taken into account to interpret sediment 231 Pa/ 230 Th, and we must identify the conditions under which one becomes dominant to interpret changes in this ratio. I developed a two-dimensional scavenging model to derive the sediment 231 Pa/ 230 Th distribution generated by simplified meridional overturning cells (chapter 2). The most striking trend generated by the model is a vertical gradient in sediment 231 Pa/ 230 Th, a trend which has now been documented in the sediments deposited in the western equatorial Atlantic (Lippold et al., 2010). This vertical trend is generated by the southward flow of NADW and the longer residence time of 231 Pa with respect to reversible scavenging compared to 230 Th (Francois et al., 2007; Gherardi et al., 2010; chapter 2). Building on these results, we have assembled a large 231 Pa/ 230 Th and opal sediment database (Appendix B) to establish the extent to which sediment 231 Pa/ 230 Th in the Atlantic reflects the vertical trend predicted by the 2D scavenging model, to identify areas where enhanced scavenging obscures the circulation signal, and to reconstruct the strength and geometry of the AMOC during the LGM.    83  3.2 Materials and methods 3.2.1 Core locations and chronology Sediment 231 Pa/ 230 Th from 51 Holocene and 47 last glacial maximum core sections are reported in Appendix B, along with core locations, water depths, sampling intervals, age information (Holocene 231 Pa/ 230 Th data are averaged over 0-7 ka and LGM data over 19-24 ka.), % opal, 230 Th-normalized opal fluxes, analytical errors and references. The cores are broadly distributed within the Atlantic (Fig. 3.1) to establish whether the basin-scale distribution patterns in sediment 231 Pa/ 230 Th expected from the impact of variations in the AMOC (chapter 2) can be broadly recognized in Atlantic sediments.  84   Figure 3.1: Core locations for Holocene and LGM 231 Pa/ 230 Th compilation identifiable by core number (first column) in Appendix B. Colour code indicates data available for the two time periods (black), the Holocene only (red), or the LGM only (blue). Open squares indicate cores influenced by opal (preserved opal flux > 0.2 g/cm 2  ka) and open triangles indicate cores affected by boundary scavenging. 3.2.2 Analytical Methods In addition to compiling literature data, additional cores were analyzed (Appendix B) using the following methods.  85  3.2.2.1 231 Pa/ 230 Th data: My colleague Joerg Lippold and Jeanne Gherardi measured the 231 Pa and 230 Th on all the new samples reported in Tables a, b, Appendix B. Their method is also included in Appendix B. 3.2.2.2 Biogenic silica: Biogenic silica was measured following the method described by Mortlock and Froelich, (1989) and Müller and Schneider, (1993). Samples were weighed (in the 20 mg range) in centrifuge tubes. Hydrogen peroxide (H2O2) and HCl were added to samples in order to remove organic and inorganic carbon, respectively. After removing the dissolved phase by centrifuging the samples, opal was extracted from the samples by adding Na2CO3 and heating to 85ºC in a water bath. Dissolved silica was measured colorimetrically at 812 nm on a LKB spectrophotometer. This method has a precision of about 10%. 3.2.3 Modeling To establish the extent to which the distribution of sediment 231 Pa/ 230 Th could be explained by lateral 231 Pa transport with the AMOC, the data (Appendix B) are compared with the output of two-dimensional scavenging models (chapter 2). The Holocene data were compared to the model output with a circulation scheme based on modern circulation as described in chapter 2. The circulation in the Holocene Atlantic presented here (Fig. 3.2) has been slightly modified from chapter 2 to add a representation of the  86  Antarctic Intermediate Water (AAIW). The scavenging parameters (Table 3.1) were established by fitting model output to dissolved 230 Th and 231 Pa profiles measured in the Atlantic (see chapter 2)  Figure 3.2: Holocene overturning scheme used for the model-data comparison shown in Fig. 3.8, NADW: 20.5 Sv, AABW: 8 Sv, AAIW: 10 Sv 5 5 55 5 5 5 5 5 1 0 1 0 1010 1 0 10 10 10 1 5 1515 1 5 15 15 15 0 0 0 0 0 202020 20 20 20-5 -5 -5 -5 -5 Latitude D e p th Atlantic Flux and Velocity Vector   60S 50S 40S 30S 20S 10S 0 10N 20N 30N 40N 50N 60N 4750 4250 3750 3250 2750 2250 1750 1250 750 250 -20 -15 -10 -5 0 5 10 15 20  87   Figure 3.3: LGM overturning scheme used for the model-data comparison shown in Fig. 3.8, GNAIW: 25 Sv, AABW: 4 Sv Table 3.1: List of abbreviations and values for the Holocene and LGM scavenging parameters (chapter 2). Higher K1 Pa  south of 47.5° S and between 55° and 60° N (Holocene only) represent the higher opal concentrations of particles settling in the Southern Ocean and in the Northern North Atlantic. The change in the position of the Southern opal belt during the LGM is represented by a northward shift of the southern region with higher K1 Pa . Lower K1 Th  at 50°S-55°S for the LGM accounts for the high percentage of opal in the settling material of this region (Asmus et al., 1999) Variables Symbol Holocene LGM Units 231 Pa production rate PPa 0.00246  same dpm/(m 3∙yr) 230 Th production rate PTh 0.0267  same dpm/(m 3∙yr) Particle sinking rate  S 500  same m/yr 230 Th adsorption rate (70°N-50°S) 0-250 m K1 Th 1.0 same 1/yr 5 5 5 5 5 5 5 5 0 0 0 0 0 0 0 0 10 10 10 1 0 10 10 1 0 1 5 15 15 1 5 15 15 1 5 2 0 20 20 2 0 20 20 -5 -5 -5 -5-1 0 -10 -1 0-1 5 -15 -1 5 25 25 252525 Latitude D e p th Atlantic Flux and Velocity Vector   60S 50S 40S 30S 20S 10S 0 10N 20N 30N 40N 50N 60N 4750 4250 3750 3250 2750 2250 1750 1250 750 250 -20 -15 -10 -5 0 5 10 15 20  88  Variables Symbol Holocene LGM Units 230 Th adsorption rate (50°S-55°S) 0-250 m K1 Th 1.0 0.6 1/yr 230 Th adsorption rate (55°S-70°S) 0-250 m K1 Th 0.6 same 1/yr 230 Th adsorption rate 250-500 m K1 Th 75% of 0-250 m value same 1/yr 230 Th adsorption rate > 500 m K1 Th 50% of 0-250 m value same 1/yr 230 Th desorption rate (70°N-70°S) all depths K-1 Th  1.6 same 1/yr 231 Pa adsorption rate (70°N-60°N) 0-250 m K1 Pa 0.08 same 1/yr 231 Pa adsorption rate (60°N-55°N) 0-250 m K1 Pa 0.16 0.08 1/yr 231 Pa adsorption rate (55°N-42.5°S) - K1 Pa 0.08 same 1/yr 231 Pa adsorption rate (42.5°S-45°S) 0-250 m K1 Pa 0.08 0.2 1/yr 231 Pa adsorption rate (45°S-47.5°S) 0-250 m K1 Pa 0.08 0.44 1/yr 231 Pa adsorption rate (47.5°S-50°S) 0-250 m K1 Pa 0.2 0.44 1/yr 231 Pa adsorption rate (50°S-70°S) - K1 Pa 0.44 same 1/yr 231 Pa adsorption rate 250-500 m K1 Pa 75% of 0-250 m value same 1/yr 231 Pa adsorption rate > 500 m K1 Pa 50% of 0-250 m value same 1/yr 231 Pa desorption rate (70°N-70°S) all depths K-1 Pa  1 same 1/yr  The overturning circulation scheme for the LGM (Fig. 3.3) was obtained by systematically changing its strength and geometry to optimize the fit with the LGM sediment data (Appendix B). The optimal LGM geometry consists of a shallower northern overturning cell reaching down to 3500 m depth in the North Atlantic and gradually shoaling to 2000 m in the South Atlantic. The LGM 2-D scavenging model also accounts for shifts in the position of the southern opal belt by 5 degrees to the north (Gersonde et al., 2003) and northern opal region was removed on account of sea ice cover (Table 3.1, Table 3.2). Fractionation factors were kept constant over the rest of the Atlantic Ocean (52.5°N – 40°S; Table 3.2). The position of the zone of deep water formation in the North Atlantic was also shifted to the south by 10 degrees (Labeyrie et al., 1992; Sarnthein et al., 2003)  89  Table 3.2: Latitudinal variations of the equilibrium Fractionation Factors (FF) used in the model. FF is the fractionation factor that would be measured if particles were in equilibrium with surrounding seawater and is calculated from the 2 3 1 Pa and 2 3 0 Th adsorption and desorption rate constant: FF = ( 231 Pa/ 230 Th)diss / ( 231 Pa/ 230 Th)part. Differences between Holocene and LGM account for shifts in the position of high biogenic opal flux regions. For the LGM, the southern opal belt was shifted ~5° to the north (Asmus et al., 1999; Sarnthein et al., 2003)  and the northern opal region was removed to account for the lower preserved opal fluxes due to sea ice cover. Fractionation factors were kept constant over the rest of the Atlantic Ocean (a reasonable assumption considering that the mean concentrations of opal in the Holocene and LGM sediment reported in Appendix B are similar: Holocene mean: 2.6 %, n=37, 1 SD=1.6 %; LGM mean=3.7 %, n=32, 1 SD=1.9 %). The position of the zone of deep water formation was also shifted to the south by 10 degrees during the LGM (Labeyrie et al., 1992). Latitude FF Holocene FF LGM 70°N – 60°N 7.8 same 60°N – 52.5°N 3.9 7.8 52.5°N – 40°S 7.8 same 40°S – 42.5°S 7.8 5.2 42.5°S – 45°S 7.8 3.1 45°S – 47.5°S 7.8 1.4 47.5°S – 50°S 3.1 1.4 50°S – 55°S 1.4 0.9 50°S – 70°S 0.9 same  One of the alternate geometries tested against the sediment database is shown in Fig. 3.4, with a northern overturning cell restricted to the upper 2000 m of the water column. This geometry was recently suggested based on a model – data comparison for benthic 13C (Hesse et al., 2011).  90   Figure 3.4: Overturning scheme used to test the very shallow (within the upper 2000 m of the water column) Glacial North Atlantic Intermediate Water (GNAIW) deduced by comparing data vs model results for benthic foraminifera  13 C (Hesse et al., 2011).  3.3 Results and Discussion 3.3.1 The influence of biogenic silica on the distribution of sediment 231 Pa/ 230 Th in Atlantic sediments When plotting 231 Pa/ 230 Th versus biogenic silica flux for Holocene and LGM core sections, a correlation is found only when including sites from the southern ocean and North Atlantic with biogenic opal flux exceeding 0.2 g/cm 2 ·ka (Fig. 3.5). Sedimentary 231 Pa/ 230 Th is also higher compared to other cores from similar water depths at the African margins (Scholten et al., 2008; Christl et al., 2009; Lippold et al., 2012) even though opal fluxes are relatively 0 0 0 0 000 0 5 5 5 555 5 1 0 10 10 1 0 1010 1 0 15 15 15 151515 -5 -5 -5 -5 -1 0 -10 -1 0 -1 5 -15 -1 5 20 20 202020 Latitude D e p th Atlantic Flux and Velocity Vector   60S 50S 40S 30S 20S 10S 0 10N 20N 30N 40N 50N 60N 4750 4250 3750 3250 2750 2250 1750 1250 750 250 -15 -10 -5 0 5 10 15  91  small, reflecting the effect of boundary scavenging (Anderson et al., 1983). However, the lack of correlation between opal flux and 231 Pa/ 230 Th for most of the Atlantic sites indicates that variations in scavenging intensity in most of the Atlantic Ocean, where opal fluxes are < 0.2 g/cm 2 ·ka, has a negligible impact on the distribution of  sediment 231 Pa/ 230 Th.  Figure 3.5: Correlation between sediment 231 Pa/ 230 Th and 230 Th-normalized opal flux is found during the Holocene (red) and LGM (blue) when including cores with high opal flux (black regression lines). The correlation disappears when only considering the main Atlantic basin (red, blue regression lines), where opal fluxes are lower than 0.2 g/cm 2  ka. Error bars indicate 2 SE for both opal flux and 231 Pa/ 230 Th. Significance of correlation is estimated by the linear correlation coefficient r and the p-value. Breaks in axes are inserted to display the Southern Ocean core.  92  A similar result is obtained when plotting sediment 231 Pa/ 230 Th vs % opal (Fig. 3.6), indicating that opal starts to significantly influence sediment 231 Pa/ 230 Th only when its concentration exceeds 8%. Since most Atlantic sediments have biogenic silica concentration below this threshold, the large scale distribution of sediment 231 Pa/ 230 Th in most of the Atlantic must be controlled by another factor.   93   Figure 3.6: Correlation between sediment 231 Pa/ 230 Th and opal concentration is found during the Holocene (a) and LGM (b) when including cores with opal concentration > 8% (black regression lines). The correlation disappears when only considering the main Atlantic basin (red, blue), where opal concentrations are lower. Error bars indicate 2 SE for both 231 Pa/ 230 Th and opal flux. 3.3.2 The influence of AMOC on the distribution of sediment 231 Pa/ 230 Th in Atlantic sediments 3.3.2.1 Holocene: The 231 Pa/ 230 Th ratios measured in most of the Holocene sediment samples clearly decrease with water depth down to 4000 m (Fig. 3.7a), consistent with the prediction of the 2D scavenging model (chapter 2). Below ~ 4000 m, the trend disappears or reverses, reflecting the influence of Antarctic Bottom Water (AABW).  94  Figure 3.7: 231 Pa/ 230 Th versus water depth. (a) Holocene sediment 231 Pa/ 230 Th generally decreases with water depth (red circles). (b) LGM 231 Pa/ 230 Th in the main Atlantic basin (blue squares) decreases with depth down to 2500 m and then increase in deeper waters. Values deviating from these general trends are from regions with high opal flux (open symbols) or from the African margin (black triangles). Note the break in the horizontal axis. Individual measurements have been averaged for both time periods (Appendix B). Error bars in 231 Pa/ 230 Th indicate 2 SE.  95  There are, however, several samples that deviate from this general trend and display higher ratios: samples immediately south of Greenland/Iceland, from the African margin, and the only sample from the southern ocean opal belt included in this study. At these sites, higher scavenging intensity by opal or particle flux increases 231 Pa/ 230 Th above the value expected from the overturning circulation alone. Figure 3.8: Correlation between Holocene (a) and LGM (b) sediment 231 Pa/ 230 Th and model outputs from grid cells closest to core locations. To further test the importance of the AMOC in controlling sediment 231 Pa/ 230 Th in the Atlantic and eventually to assess the strength of the glacial AMOC, we compare the Holocene 231 Pa/ 230 Th database to the output of the Holocene 2-D scavenging model (Fig. 3.8). In this model, we use 20.5 Sv NADW and 8 Sv AABW (Talley et al., 2003), and adsorption/desorption rate constants vary with latitude to reflect the presence of opal in the North Atlantic and the Southern Ocean (Table 3.1; 3.2). The model generates a decreasing  96  trend in sediment Pa/Th with depth (chapter 2), as observed in the data (Fig. 3.7a), and we find a good correlation between sediment data and model output generated at the closest locations with respect to latitude and depth (Fig. 3.8a). This result supports the interpretation that the distribution of 231 Pa/ 230 Th in Atlantic sediments deposited in regions with low opal flux can be largely attributed to the AMOC. The offset between data and model output is attributed to boundary scavenging, which is not included in the model. Boundary scavenging removes to the margins some of the 231 Pa produced in the water of the NADW as it transits through the Atlantic, accounting for the systematically lower 231 Pa/ 230 Th in the data compared to model output. 3.3.2.2 LGM: In contrast to the Holocene, LGM sediment 231 Pa/ 230 Th decreases with depth to 2500 m only and starts increasing below (Fig. 3.7b). As for the Holocene, sites south of Iceland and Greenland diverge for the LGM from the general trend but less so than during the Holocene, reflecting lower opal fluxes. Plotting 231 Pa/ 230 Th versus biogenic silica flux in glacial sediments (Fig. 3.5) reveals a weak but significant correlation with the inclusion of the Southern Ocean site. Three high 231 Pa/ 230 Th cores south of Iceland have high opal fluxes, suggesting an opal influence. However, two cores from the NW Atlantic with similar opal fluxes have low 231 Pa/ 230 Th, casting doubt that opal is the prevailing cause for these higher values. High 231 Pa/ 230 Th in the cores south of Iceland may also indicate a zone of deep water formation, where sediment 231 Pa/ 230 Th ratios are expected to be higher (chapter 2). Two cores taken in the upwelling region off Namibia also clearly deviate from the general trend due to boundary scavenging. As for the Holocene, there is no significant  97  correlation between preserved opal flux and sediment 231 Pa/ 230 Th where the flux of opal is below 0.2 g/cm 2 ·ka or where % opal is less than 8% (Fig. 3.6b). The change in the depth gradients in sediment 231 Pa/ 230 Th during the LGM (Fig. 3.7b vs Fig. 3.7a) is generally consistent with a shallower glacial AMOC driven by Glacial North Atlantic Intermediate Water (GNAIW). We tested several glacial AMOC geometries in the 2D model and found that the best fit with the data is obtained when the core of the glacial northern sourced AMOC is at 2000 m. However, to generate the comparatively low 231 Pa/ 230 Th at North Atlantic sites between 3000 and 3600 m depth (Fig. 3.7b) and produce a good fit between data and model output (Fig 3.8b), the base of the overturning cell must have reached down to this depth before gradually shoaling to the south (Fig. 3.3). Paleonutrient proxies suggest that GNAIW extended down to ~2500-3000 m depth only (Curry and Oppo, 2005; Marchitto and Broecker, 2006), and recently Hesse et al. (2011) inferred an even shallower glacial overturning by comparing the output of a 3-D circulation model and a δ13C database (Fig. 3.4). If our interpretation is correct, the low benthic δ13C of these cores must reflect high preformed nutrients in the deepest and densest waters involved in the glacial northern AMOC. The fit between sediment 231 Pa/ 230 Th generated by the optimal 2D glacial overturning geometry (Fig. 3.3) and the sediment database is shown in Fig. 3.8b. As for the Holocene, the offset from the 1:1 line is attributed to boundary scavenging, which is not represented in the model. This offset is less pronounced for the LGM, suggesting a weaker boundary scavenging effect at the African margin (Lippold et al., 2012).  98  3.3.3 Sensitivity test on the LGM 2-D scavenging model The sensitivity of the LGM model output to the scavenging parameters (sinking rate, fractionation factor), and to the strength and geometry of the overturning cells (GNAIW and AABW) was tested by comparing data vs model output, as shown in Fig. 3.8, under different configurations, and reporting the fit for each run either as the linear correlation coefficient (r) or the mean square weighed deviation (mswd, Powell et al., 2002) of the linear regression between data and model output. 3.3.3.1 Sensitivity to sinking rate and fractionation factor: Using the optimal geometry of the glacial AMOC (Fig.3.3), we systematically varied the sinking rate of particles and the fractionation factor in the main Atlantic basin, and compared sediment data with model outputs as done for Figure 3.8. Plotting the linear correlation coefficient (r; Fig. 3.9) and the mean square weighed deviation (mswd; Fig. 3.10) obtained when regressing the output for each model run against the sediment database indicates that the best fit is obtained with a sinking rate of 500m/y and a fractionation factor of 7.8. These parameters were thus used to assess the effect of overturning strength and geometry.  99   Figure 3.9: Variations in the linear correlation (r) obtained between the sediment 231 Pa/ 230 Th database (Appendix B) and model output using the overturning geometry and strength shown in Fig. 3.3 and systematically varying the sinking rate of particles and fractionation factor (FF).  Figure 3.10: Variations in the mean square weighed deviation (mswd) between the sediment 231 Pa/ 230 Th database (Appendix B) and model output using the overturning geometry and strength shown in Fig. 3.3 and systematically varying the sinking rate of particles and fractionation factor (FF). sensitivity test of model scavenging parameters 0.00 0.20 0.40 0.60 0.80 1.00 0 100 200 300 400 500 600 700 800 900 1000 particle sinking rate [m/year] r FF=3.4 FF=5.1 FF=7.8 FF=10.5 FF=15.6 sensitivity test of model scavenging parameters 0 50 100 150 200 250 0 100 200 300 400 500 600 700 800 900 1000 particle sinking rate [m/year] m sw d FF=3.4 FF=5.1 FF=7.8 FF=10.5 FF=15.6  100  3.3.3.2 Sensitivity to the strength of the overturning circulation cells (GNAIW; AABW) Using the optimal geometry of the glacial AMOC (Fig. 3.3) and the scavenging parameters determined above, we systematically varied the strength of the northern overturning cell, initiated by the formation of GNAIW, and the southern overturning cell, initiated by the formation of AABW, and quantified the fit between sediment data and model outputs using the mean square weighed deviation (mswd) of the regression.  Figure 3.11: Fit between observations and model outputs generated with the optimal LGM model geometry with varying GNAIW and AABW strengths. Agreement between measurements and model is  101  quantified using mean square weighted deviation (mswd). Smaller deviations are obtained with GNAIW > 20 Sv. The fit is less sensitive to variations in AABW. The model output with the Holocene circulation scheme yields a weaker agreement to the LGM observations (black square) further indicating that a stronger and shallower AMOC is necessary to explain the LGM measurements. The mswd between data and model increases sharply as soon as the rate of GNAIW drops below 22 Sv and is optimal for GNAIW rates ranging between 22 and 30 Sv (Fig. 3.11). The model-data fit provides a sharp lower limit for the rate of GNAIW (20-22 Sv), but the upper limit is less precisely constrained. This is because sediment 231 Pa/ 230 Th is more sensitive to changes in the rate of AMOC when it is weaker (Yu et al., 1996). When GNAIW drops below 22 Sv, the model is unable to generate the low sediment 231 Pa/ 230 Th observed in the glacial sediments deposited in the equatorial region. Considering that the cross section of the GNAIW overturning cell was smaller than today, these results imply a stronger mean flow of northern-sourced deep water during the LGM. In contrast, limits on the strength of the AABW could not be clearly established. Although a weak AABW overturning cell produces a slightly better agreement with observations (Fig. 3.11), the differences are very small. This may be because the available database is skewed to the northern and equatorial region of the Atlantic (Fig. 3.1) where the distribution of sediment 231 Pa/ 230 Th is dominated by the strength (and geometry) of GNAIW. Increasing the sediment database in the South Atlantic may improve our estimates of the strength of the AABW during the LGM.  102  3.3.3.3 Sensitivity to the geometry of the overturning circulation cells (GNAIW; AABW) The distribution of 231 Pa/ 230 Th in glacial Atlantic sediments confirms that the glacial AMOC was shallower but suggests a deeper overturning in the North Atlantic than inferred from nutrient proxies (Hesse et al., 2011).  Figure 3.12: Fit between observations and model outputs generated with a very narrow GNAIW (Fig. 3.4) as a function of GNAIW and AABW strengths. Agreement between measurements and model is quantified using mean square weighted deviation (mswd), which is invariably large compared to the optimal LGM circulation scheme.  103  To further test whether the glacial sediment 231 Pa/ 230 Th database is compatible with the consistently shallow overturning cell inferred from 13C, we calculated the msdw of the correlation between sediment data and model output using the geometry shown in Fig. 3.4, in which the penetration of GNAIW is restricted to the upper 2000m over the entire Atlantic and the core of the overturning cell is at 750 m. The results (Fig. 3.12) indicate a very poor fit, primarily because the low 231 Pa/ 230 Th values observed below 2000m in the glacial sediment database cannot be reproduced using this geometry. This confirms that a relatively deep overturning in the North Atlantic is needed to explain the observations and the discrepancy with 13C is best explained by invoking changes in preformed nutrients in the densest water of the GNAIW, maybe because of extensive ice cover in the zone where these denser waters formed. If this interpretation is correct, it could have important implications for our understanding of the marine carbon cycle and the estimation of CO2 uptake during the LGM (Kwon et al., 2012). The fit between observational data and model output is also shown on the sections reported in Fig. 3.13 showing sediment 231 Pa/ 230 Th sections generated by the three model geometries (Fig. 3.2 - Holocene circulation; Fig. 3.3 - optimal LGM circulation; and Fig. 3.4 - very shallow GNAIW) and superimposing on them the sediment database (plotted at a position defined by their depth and latitude) using the same colour scale. Fig. 3.13a clearly shows that the very shallow overturning cell derived from the distribution of 13C alone is unable to generate the low 231Pa/230Th in the sediment of the North Atlantic observed below 2000m, even if the GNAIW is increased to 30 Sv. On the other hand, Fig.  104  3.13b shows that the optimal LGM circulation scheme (Fig. 3.3) is able to produce these low values, as already indicated by the fit in Fig.3.8b. This figure also shows that if GNAIW is reduced, the model cannot generate the low 231 Pa/ 230 Th observed in the equatorial region. Finally, Fig. 3.13c shows that while the Holocene model (Fig. 3.2) reproduces well the Holocene database, it is not able to generate the distribution observed in the glacial core sections. In particular the Holocene model cannot reproduce the relatively high sediment 231 Pa/ 230 Th found in the deep North Atlantic during the LGM. This misfit is also shown on Fig. 3.11, where the mswd of the correlation between the LGM sediment database and the Holocene model input show a relatively poor fit compared to the optimal LGM model output. 3.3.4 Can the Holocene circulation scheme explain the LGM observations by varying the scavenging parameters? Thus far, we have shown that when using the scavenging parameters that best reproduce the LGM data with the optimal LGM circulation, the Holocene circulation scheme generates a distribution of sediment 231 Pa/ 230 Th that fits the LGM data base poorly (Fig. 3.11). The question remains, however, whether the Holocene circulation scheme could generate a 231 Pa/ 230 Th distribution similar to LGM observations with a different set of scavenging parameters.  105   Figure 3.13: Sediment 231 Pa/ 230 Th superimposed to the sediment 231 Pa/ 230 Th section generated by the 2-D scavenging model.  106  a  b  Figure 3.14: Linear correlation coefficient (a) and mean square weighed deviations (b) obtained when correlating the sediment database with sediment 231 Pa/ 230 Th generated by the Holocene circulation scheme with varying particle sinking rates and fractionation factors. -0.20 0.00 0.20 0.40 0.60 0.80 0 100 200 300 400 500 600 700 800 900 1000 particle sinking rate [m/year] lin ea r co rr el at io n co ef fi ci en t r FF=3.4 FF=5.1 FF=7.8 FF=10.5 FF=15.6 LGM FF=7.8 5 25 45 65 85 105 125 145 165 185 0 100 200 300 400 500 600 700 800 900 1000 particle sinking rate [m/year] m sw d FF=3.4 FF=5.1 FF=7.8 FF=10.5 FF=15.6 LGM FF=7.8  107  To test this possibility, we determined the linear correlation coefficient (r) and the mswd of the correlation obtained between the LGM dataset and the model output generated by the Holocene circulation scheme with varying sinking rates and fractionation factors. The results (Fig. 3.14) indicate a much poorer fit over a wide range of sinking rates and FF than can be obtained with the optimal LGM circulation scheme, reinforcing the conclusion that the LGM database reflects a change a circulation as illustrated in Fig. 3.3.  3.4 Conclusion The results from this study indicate that while the concentration of biogenic silica in particles can affect sediment 231 Pa/ 230 Th, its influence becomes significant only when the preserved opal flux exceeds 0.2g/cm 2 ka or when the opal concentration in sediment exceeds 8%. Below these threshold values, the distribution of sediment 231 Pa/ 230 Th in the Atlantic is controlled by the differential transport of the two isotopes by the AMOC. Since the opal content of most Atlantic sediments is below the threshold at which it controls 231 Pa/ 230 Th, this study confirms that past changes in Atlantic sediment 231 Pa/ 230 Th can be interpreted as changes in the strength and geometry of the AMOC, if the cores analysed do not include those from the African coastal upwelling regions and contain less than 8% opal. The main conclusions derived by comparing the extended database reported in Appendix B with output from our 2-D scavenging models are as follows.  108  (1) The GNAIW formation rate during the LGM was > 20Sv compared to 20Sv for the modern NADW. Because of its smaller cross section, the flow of water must have been faster and the glacial AMOC more vigorous during the LGM. However, while sediment 231 Pa/ 230 Th provides a sharp lower limit for the rate of GNAIW (20-22 Sv), the upper limit is less precisely constrained (~25-30 Sv). (2) The glacial AMOC was shallower than during the Holocene, but the GNAIW sunk to greater depth in the North Atlantic than suggested by 13C, which implies higher preformed nutrients in densest (deepest) waters of the GNAIW. The existing database is, however, unable to constrain changes in the rate of AABW. This is in part because of the relatively poor data coverage from the deep South Atlantic, which could be remedied by analysing additional cores from this region.         109   Chapter 4 The influence of deep water circulation on the distribution of 231Pa and 230Th in the water column and sediments of the Pacific Ocean  4.1 Introduction The ocean’s deep water circulation plays a prominent role in climate regulation. The overturning circulation of the Atlantic is a major contributor to heat transport to northern latitudes, while the relative isolation of deep water in the Pacific results in greater sequestration of carbon, contributing to lowering atmospheric CO2. It is therefore important to document past changes in ocean circulation in both oceans to understand climate evolution, especially on glacial-interglacial time scales and during abrupt climate changes. Past changes in ocean circulation have been inferred from nutrient proxies (e.g., Boyle and Keigwin, 1987), but poor carbonate preservation severely limits their application in the Pacific Ocean. In addition, although these tracers provide crucial information on changes in water mass distribution, they cannot constrain changes in the rate of deep water circulation (Legrand and Wunsch, 1995). In response to the latter  110  problem, attempts are being made to develop kinematic tracers of past ocean circulation (Lynch-Stieglitz et al., 2007). In particular, sedimentary 231 Pa/ 230 Th (activity ratios of 231 Pa and 230 Th not supported by U decay in the sediment mineral lattice and decay-corrected to the time of deposition, Pa/Th hereafter) has been used to evaluate past changes in the rate of the Atlantic Meridional Overturning circulation (AMOC) (Yu et al., 1996; Marchal et al., 2000, Mcmanus et al., 2004; Hall et al., 2006; Gherardi et al., 2005, 2009; Negre et al., 2010; chapter 3). This approach stems from the observation that Atlantic sediments have, on average, Pa/Th ratios lower than the fixed production rate ratio of 0.092 (Yu et al., 1996). Pa-231 and Th-230 are produced at constant rates in seawater from the decay of dissolved 235 U and 234 U, respectively. Pa-231 is less particle-reactive than Th-230 and has a longer residence time in the water column before removal to the sediment by scavenging. As a result, the AMOC exports 231 Pa into the southern ocean more effectively, which produces a 231 Pa deficit in Atlantic sediments controlled by the rate of overturning (Yu et al., 1996; Marchal et al., 2000; Siddall et al., 2007). This simple interpretation, however, can be obscured by changes in particle flux and opal content, which affect the residence time of 231 Pa in the water column and overprint sediment Pa/Th independently of circulation (Walter et al., 1997; Chase et al., 2002, 2003; Gil et al., 2009; Guo et al., 2002; Lippold et al., 2009). To further address this question, we recently developed a 2D scavenging model to investigate the influence of the overturning circulation on the distribution of Pa/Th in  111  Atlantic sediments (chapter 2). The model was tuned using vertical profiles of dissolved 231 Pa and 230 Th measured in the Atlantic and it predicts variations in Pa/Th that are recognized in the sediment, in particular, a decrease with depth which cannot be readily explained by invoking particle or opal flux as the main controlling factors (Gherardi et al., 2005; 2009; Lippold et al., 2011; chapter 3). Here, we use a similar approach to argue that deep water circulation must also be considered when interpreting the sedimentary Pa/Th record in the Pacific Ocean. In the Pacific, sediment Pa/Th tends to be low in the central basins and higher near the continental margins (Yang et al., 1985; Walter et al., 1999). This distribution of sediment Pa/Th is commonly attributed to ”boundary scavenging”, a process that enhances the scavenging of the less particle-reactive 231 Pa in zones of higher particle flux (Yang et al., 1985; Bacon, 1988; Taguchi et al., 1989; Lao et al., 1992; Yu et al., 2001; Roy-Barman, 2009), higher opal flux (Pichat et al., 2004; Bradtmiller et al., 2006), or higher opal and MnO2 concentration in settling particles (Anderson et al., 1983; Walter et al., 1999; Chase et al., 2002). These areas often coincide with continental margins. In contrast to the Atlantic, the residence time of deep water in the Pacific (~ 600 years) is much longer than the lateral diffusive mixing time in ocean basins (~ 100 years; Anderson et al., 1990), allowing a full expression of boundary scavenging (Yu et al., 2001). The Pacific Meridional Overturning Circulation (PMOC) has, thus far, not been considered as a significant factor affecting the large scale distribution of Pa/Th in Pacific sediments, in sharp contrast to the on-going debate concerning the influence of the AMOC in Atlantic sediments (e.g., Peacock et al., 2010; Gherardi et al., 2010). Building on chapter 2, we  112  have developed a Pacific 2D scavenging model with a schematic overturning circulation based on observations and tuned to produce water column profiles of dissolved 230 Th and 231 Pa similar to those measured in the Pacific Ocean. The results show that the PMOC, in addition to boundary scavenging, has a significant impact on the distribution of Pa/Th in Pacific sediments.  4.2 Model descriptions 4.2.1 Overturning circulation. The 2-D Pacific Meridional Overturning Circulation (PMOC) scheme (Figure 4.1) connects the meridional section in the Atlantic Ocean (constant 5000m depth from 70°N to 70°S, evenly divided into 20 layers and 56 columns; chapter 2) to a meridional section of the Pacific Ocean (constant 5000m depth and 4000km width from 55°N to 70°S evenly divided into 20 layers and 50 columns).  113                    ATLANTIC         PACIFIC Figure 4.1: Velocity vector plot for the Atlantic (Antarctic Intermediate Water (AAIW): 10 Sv; North Atlantic Deep Water (NADW): 20.5 Sv; Antarctic Bottom Water (AABW): 8 Sv) and the Pacific (Antarctic Intermediate Water (AAIW): 3 Sv; North Pacific Intermediate Water (NPIW): 4 Sv; Lower Circumpolar Deep Water (LCDW)/Antarctic Bottom Water (AABW): 26 Sv). The circulation in the Atlantic section is reported in chapter 3. The Atlantic and the Pacific sections are connected by a single mixing box from 0 to 1000m at 57.5°S-70°S in the Atlantic and 62.5°S-70°S in the Pacific. The geometry and transport rates of the different Atlantic water masses are based on published field observations (Friedrichs and Hall, 1993; Macdonald, 1998; Talley, 2003; Ganachaud and Wunsch, 2000) and discussed in chapter 2. Pacific Meridional Overturning Circulation (PMOC) is initiated by 26Sv of surface and intermediate water sinking in the southernmost column (67.5°S – 70°S) and 0 0 0 0 0 0 0 0 0 0 -5 -5 -5 -5 -5 -5 -5 -5 -5 -5 -5 -5 -5 -1 0 -10 -1 0 -1 0 -10 -10 -1 0 -1 0 -1 0 -10 -1 0 -10 -10 -15 -1 5 -1 5 -1 5 -15 -1 5 -1 5 -15 -1 5 -15 -15 -20 -2 0 -2 0 -2 0 -20 -2 0 -2 0 -20 -20 -20 5 5 5 5 5 -2 5 -2 5 -2 5 -2 5 Latitude D e p th Flux and Velocity Vector   N.A. 60N 50N 40N 30N 20N 10N 0 10S 20S 30S 40S 50S 60S S.O. 60S 50S 40S 30S 20S 10S 0 10N 20N 30N 40N N.P. 4750 4250 3750 3250 2750 2250 1750 1250 750 250 -25 -20 -15 -10 -5 0 5  114  flowing northward below 3000m to represent the Lower Circumpolar Deep Water (LCDW) and Antarctic Bottom Water (AABW) (Sloyan and Rintoul, 2001). This Southern Component Water mass (SCW) gradually upwells in the north Pacific to support the southward flow of North Pacific Deep Water (NPDW) between 1000m and 3000m. Most of the SCW (25Sv) upwells south of 45°N (Macdonald, 1998) while the remaining 1Sv is set to upwell further north. North Pacific Deep Water (NPDW) begins to upwell at 45°S toward the surface and subsurface of the southern ocean. At shallower depths, the North Pacific Intermediate Water is represented by a 4Sv meridional overturning cell above 1000m north of the equator (Kawabe and Fujio, 2010). In the south Pacific, the upper 1000m are occupied by a 3Sv overturning cell initiated by the formation of AAIW at 40°S (Talley, 2003). 4.2.2 Formulation of the two-dimensional scavenging model in the Pacific Ocean The scavenging model used in our study (Fig. 4.2) is based on the principle of reversible scavenging (Bacon and Anderson, 1982; Nozaki et al., 1987) and builds on the Atlantic model in chapter 2. In contrast to the Atlantic, however, where boundary scavenging can be neglected to a first approximation, boundary scavenging is much more pronounced in the Pacific Ocean and must be taken into account. Since this removal process cannot be explicitly represented in the 2-D meridional overturning scheme, we have added a removal term in each Pacific grid between 40°S and 35°N to reflect the lateral removal of 230 Th and 231 Pa from the central gyres to the margins. Dissolved 230 Th and 231 Pa concentrations in the Pacific 2-D model ([Xd] in dpm.m -3 ) are thus dictated by their  115  production rates from uranium decay (PX; dpm.m −3 .y −1 ), the adsorption (K X 1) and desorption (K X −1) rate constants (y −1 ), the transport rates imposed by the circulation scheme (V; m.y −1 ), and the removal rate to the margins by boundary scavenging, which we assume, as a first approximation, to be proportional to the concentration of dissolved nuclide (R[Xd]). Particulate  230 Th and 231 Pa concentrations ([Xp]) are controlled by the adsorption/desorption rate constants, transport rates and the sinking rate (S; m.y −1 ) of the particles that scavenge the two nuclides from the water column. At steady-state, we can write: Px − K1x [Xd] + K−1x [Xp] + VΔ[X]d - R [Xd] = 0  (4.1) K1x [Xd] − K−1x [Xp] + VΔ[Xp] + dFlux/dZ = 0   (4.2) dFlux = S( [Xp] (i+1)  − [Xp] (i)  )       (4.3) where X represents 230 Th or 231Pa, Z is water depth (m), i is the vertical index, Δ is an “upwind” difference divided by the grid spacing (Press et al., 1992), and R (y-1) is the effective removal rate constant to the margins, taken to be invariant between 40°S and 35°N. Using the upwind scheme with a horizontal velocity u=4×10 −3  m/s and a horizontal grid spacing Δx = 278×103 m, the inherent mixing in our model (Kdiff) is ~600 m 2 s -1  (Kdiff = u x/2); based on equivalence of the upwind scheme applied to an advective-reactive equation and an analytic diffusive-advective-reactive equation (e.g., Press et al., 1992)]. This is in the range of the along-isopycnal tracer diffusivities reported for the southern ocean (100~800m 2 s −1 ; Zika et al., 2009). Unlike lateral transport to the margins, the meridional lateral transport by turbulent mixing is therefore implicit in the model.  116   Figure 4.2: Scavenging and transport model in each model grid box. Xd = the concentration of dissolved 230 Th or 231 Pa (dpm/m 3 ). Xp = the concentration of particulate 230 Th or 231 Pa (dpm/m 3 ). Px = production rate of 230 Th or 231 Pa (dpm/m 3 /y). K1/K-1 = adsorption/desorption rate of 230 Th or 231 Pa (1/y). S = sinking rate of scavenging particles (m/y). V = transport rate by circulation (m/y). R is the effective removal rate constant to the margins. 4.2.3 Parameterization of the scavenging model. The parameters for the scavenging model in the Pacific have been selected following chapter 2, but the adsorption rate constants for 231 Pa and 230 Th had to be lowered in order to generate dissolved 231 Pa and 230 Th profiles similar to water column profiles measured in the North Pacific (Table 4.1). This is consistent with the general view that particle scavenging is, on average, more pronounced in the Atlantic compared to the Pacific because of its smaller volume and stronger aeolian input. The Southern Ocean’s opal belt is represented by higher K1 Pa  south of 45°S (Table 1b).  117  Table 4.1a: List of abbreviations and values for the Atlantic model parameters. Variables Symbo l Control run Units 231 Pa production rate PPa 0.00246 dpm/(m 3 *y r) 230 Th production rate PTh 0.0267 dpm/(m 3 *y r) Particle sinking rate  S 500 m/yr 230 Th adsorption rate (70°N-55°S) 0-250m K1 Th  1 1/yr 230 Th adsorption rate (55°S-70°S) 0-250m K1 Th  0.6 1/yr 230 Th adsorption rate  250-500m  K1 Th  75% of 0-250 m value 1/yr 230 Th adsorption rate > 500m  K1 Th  50% of 0-250 m value 1/yr 230 Th desorption rate (70°N-70°S) All depths K-1 Th  1.6 1/yr 231 Pa adsorption rate (70°N-60°N & 55°N -45°S) 0-250m K1 Pa  0.08 1/yr 231 Pa adsorption rate (60°N-55°N & 45°S -47.5°S) 0-250m K1 Pa  0.2 1/yr 231 Pa adsorption rate (47.5°S-50°S) 0-250m K1 Pa  0.32 1/yr 231 Pa adsorption rate (50°S-57.5°S) 0-250m  K1 Pa  0.44 1/yr 231 Pa adsorption rate (57.5°S-60°S) 0-250m  K1 Pa  0.36 1/yr 231 Pa adsorption rate (60°S-70°S) 0-250m K1 Pa  0.28 1/yr 231 Pa adsorption rate 250-500m  K1 Pa  75% of 0-250 m value 1/yr 231 Pa adsorption rate > 500m  K1 Pa  50% of 0-250 m value 1/yr 231 Pa desorption rate (55°N-70°S)  All depths K-1 Pa  1 1/yr  Table 4.1b: List of abbreviations and values for the Pacific model parameters. Variables Symbo l Control run Units 231 Pa production rate PPa 0.00246 dpm/(m 3 *yr ) 230 Th production rate PTh 0.0267 dpm/(m 3 *yr ) Particle sinking rate (55°N-35°N) S 750 m/yr Particle sinking rate (35°N-70°S) S 500 m/yr  118  Variables Symbo l Control run Units Boundary scavenging removal rate constant (40°S-55°N) R 0.0025 (AVG) 1/yr 230 Th adsorption rate (55°N-70°S) 0-250m K1 Th  0.6 1/yr 230 Th adsorption rate (55°N-70°S) 250-500m  K1 Th  75% of 0-250 m value 1/yr 230 Th adsorption rate (55°N-70°S) > 500m  K1 Th  50% of 0-250 m value 1/yr 230 Th desorption rate (55°N-70°S) All depths K-1 Th  1.6 1/yr 231 Pa adsorption rate (55°N-50°N) 0-250m K1 Pa  0.2 1/yr 231 Pa adsorption rate (50°N-35°N) 0-250m K1 Pa  0.16 1/yr 231 Pa adsorption rate (35°N-45°S) 0-250m K1 Pa  0.048 1/yr 231 Pa adsorption rate (45°S-47.5°S)  0-250m K1 Pa  0.2 1/yr 231 Pa adsorption rate (47.5°S-50°S) 0-250m K1 Pa  0.32 1/yr 231 Pa adsorption rate (50°S-57.5°S)  0-250m K1 Pa  0.44 1/yr 231 Pa adsorption rate (57.5°S-60°S)  0-250m K1 Pa  0.36 1/yr 231 Pa adsorption rate (60°S-70°S) 0-250m K1 Pa  0.28 1/yr 231 Pa adsorption rate  250-500m  K1 Pa  75% of 0-250 m value 1/yr 231 Pa adsorption rate > 500m  K1 Pa  50% of 0-250 m value 1/yr 231 Pa desorption rate (55°N-70°S) All depths K-1 Pa  1 1/yr  Similarly, K1 Pa  is higher between 35°N and 55°N to reflect higher opal fluxes in the North Pacific. The equilibrium fractionation factors F = (K-1 Pa  K1 Th )/(K1 Pa  K-1 Th ) generated by these rate constants range from 7.8 between 45°S and 35°N to 2 north of 50°N to a minimum of 0.9 between 50°S and 57.5°S (Table 4.2). The fractionation factor in the Southern Ocean opal belt reaches a minimum between 50°S and 57.5°S in both Atlantic and Pacific sector, to reflect the latitude of the region of highest opal flux. There is also an opal region in the North Atlantic (60°N-52.5°N) (Chapter 3).  119  Table 4.2: “Equilibrium” Fractionation Factors for the Pacific Model. Latitude “Equilibrium Fractionation Factor” 55°N – 50°N 1.88 50°N – 35°N 2.34 35°N – 45°S 7.8 45°S – 47.5°S 2 47.5°S – 50°S 1.24 50°S – 57.5°S 0.9 57.5°S – 60°S 1.1 60°S – 70°S 1.65  4.2.4 Estimating removal by “boundary scavenging” in the 2-D scavenging model This section is the contribution from my PhD committee member Susan Allen and it is therefore reported in Appendix C.  4.3 Results and discussion 4.3.1 Dissolved 230 Th and 231 Pa water column profiles 4.3.1.1. Data-model comparison We use published (Nozaki et al., 1987; Chase et al., 2003) and new (Table 4.3) 231 Pa and 230 Th seawater profiles (Fig. 4.3) to constrain the circulation and scavenging parameters in  120  our 2-D model. The dissolved 231 Pa and 230 Th profiles reported in Table 3 were measured by isotope dilution ICP-MS as described by Choi et al. (2000). In the absence of circulation, the reversible scavenging model predicts linear increases in dissolved and particulate 231 Pa and 230 Th concentrations with depth if the scavenging parameters (K1, K-1, S) remain constant (Bacon and Anderson, 1982; Nozaki et al., 1987): [X]p = [PX/S]*Z         (4.4) [X]d = [PX/K1] + [(K−1PX)/(K1S)]*Z     (4.5) We use these linear profiles as reference (“no-circulation” profiles) in Figure 4.4-4.5 to highlight the influence of circulation on the 230 Th and 231 Pa profiles. Profiles deviate from linearity when influenced by circulation and relax gradually back to linearity with an e-folding time equivalent to the residence time of the nuclide with respect to scavenging (e.g., Francois, 2007). When nuclide concentrations in seawater exceed the value predicted by the “no-circulation” profile, they tend to gradually decrease as water masses age (because the rate of adsorption on particles exceeds the rate of desorption). On the other hand, when nuclide concentrations are lower than predicted by the “no-circulation” profiles, they tend to gradually increase as water masses age.  121   Figure 4.3: Station locations for the water column profiles used to constrain the parameters in the model: North Pacific: CE-8, CE-13 (Nozaki et al. 1987); ALOHA (Table 4.3); South Pacific: SAZ (Table 4.3); and AESOPS stations MS-1 to MS-5 (Chase et al., 2002) Table 4.3: 230 Th and 231 Pa activities in sea water (dpm/1000kg). Station PaPa (50°N, 145°W) depth Total  230 Th Total  231 Pa m dpm/1000kg (± 95% CI) 200 0.092 ± 0.005 -0.002 ± -0.005 400 0.134 ± 0.008 0.005 ± 0.004 700 0.253 ± 0.014 0.026 ± 0.006  122  depth Total  230 Th Total  231 Pa depth Total  231 Pa Total  231 Pa depth m dpm/1000kg (± 95% CI) m dpm/1000kg (± 95% CI) m 1000 0.302 ± 0.019 0.056 ± 0.006 1300 0.336 ± 0.021 0.088 ± 0.011 1600 0.448 ± 0.025 0.098 ± 0.010 1900 0.543 ± 0.03 0.128 ± 0.012 2200 0.543 ± 0.03 0.132 ± 0.013 2500 0.550 ± 0.033 0.167 ± 0.020 2800 0.638 ± 0.035 0.153 ± 0.014 3100 0.720 ± 0.038 0.184 ± 0.015 3400 0.719 ± 0.039 0.192 ± 0.015 3700 0.799 ± 0.038 0.209 ± 0.018 4000 0.907 ± 0.042 0.209 ± 0.020 4315 1.165 ± 0.064 0.217 ± 0.016  Station Aloha (22°45’N, 158°W) depth Total  230 Th Total  231 Pa m dpm/1000kg (± 95% CI) 25 0.04949 ± 0.00222 0.011 ± 0.015 100 0.06284 ± 0.00252 0.045 ± 0.014 150 0.05663 ± 0.00766 0.025 ± 0.015 200 0.08648 ± 0.00438 0.038 ± 0.015 250 0.10694 ± 0.00436 0.028 ± 0.014 300 0.11576 ± 0.00362 0.041 ± 0.014 400 0.15316 ± 0.00349 0.065 ± 0.015 500 0.17743 ± 0.00426 0.096 ± 0.015  123  depth Total  230 Th Total  231 Pa depth Total  231 Pa Total  231 Pa depth m dpm/1000kg (± 95% CI) m dpm/1000kg (± 95% CI) m 650 0.26129 ± 0.00521 0.133 ± 0.017 800 0.32618 ± 0.00804 0.236 ± 0.019 1150 0.47534 ± 0.00839 0.401 ± 0.023 1500 0.59073 ± 0.00858 0.509 ± 0.022 1800 0.77509 ± 0.01 0.549 ± 0.019 2100 0.86417 ± 0.01218 0.581 ± 0.022 2450    0.581 ± 0.023 2800 1.09756 ± 0.01817 0.568 ± 0.023 3200 1.37107 ± 0.01983 0.568 ± 0.023 3600 1.58197 ± 0.02386 0.508 ± 0.022 4000 1.64832 ± 0.01838 0.466 ± 0.022 4410 1.66059 ± 0.02484 0.427 ± 0.019 4610 1.69769 ± 0.02938 4710 1.69085 ± 0.02694 0.430 ± 0.023 4760 1.63658 ± 0.02988 0.395 ± 0.021 4790 1.7175 ± 0.032 0.392 ± 0.020  Station SAZ2002-54S (54°S, 142°E) depth Total  230 Th Total  231 Pa m dpm/1000kg (± 95% CI) 30 0.096 ± 0.003 0.090 ± 0.012 100 0.094 ± 0.004 0.087 ± 0.011 200 0.230 ± 0.007 0.127 ± 0.009 400 0.328 ± 0.011 0.205 ± 0.017  124  depth Total  230 Th Total  231 Pa depth Total  231 Pa Total  231 Pa depth m dpm/1000kg (± 95% CI) m dpm/1000kg (± 95% CI) m 599 0.350 ± 0.010 0.213 ± 0.011 800 0.435 ± 0.009 0.246 ± 0.008 1102 0.515 ± 0.013 0.298 ± 0.022 1400 0.594 ± 0.010 0.308 ± 0.009 1702 0.641 ± 0.010 0.330 ± 0.015 2101 0.733 ± 0.025 2601 0.792 ± 0.013 0.363 ± 0.014 3183 0.861 ± 0.022 0.384 ± 0.010  Station SAZ2002-61S (61°S, 142°E) depth m Total  230 Th Total  231 Pa dpm/1000kg (± 95% CI) 21    0.271 ± 0.027 152    0.149 ± 0.012 401 0.405 ± 0.012 0.220 ± 0.012 801 0.566 ± 0.010 0.273 ± 0.011 1200 0.712 ± 0.022 0.315 ± 0.013 1600 0.774 ± 0.020 0.330 ± 0.012 2002 0.826 ± 0.021 0.330 ± 0.011 2399 0.917 ± 0.019 0.363 ± 0.012 2802 1.007 ± 0.029 0.365 ± 0.013 3200 1.091 ± 0.016 0.372 ± 0.010 3599 1.174 ± 0.021 0.397 ± 0.018 4003 1.194 ± 0.032 0.447 ± 0.015   125  Station SAZ2002-64S (64°S, 142°E) depth m Total  230 Th Total  231 Pa dpm/1000kg (± 95% CI) 30 0.183 ± 0.006 0.089 ± 0.006 152 0.143 ± 0.007 0.186 ± 0.007 401 0.599 ± 0.013 0.245 ± 0.013 599 0.686 ± 0.012 0.276 ± 0.012 801 0.737 ± 0.012 0.299 ± 0.012 1200 0.756 ± 0.012 0.323 ± 0.012 1601 0.912 ± 0.013 0.362 ± 0.013 2002 0.995 ± 0.017 0.350 ± 0.017 2401 1.079 ± 0.017 0.370 ± 0.017 2803 1.110 ± 0.012 0.378 ± 0.012 3201 1.107 ± 0.016 0.386 ± 0.016 3549 1.023 ± 0.012 0.367 ± 0.012  Station SAZ2002-66S (66°S, 142°E) depth m Total  230 Th Total  231 Pa dpm/1000kg (± 95% CI) 50 0.459 ± 0.01 0.180 ± 0.011 200 0.407 ± 0.010 0.158 ± 0.011 400 0.422 ± 0.010 0.157 ± 0.020 600 0.434 ± 0.011 0.167 ± 0.015 749 0.486 ± 0.012 0.168 ± 0.011   126  The 230 Th and 231 Pa profiles produced by the Pacific model between 54°S and 66°S (Figure 4.4c, f) exhibit convex shapes similar to those measured in Pacific sector of the southern ocean (Fig. 4.4a, b, d, e). The measured and modeled concentrations in this region are higher than expected for the “no-circulation” profiles, particularly for 231Pa. These high concentrations can be explained by upwelling of Upper Circumpolar Deep Water (UCDW), which brings deep waters with high nuclide concentrations to shallower depths (Rutgers van der Loeff and Berger, 1993; Chase et al., 2003; Francois, 2007), and by the lower affinity of marine particles for 231 Pa north of the opal belt (Walter et al., 1997; Chase et al., 2002), allowing 231 Pa to build-up in the NPDW before it returns to the Southern Ocean. The formation of AABW from shallow waters also influences the 230 Th and 231 Pa profiles by depressing concentrations in bottom waters. Profiles measured in the western tropical Pacific (Fig. 4.5) and at station ALOHA (Fig. 4.6) also show curvatures that are reproduced in the model and can be explained by deep water upwelling in the North Pacific. Southern Component Water (SCW) spreads northwards and shoals to intermediate depths in the north Pacific to form NPDW which flows to the south between 1000 and 3000 m depth. The maximum dissolved 231 Pa concentration at intermediate depth is due to upwelling, which brings deep seawater with relatively high 231 Pa to intermediate depth and decreases the sinking rate of small scavenging particles.  127   0 500 1000 1500 2000 2500 3000 3500 4000 4500 0 0.2 0.4 0.6 0.8 1 1.2 1.4 D ep th  ( m ) 230Th (dpm/T) SAZ2002 66S,142E SAZ2002 64S,142E SAZ2002 61S,142E SAZ2002 54S,142E Standard 0 500 1000 1500 2000 2500 3000 3500 4000 4500 0 0.2 0.4 0.6 0.8 1 1.2 1.4 D ep th  ( m ) 230Th (dpm/T) MS-1 53S,17.5W MS-2 57S,170W MS-3 60S,170W MS-4 63S,170W MS-5 66S,169W Standard 0 500 1000 1500 2000 2500 3000 3500 4000 4500 0 0.2 0.4 0.6 0.8 1 1.2 1.4 D ep th  ( m ) 230Th (dpm/T) Model 54S Model 59S Model 64S Model 66S Standard  128   Figure 4.4: Total 231 Pa and 230 Th profiles measured in the Southern Ocean (a) & (d) at 142°E longitude from 54°S to 66°S during the 2002 Sub-Antarctic Zone program (SAZ) (Table 4.2), (b) & (e) during the AESOPS program (Chase et al., 2003), and (c) & (f) predicted by our 2-D scavenging model between 54°S and 66°S. The linear profiles are concentrations predicted by the reversible scavenging 0 500 1000 1500 2000 2500 3000 3500 4000 4500 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 D ep th  ( m ) 231Pa (dpm/T)   Standard K1=0.024 Standard K1=0.14 Standard K1=0.22 SAZ2002 54S,142E SAZ2002 61S,142E SAZ2002 64S,142E SAZ2002 66S,142E 0 500 1000 1500 2000 2500 3000 3500 4000 4500 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 D ep th  ( m ) 231Pa (dpm/T) Standard K1=0.024 Standard K1=0.14 Standard K1=0.22 MS-1 53S,17.5W MS-2 57S,170W MS-3 60S,170W MS-4 63S,170W MS-5 66S,169W 0 500 1000 1500 2000 2500 3000 3500 4000 4500 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 D ep th  ( m ) 231Pa (dpm/T) Standard K1=0.024 Standard K1=0.14 Standard K1=0.22 Model 49S Model 54S Model 59S Model 66S  129  model in absence of circulation with K1 = 0.3 y -1  for 230 Th, and with 0.14 < K1 < 0.22 (southern ocean) and K1 = 0.024 y -1  (47.5°N – 45°S) for 231 Pa. Figure 4.5b and 4.6b indicate that 231Pa concentrations are lower than the “no circulation” values below ~2500 m. We thus expect 231 Pa concentration to increase as deep water is upwelled to intermediate depths. The 230 Th profile at Station ALOHA also shows a convex shape because of the upwelling, but the curvature is not as pronounced as for 231 Pa due to the greater particle reactivity and shorter residence time of Th in the water column, allowing it to relax towards linearity more quickly.  Figure 4.5: Dissolved 230 Th (a) and 231 Pa (b) at station CE-8 (12.45°N, 173°E) and CE-13 (13.12°N, 152°E; Nozaki et al., 1987) compared to model results at 14°N. 0 1000 2000 3000 4000 5000 6000 0 0.4 0.8 1.2 1.6 2 D ep th  ( m ) 230Th (dpm/T) CE-8 12.45N,173E CE-13 12N,152E Model 14N Standard 0 1000 2000 3000 4000 5000 6000 0 0.2 0.4 0.6 0.8 1 1.2 D ep th  ( m ) 231Pa (dpm/T) CE-8 12.5N,173E CE-13 12N,152E Model 14N Standard  130   Figure 4.6: Dissolved 230 Th  (a) and 231 Pa (b) at station Aloha compared to model results at 21°N and 37.5°N. 4.3.1.2. The influence of boundary scavenging on the curvature of the 231 Pa seawater profiles in the North Pacific The dissolved 231 Pa seawater profiles generated by the 2D scavenging model suggest that PMOC is the primary factor producing the curvature often reported in the dissolved 231 Pa seawater profiles in the North Pacific. To test whether boundary scavenging could also influence the shape of these profiles, the scavenging model was run with PMOC at a fixed rate of 26Sv while varying the effective removal rate constant to the margins (R) by a constant multiplier (Fig. 4.7). The results indicate that the curvature generated by the 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 0 0.4 0.8 1.2 1.6 2 D ep th  ( m ) 230Th (dpm/T) Aloha Feb/2002 Model 21N Model 31N Standard Standard * 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 0 0.2 0.4 0.6 0.8 1 D ep th  ( m ) 231Pa (dpm/T) Aloha Feb/2002 Model 21N Model 31N Standard Standard *  131  PMOC is more pronounced when R is smaller indicating that boundary scavenging tends to reduce the curvature generated by circulation.  Figure 4.7: Changes in the curvature of the dissolved 231 Pa profile generated at 21°N with 26Sv PMOC and varying effective rate constant for removal to the margins. 4.3.2 231 Pa and 230 Th sections generated by the Pacific 2-D scavenging model The sections generated by the Atlantic section of the 2-D scavenging model have been discussed in details in previous chapters. Here, we are discussing the influence of the overturning circulation in the Pacific section.  0 1000 2000 3000 4000 5000 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 D ep th  ( m ) Dissolved 231Pa (dpm/T) Aloha Feb/2002 Standard Standard* Model 21N (No R) Model 21N (1/2*R) Model 21N (CTR run) Model 21N (2*R)  132  4.3.2.1 Dissolved 230 Th The dissolved 230 Th generated by our 2D scavenging model (Fig. 4.8) gradually increases in the SCW as this water mass spreads northward, reaching a maximum in the zone of deep water upwelling in the North Pacific.  Figure 4.8: Dissolved 230 Th section generated by the model. Comparing the dissolved 230 Th concentrations generated by the model to the dissolved 230 Th concentrations expected in absence of circulation (Fig. 4.9) indicates that, while SCW starts at concentrations lower than the “no-circulation” values in the southern ocean, its dissolved 230Th concentrations quickly reach the “no-circulation” value and then increase slowly to the north. The reason for this unexpected result (i.e., that dissolved 230 Th concentration continues to increase as SCW moves northward even though it already exceeds the concentration at steady state with respect to scavenging) is the deep water upwelling in the North Pacific, which generates NPDW with a large dissolved 230 Th 0.2 0.2 0.2 0 .4 0.4 0.4 0.4 0 .6 0.6 0.6 0.6 0.6 0 .8 0 .8 0 .8 0.8 0.8 0.8 0 .8 1 1 1 1 1 1 11.2 1.2 1.2 1.2 1 .2 1 .4 1.4 1 .4 1 .4 1 .6 1. 6 1 .6 1. 2 Latitude D e p th Pacific dissolved 230Th   65s 55s 45s 35s 25s 15s 5s 5N 15N 25N 35N 45N 4500 4000 3500 3000 2500 2000 1500 1000 500 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8  133  excess, which is transmitted to deeper waters by scavenging. The NPDW entrains high 230 Th water southward between 1000m and 3000m, but the high 230 Th dissolved concentration is quickly attenuated because of the thorium particle reactivity . The AAIW and NPIW above 1000m in the model have limited influence on the dissolved 230 Th distribution due to the short residence time of 230 Th in shallow waters.  Figure 4.9: Difference between the dissolved 230 Th concentration generated by the 2-D model and the concentration predicted in the absence of circulation. Positive values indicate that dissolved 230 Th concentration in the presence of circulation exceeds the steady state value with respect to scavenging. 4.3.2.2 Particulate 230 Th The section of particulate 230 Th concentration (Fig. 4.10) is similar to that of dissolved 230 Th. The most conspicuous feature is the maximum in the deep North Pacific. -0 .4 -0 .3 -0.2 -0 .2 -0.1 -0 .1 0 0 0 0 0 0 0 0 0 0 0 0 .1 0 .1 0 .1 0.1 0 .1 0 .2 0.2 0 .2 0 .2 0 .2 0.3 0 .3 0 .3 0 .3 0 .4 0.4 0 .4 0 .4 0.1 0 .1 0.1 0 .1 0 .5 0 .5 0 .5 0 .2 0 .2 0 .2 0 .3 0 .3 0 .3 Latitude D e p th Dissolved Th, circu-no circu   65s 55s 45s 35s 25s 15s 5s 5N 15N 25N 35N 45N 4500 4000 3500 3000 2500 2000 1500 1000 500 0 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5  134   Figure 4.10: Particulate 230 Th section generated by the model.  Figure 4.11: Fraction of total 230 Th associated with particles generated by the model. The fraction of total 230 Th associated with particles ([ 230 Th]p/[ 230 Th]t) ranges from 0.12 to 0.15 over most of the Pacific (Fig. 4.11), which is lower than generated by the model in the Atlantic (0.18-0.22; chapter 2) but still conforms to field observations (Bacon and Anderson, 1982; Moran et al., 2002). Higher fractions are generated in the upper (<1000m) 0.05 0.05 0.05 0.05 0.1 0.1 0.1 0.1 0 .1 5 0 .1 5 0 .1 5 0.15 0.1 5 0.15 0 .1 5 0.2 0.2 0. 2 0 .2 0 .2 0 .2 50 .25 0 .2 5 0 .2 5 0 .3 0 .3 Latitude D e p th Pacific particulate 230Th   65s 55s 45s 35s 25s 15s 5s 5N 15N 25N 35N 45N 4500 4000 3500 3000 2500 2000 1500 1000 500 0 0.05 0.1 0.15 0.2 0.25 0.3 0.12 0 .1 2 0.1 3 0.13 0.13 0 .1 3 0.1 40.140.14 0.14 0.14 0.14 0 .1 4 0 .1 4 0 .1 5 0 .1 5 0.15 0. 15 0. 15 0 .1 5 0.1 3 0.13 0.130.13 0.16 0 .1 6 0 .1 6 0 .1 70. 180 .1 90 .20 .2 1 0 .2 2 0.12 0 .1 2 0.12 0.11 0.12 0 .1 7 0.12 0. 14 Latitude D e p th Particulate 230Th/Total 230Th   65s 55s 45s 35s 25s 15s 5s 5N 15N 25N 35N 45N 4500 4000 3500 3000 2500 2000 1500 1000 500 0 0.11 0.12 0.13 0.14 0.15 0.16 0.17 0.18 0.19 0.2 0.21  135  water column of the Southern Ocean (>0.22), reflecting the convection and the resulting longer residence time of particles in the water column. 4.3.2.3 Dissolved 231 Pa Compared to dissolved 230 Th, the distribution of dissolved 231 Pa concentration is more strongly influenced by PMOC, because of its longer residence time in the water column. Pa-231 concentrations gradually increase along the path of the SCW, reaching a maximum at mid-depth in the North Pacific where deep waters upwell (Fig. 4.12). This is consistent with the observation that dissolved 231 Pa concentrations in the deep Pacific remain well below the concentrations predicted in the absence of circulation below 2500m depth (Figure 4.5b, 4.6b, 4.13). As the SCW spreads northward and leaves the opal dominated region of the southern ocean, the “no-circulation” concentrations for 231Pa increase sharply, as K1 Pa  decreases north of the opal belt (Table 4.1b). Because the “no-circulation” concentrations exceed the 231 Pa concentrations in deep water, the latter increase northward. Above 3000m, the initially high dissolved 231 Pa concentration of NPDW decreases to the south as dissolved 231 Pa is slowly scavenged from the southward flowing water mass. In contrast to 230 Th, however, mid-depth waters with relatively high 231 Pa concentration propagate far into the South Pacific.  136   Figure 4.12: Dissolved 231 Pa section generated by the model.  Figure 4.13: Difference between the dissolved 231 Pa concentration generated by the 2-D model and the concentration predicted in the absence of circulation. Negative values indicate that dissolved 231 Pa concentrations are lower than the steady state (non-circulating) 231 Pa concentration with respect to scavenging, allowing 231 Pa to increase along the path of the water masses. 0 .1 0.1 0.1 0 .1 0 .2 0.2 0.2 0.2 0 .2 0 .2 0 .3 0 .3 0 .3 0.3 0.3 0 .3 0 .4 0.4 0.4 0.4 0.4 0 .4 0 .4 0 .5 0.5 0.5 0.5 0.5 0 .5 0 .3 0.3 0 .6 0. 6 Latitude D e p th Pacific dissolved 231Pa   65s 55s 45s 35s 25s 15s 5s 5N 15N 25N 35N 45N 4500 4000 3500 3000 2500 2000 1500 1000 500 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55 -0 .7 -0 .6 -0 .5 -0 .5 -0 .4 -0 .4 -0.3 -0.3 -0 .3 -0.2 -0.2 -0 .2 -0 .2 -0 .1 -0.1 -0.1 -0 .1 -0 .1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 .1 0 .1 0 .1 0.1 0.1 0.1 0 .1 0 .1 0 .1 0 .1 0 .1 0 .1 0 .1 0. 2 0 .2 0.2 0 .2 0 .2 0 .2 0 .2 0 .2 0 .3 0 .3 0 .2 0 .2 -0 .1 Latitude D e p th Dissolved Pa, circu-no circu   65s 55s 45s 35s 25s 15s 5s 5N 15N 25N 35N 45N 4500 4000 3500 3000 2500 2000 1500 1000 500 0 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5  137  4.3.2.4 Particulate 231 Pa Particulate 231 Pa concentrations (Fig. 4.14) are mostly controlled by the presence of opal, which has a high affinity for Pa (Walter et al., 1997; Chase et al., 2002). Accordingly, concentrations are highest in the southern ocean and in the North Pacific, resulting from the higher K1 Pa  in these regions.  Figure 4.14: Particulate 231 Pa section generated by the model. The fraction of particulate 231 Pa generated by the model between 55°S and 45°N ranges between 0.02 and 0.04 (Fig. 4.15) with higher values in the North Pacific (0.06-0.08) and the Southern Ocean (0.10–0.18). The latter are consistent with the values from the Atlantic sector of the southern ocean reported by Rutgers van der Loeff and Berger (1993). 0 .0 1 0.01 0.01 0 .0 1 0.01 0 .0 1 0.0 2 0 .0 2 0 .0 2 0 .0 3 0 .0 3 0 .0 30.0 3 0 .0 3 0 .0 4 0 .0 4 0 .0 4 0 .0 4 0 .0 5 0 .0 50 .0 5 0 .0 2 0 .0 2 0 .0 2 0 .0 6 0 .0 6 0 .0 6 0 .0 3 0 .0 30 .0 4 0 .0 7 Latitude D e p th Pacific particulate 231Pa   65s 55s 45s 35s 25s 15s 5s 5N 15N 25N 35N 45N 4500 4000 3500 3000 2500 2000 1500 1000 500 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07  138   Figure 4.15: Fraction of total 231 Pa associated with particles generated by the model. 4.3.2.5 Dissolved Pa/Th Between the northern and southern opal belts, modeled dissolved Pa/Th ratios decrease with water depth below 1000m (Fig. 4.16).  Figure 4.16: Dissolved 231 Pa/ 230 Th section generated by the model. 0 .0 2 0.02 0.02 0 .0 2 0.02 0.02 0 .0 4 0 .0 4 0 .0 4 0 .0 6 0 .0 6 0 .0 6 0 .0 8 0 .0 8 0 .0 8 0 .0 8 0 .1 0 .1 0 .1 0 .1 0 .0 4 0 .0 4 0 .0 4 0 .1 2 0 .1 2 0 .1 2 0 .1 2 0 .1 2 0 .1 4 0 .1 4 0 .1 4 0 .1 4 0 .0 6 0 .0 6 0 .0 6 0 .0 6 0 .1 6 0 .1 6 0 .1 6 0. 18 0 .1 8 0 .1 2 0. 14 0.12 0 .1 2 0 .0 8 0 .0 6 Latitude D e p th Particulate 231Pa/Total 231Pa   65s 55s 45s 35s 25s 15s 5s 5N 15N 25N 35N 45N 4500 4000 3500 3000 2500 2000 1500 1000 500 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 0 .2 0 .2 0. 4 0.4 0 .4 0 .4 0.4 0.4 0 .4 0. 4 0.60.6 0.6 0.6 0. 6 0.6 0 .6 0 .6 0.80.8 0 .8 1 1 1.21 Latitude D e p th Pacific dissolved 231Pa/230Th   65s 55s 45s 35s 25s 15s 5s 5N 15N 25N 35N 45N 4500 4000 3500 3000 2500 2000 1500 1000 500 0 0.2 0.4 0.6 0.8 1 1.2 1.4  139  While this decrease within the intermediate waters is found over the entire Pacific, it is more pronounced south of the equator (Fig. 4.17a). In the opal belts of the southern ocean and the North Pacific, the Pa/Th ratios are lower (Fig. 4.16) because of the higher Pa adsorption rates and the ratios are nearly uniform with depth (Fig. 4.17c). Available field data show similar vertical trends (Fig. 4.17b, d). In the southern Pacific, maximum dissolved Pa/Th ratios occur at the surface just north of the opal belt and a sharply decrease with depth, while within the opal belt, dissolved Pa/Th is low and nearly constant with depth (Fig. 4.17d). In the subtropical Pacific, dissolved Pa/Th ratios decrease with depth from a maximum of ~0.8 at 1000 m in both the model (Fig. 4.17a) and the field data (Fig. 4.17b). Model and field data diverge in shallower water and may reflect the larger errors associated with field measurements of very low concentrations or misrepresentation of the shallow overturning cells in the model. 4.3.2.6 Particulate and sedimentary Pa/Th The distribution of particulate Pa/Th produced by the model is shown in Figure 4.18. Between the southern and northern opal belts, the model predicts a decrease with depth below 1500m south of the equator and a northward decrease below 1500m in the North Pacific.  140    Figure 4.17: Dissolved 231 Pa/ 230 Th (a) generated by the model between the two opal belts below 1000m; (b) measured in the subtropical North Pacific below 1000m; (c) generated by the model in the 1000 1500 2000 2500 3000 3500 4000 4500 5000 0.2 0.4 0.6 0.8 1 D ep th  ( m ) Dissolved 231Pa/230Th Model 24S Model 9S Model 14N Model 21N 1000 1500 2000 2500 3000 3500 4000 4500 5000 0.2 0.4 0.6 0.8 1 D ep th  ( m ) Dissolved 231Pa/230Th CE-13 12N,152E CE-8 12.5N,173E Aloha Feb/2002 1000 1500 2000 2500 3000 3500 4000 4500 5000 0 0.2 0.4 0.6 0.8 1 D ep th  ( m ) Dissolved 231Pa/230Th Model 66S Model 61S Model 56S Model 53S Model 44S 1000 1500 2000 2500 3000 3500 4000 4500 5000 0 0.2 0.4 0.6 0.8 1 D ep th  ( m ) Dissolved 231Pa/230Th MS-1 53S,17.5W MS-2 57S,170W MS-3 60S,170W MS-4 63S,170W MS-5 66S,169W SAZ2002 54S,142E SAZ2002 61S,142E SAZ2002 64S,142E SAZ2002 66S,142E  141  Southern opal belt (red) and just to the north (blue), (d) measured in the Pacific section of the southern ocean (MS stations are from Chase et al., 2002; SAZ stations are reported in Table 3) to the north and south of the Polar Front. To generate sections of sediment Pa/Th in the 2D model, we take into account the fact that sinking particles are not in equilibrium with seawater (chapter 2). Upon reaching the seafloor, particles are not usually buried immediately and reach equilibrium with bottom waters (chapter 2). We can calculate the Pa/Th of sediment deposited at a given latitude and depth and at equilibrium with bottom waters using [X]p/[X]d=K X 1/K X −1 for 230 Th and 231 Pa (Fig. 4.19).  Figure 4.18: Particulate 231 Pa/ 230 Th section generated by the model. Partial equilibration would result in sediment Pa/Th intermediate between values reported in Figs. 4.18 and 4.19. The distribution pattern of sediment Pa/Th is very similar to that of particulate Pa/Th but with more pronounced vertical gradients. This is because settling 0 .0 4 0.04 0 .0 5 0 .0 5 0.0 5 0.05 0 .0 5 0 .0 6 0 .0 6 0 .0 6 0. 06 0.06 0 .0 6 0 .0 7 0 .0 7 0.07 0.07 0. 07 0.07 0 .0 7 0 .0 8 0 .0 8 0.080.08 0 .0 8 0. 08 0 .0 8 0 .0 8 0. 09 0 .0 9 0 .0 9 0. 1 0 .1 0 .1 0 .0 9 0 .0 9 0 .0 9 0. 09 0. 09 0 .1 1 0 .1 1 0 .1 1 0 .1 2 0 .1 2 0 .1 2 0 .1 3 0 .1 3 0 .1 3 0. 14 0 .1 4 0 .1 4 0 .1 0 .1 0 .1 0 .1 0 .1 0.1 0.1 0. 11 0 .1 1 0 .1 1 0 .1 1 0 .1 1 0 .1 2 0 .1 2 0 .1 2 0 .1 2 0 .1 2 0 .1 3 0 .1 3 0 .1 3 0 .1 3 0 .1 3 0 .1 4 0 .1 4 0 .1 4 0 .1 4 0.1 1 0.11 0.120.1 2 0 .0 7 Latitude D e p th Pacific particulate 231Pa/230Th   65s 55s 45s 35s 25s 15s 5s 5N 15N 25N 35N 45N 4500 4000 3500 3000 2500 2000 1500 1000 500 0 0.04 0.05 0.06 0.07 0.08 0.09 0.1 0.11 0.12 0.13 0.14  142  particles are farther from chemical equilibrium with ambient seawater at shallower depths (chapter 2), reflecting the trend generated by the model for dissolved Pa/Th, the vertical gradients of sediment Pa/Th are stronger south of the equator (Fig. 4.20).  Figure 4.19: Sedimentary 231 Pa/ 230 Th section generated by the model.  Figure 4.20: Sediment 231 Pa/ 230 Th generated by the 2D model at 21°S and 21°N. 0 .0 4 0 .0 4 0.04 0.04 0 .0 4 0 .0 5 0 .0 5 0.05 0.05 0 .0 5 0 .0 6 0 .0 6 0 .0 6 0.06 0.06 0 .0 6 0 .0 7 0 .0 7 0. 07 0.07 0.07 0 .0 7 0 .0 8 0 .0 8 0.08 0.080 .0 8 0.08 0. 08 0.08 0 .0 8 0 .0 9 0 .0 9 0.09 0.09 0.09 0.09 0.09 0 .0 9 0 .0 9 0 .1 0 .1 0 .1 0.10.1 0.1 0 .1 0 .1 0. 11 0 .1 1 0 .1 1 0 .1 1 0 .1 1 0.110.1 1 0. 11 0 .1 1 0 .1 1 0 .1 2 0 .1 2 0 .1 2 0 .1 3 0 .1 3 0 .1 3 0 .1 4 0 .1 4 0 .1 4 0 .1 2 0 .1 2 0 .1 2 0 .1 2 0 .1 2 0.120.12 0.13 0.13 0 .1 3 0 .1 3 0 .1 3 0.130 .13 0.07 0.07 0.0 7 0 .1 4 0 .1 4 0 .1 4 0 .1 4 0 .0 6 0 .0 6 0.14 0.05 Latitude D e p th Pacific sediment 231Pa/230Th   65s 55s 45s 35s 25s 15s 5s 5N 15N 25N 35N 45N 4500 4000 3500 3000 2500 2000 1500 1000 500 0 0.04 0.05 0.06 0.07 0.08 0.09 0.1 0.11 0.12 0.13 0.14 1500 2000 2500 3000 3500 4000 4500 5000 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 D ep th  ( m ) Sediment 231Pa/230Th Model 21S Model 1S Model 1N Model 21N  143  4.3.3 Comparison between sediment Pa/Th measured in the Pacific and model predictions We use a compilation of surface sediment Pa/Th (Fig. 4.21) reported by Lao et al. (1992) and Pichat et al. (2004) (and references therein) and recent measurements published by Bratmiller et al. (2006)  (see Appendix C) to investigate whether the depth dependency predicted by our 2D model can also be recognized in the Pacific sediment Pa/Th database. In the eastern equatorial Pacific, Pichat et al. (2004) found a linear correlation (r 2  = 0.69) between the surface sediment Pa/Th and export production estimated from satellite data. Likewise, Bradtmiller et al. (2006) found a linear correlation (r 2  = 0.66) between Holocene sediment Pa/Th and 230 Th-normalized opal flux in cores spanning the entire equatorial Pacific region. These results indicate that particle flux is an important factor controlling sediment Pa/Th, at least in the equatorial upwelling region. However, Pichat et al. (2004) also noted that variations in sediment Pa/Th could not be readily explained by variations in productivity estimated from ocean color in the Western Equatorial Pacific. They tentatively attributed this observation to underestimations of the export flux from satellite data in this region, but other factors could also be important.  144   Figure 4.21: Distribution of core-top Pa/Th data. Data taken from Lao et al. (1992), Pichat et al. (2004) and Bradtmiller et al. (2006), and references therein. Blue diamonds = cores from low productivity regions at depth < 5000m; white diamonds = cores from low productivity regions at depth > 5000m (i.e., deeper than the 2D scavenging model); Orange squares = cores from the eastern equatorial upwelling region with high Pa/Th; grey squares = cores from the eastern equatorial upwelling region with low Pa/Th; brown squares = cores from the western equatorial Pacific region with high Pa/Th; green triangles = cores from the northeast Pacific margin; red triangle = core from the southeast Pacific margin; yellow triangle = core from the western equatorial Pacific margin; magenta squares = cores affected by biogenic silica;  light blue squares: cores at the boundary between subtropical and subpolar regions; black circles = cores from low productivity sites where Pa/Th is anomalously high; green diamonds = cores from low productivity sites where Pa/Th is anomalously low (see also Fig. 24; cores affected by hydrothermal activity (Pichat et al., 2004) are not included).  145  The correlation between the entire database and our model output (i.e., sediment Pa/Th generated by the model at the same depth and latitude as the cores) is not very high (Fig. 4.22; r 2  = 0.12) and the sediment data are often higher than the model output. This is not surprising since many of the cores analyzed come from areas close enough to the coast to be directly affected by boundary scavenging, which cannot be represented in the 2D scavenging model.  Figure 4.22: Sediment Pa/Th measured in the Pacific vs. sediment Pa/Th generated by the 2D scavenging model at the same latitude and depth. Symbols are the same as in Fig. 4.21. [Blue diamonds = cores from low productivity regions at depth < 5000m; white diamonds = cores from low productivity regions at depth > 5000m (i.e. deeper than the 2D scavenging model); Orange squares = y = 1.01x R² = 0.77 0 0.05 0.1 0.15 0.2 0 0.05 0.1 0.15 0.2 m o d e l P a /T h  Measured Pa/Th Low Prod. (>5000m) Low Prod. Opal Opal Boundary N. Pac Margin S. Pac Margin Asian Margin EEP EEP - Lo WEP Hi Pa/Th Too Hi - Lo P Too Lo - Lo P  146  cores from the eastern equatorial upwelling region; grey squares = cores from the eastern equatorial upwelling region with lower Pa/Th; brown squares = cores from the western equatorial Pacific region with high Pa/Th; green triangles = cores from the northeast Pacific margin; red triangle = core from the southeast Pacific margin; yellow triangle = core from the western equatorial Pacific margin; magenta squares = cores affected by biogenic silica; green squares: cores at the boundary between subtropical and subpolar regions; black circles = cores from low productivity sites where Pa/Th is anomalously high; green diamonds = cores from low productivity sites where Pa/Th is anomalously low]. We note, however, that a group of data points plot close to the 1:1 line (blue and white diamonds in Figs. 4.21 and 4.22; Modeled Pa/Th = 1.01 (Measured Pa/Th); r 2  = 0.77), indicating that they are well reproduced by the model. These cores were taken in low productivity regions of the Pacific (Fig. 4.21), and their Pa/Th clearly decreases with depth between 2000m and 5000m (Fig. 4.23), as predicted by the 2D scavenging model. Some of these cores were taken below the maximum depth represented in our 2D model (5000m). Because observed sediment Pa/Th does not generally decrease between 5000 and 6000m, we compared Pa/Th measured in sediment below 5000m to model results obtained at 5000m (white diamonds; Fig. 4.22) but excluded them in the linear regressions (Fig. 4.23; r 2  = 0.88). Comparing this regression line with model output on Fig. 4.20 confirms that our 2D model reproduces reasonably well the vertical sediment Pa/Th gradient observed in the data, at least in low-productivity areas with low sedimentary Pa/Th ratios. The shallower low productivity cores (blue diamonds) in Fig. 4.23 come from the western Pacific Ocean. Those data are thus best compared to model predictions near the equator (Fig. 4.20; Model 1S and 1N).  147   Figure 4.23: Pa/Th in core tops versus depth (symbols are identical to Fig. 4.21). When plotting Pa/Th from the low productivity sites against satellite-derived export production, no significant correlation can be found (r 2 =0.005), possibly indicating that proximal scavenging has little effect on sediment Pa/Th in low productivity regions In the other regions of the Pacific, sediment Pa/Th is significantly higher than predicted by the low productivity depth trend (Fig. 4.23). Sediment Pa/Th from core tops in the eastern equatorial upwelling region also decreases with depth (r 2  = 0.80), but Pa/Th at a given R² = 0.8809 R² = 0.8015 0 1000 2000 3000 4000 5000 6000 7000 0.00 0.05 0.10 0.15 0.20 D ep th  ( m ) Measured sediment Pa/Th Low prod. (>5000m) Low prod. Opal Opal Boundary N. Pac Margin S. Pac. Margin Asian Margin EEP EEP-Lo WEP Hi Pa/Th Too Hi - Lo P Too Lo - Lo P  148  depth is much higher than in low productivity regions, reflecting the direct effect of boundary scavenging.  Figure 4.24a: 230 Th-normalized opal flux against sediment Pa/Th in the equatorial Pacific (Bradtmiller et al., 2006)  Figure 4.24b: Difference between Pa/Th measured and estimated from 230 Th-normalized fluxes as a function of depth and the linear regression (data from Bradtmiller et al., 2006; two yellow data points show the data from western Pacific and the others are data from eastern Pacific). y = 1.1673x - 0.0454 R 2  = 0.6646 0 0.1 0.2 0.3 0.05 0.1 0.15 0.2 231 Pa/ 230 Th 2 3 0 T h -n o rm a li z e d  o p a l fl u x y = -4E-05x + 0.1095 R 2  = 0.8751 -0.05 -0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 1500 2000 2500 3000 3500 4000 4500 Depth (m)   2 3 1 P a /2 3 0 T h  149  In addition to correlating with water depth, Pa/Th at these sites also correlates, but less tightly, with export production estimated from satellite data (r 2  = 0.59), as already noted by Pichat et al. (2004). Although export production and water depth also weakly correlate (r 2  = 0.38) at these core locations, both water depth and particle flux appear to play a role in determining sediment Pa/Th in the Pacific equatorial upwelling region. This is further confirmed when revisiting the data reported by Bradtmiller et al. (2006), who found a correlation between 230 Th-normalized opal fluxes and sediment Pa/Th using 11 cores spanning the entire equatorial Pacific (Fig. 4.24a). Plotting the differences between measured Pa/Th and the regression line reported in Figure 4. 24a against water depth (Fig. 4.24b) reveals a clear correlation (r 2  = 0.88), confirming that water depth is also important in dictating sediment Pa/Th in the equatorial Pacific. The three relatively shallow cores from the eastern equatorial Pacific with lower Pa/Th than neighboring cores (grey squares; Fig. 4.21-4.23) come from the Panama Basin just north of the equator, a region with relatively low productivity between the equatorial upwelling and the Costa Rica upwelling dome, and may thus reflect lower scavenging intensity at this particular location. Alternatively, lower Pa/Th could also be explained if the upper section of these cores is missing and the sediment analyzed is not modern. This is probably the explanation for two of the cores with very low Pa/Th taken from the low productivity region (green diamonds; Fig. 4.21-4.23), where core tops can be lost more readily due to lower sedimentation rates.  150  More intriguing, five core tops from the western equatorial Pacific (brown squares; Fig. 4.21 & 4.23) have significantly higher Pa/Th ratios than expected from their depth, even though they were collected in regions where satellite derived export production estimates are low. This observation was already discussed by Pichat et al. (2004) and is now confirmed by similarly high values from the Bradtmiller et al. (2006) data set. Pichat et al. (2004) suggest that the deep chlorophyll maximum observed in this region may account for a higher export production than is deduced from ocean color. Clearly, this feature needs further investigation by conducting water column measurements in the region. The 2D model also generates sediment Pa/Th lower than measured at three low productivity sites (black circles; Fig. 4.21-4.23). Two of the cores were taken in the vicinity of Hawaii and may reflect a local effect on particle flux scavenging or underwater volcanoes, but there is no explanation for the high value observed the central equatorial Pacific core (4810 m). The Pa/Th of sediment deposited at ocean margins and opal dominated regions (Fig. 4.21) are invariably higher than in the low productivity regions (Fig. 4.22, 4.23) reflecting the effect of enhanced scavenging. Some cores taken at the boundary between the subtropical and subpolar region (light blue squares) show the same depth dependency as the low productivity cores, suggesting that opal flux at these sites is too small to significantly affect Pa/Th in the underlying sediments.  151  4.3.4 Is the decreasing trend in sediment Pa/Th with depth below 1500m a result of the PMOC? The vertical gradient in sediment Pa/Th is one of the key features generated by the meridional overturning circulation in the Atlantic Ocean (chapter 2, 3; Lippold et al., 2011). To assess whether the vertical gradient in sediment Pa/Th observed in the Pacific and generated by the 2D scavenging model can also be attributed to the overturning circulation, we ran the scavenging model without PMOC while maintaining boundary scavenging.  Figure 4.25: Vertical sediment Pa/Th profiles generated between 35°N and 45°S by the 2D scavenging model in the absence of PMOC and with varying boundary scavenging strength. 1500 2000 2500 3000 3500 4000 4500 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10 D ep th  ( m ) Sediment Pa/Th generated by different Boundary scavenging without PMOC 0.5*R without PMOC 1*R without PMOC 2*R without PMOC  152  Under these conditions (Fig. 4.25), the model does not produce the decrease in sediment Pa/Th with depth below 1500m observed in the data or in the model when PMOC is present. As expected, increasing boundary scavenging decreases sediment Pa/Th in the low productivity regions represented in the 2D model, but the decrease happens at all depths. The vertical gradient in sediment Pa/Th in low productivity regions thus seems to be a direct consequence of the overturning circulation. 4.3.5 Sensitivity of sediment Pa/Th in low productivity regions to changes in the rate of PMOC The ultimate goal of this study is to assess whether sediment Pa/Th could potentially provide constraints on past changes in the rate of PMOC. To assess whether sediment Pa/Th in the Pacific could be used as a paleocirculation proxy, we must establish whether the vertical and latitudinal distributions of sediment Pa/Th in low productivity regions are sufficiently sensitive to PMOC strength and the extent to which they are also influenced by factors others than the PMOC. When varying PMOC from 13 Sv to 39 Sv in the model, and keeping the same water mass distribution and boundary scavenging intensity, the latitudinal gradients of sediment Pa/Th at different depths change significantly (Fig. 4.26). With 39 Sv PMOC (i.e., a 50% increase compared to today), sediment Pa/Th at 4750m is low and almost constant (~ 0.03) with latitude. As PMOC intensity decreases, sediment Pa/Th at 4750m increases gradually to a maximum at mid-latitudes in the northern hemisphere.  153  The model predicts that a 50% reduction in PMOC rate from 26Sv to 13Sv would generate an increase in sediment Pa/Th of ~ 0.01 at 20-30°N in deep water (4.26a). In contrast, at the base of the NPIW (~3000m), high rates of PMOC result in a sharp northward decrease in sediment Pa/Th from ~0.075 to ~0.035 between 40°S and 30°N (4.26b). Slowing down PMOC results in nearly constant and high (~0.075) sediment Pa/Th south of the equator and a sharp decrease further north, where SCW upwells. These model outputs suggest that the rate of PMOC would be more easily estimated from changes in sediment Pa/Th in the northern subtropical Pacific than in the South Pacific. Slower rates of PMOC would generate higher sediment Pa/Th at all depths in the North Pacific. Confirmation that the signal is generated by changes in PMOC rate could be obtained by determining the latitudinal gradients of sediment Pa/Th at a given depth. In addition to generating higher sediment Pa/Th in the northern subtropical Pacific, slower rates of PMOC should also produce an increase in the northward gradient of sediment Pa/Th in deep water. On the other hand, faster rates of PMOC would decrease sediment Pa/Th in the northern subtropical Pacific and reduce the latitudinal gradient in deep water but generate a sharp northward decrease in sediment Pa/Th at mid-depth. As already noted in Fig. 4.19 and Fig. 4.20, the vertical gradient of sediment Pa/Th between 5000m and 3000m is larger in the South Pacific and decreases gradually to the North (Fig. 4.27).   154     Figure 4.26: Variations in sediment Pa/Th as a function of latitude in low productivity regions predicted by the 2D scavenging model with varying PMOC rates. Squares are measured sediment Pa/Th from the low productivity zone. 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 -50 -40 -30 -20 -10 0 10 20 30 P a/ T h  Longitude (a) Sedi Pa/Th @ 4500-4750m 13Sv 26Sv 39Sv Data 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 -50 -40 -30 -20 -10 0 10 20 30 P a/ T h  Longitude (b) Sedi Pa/Th @ 2750-3000m 13Sv 26Sv 39Sv Data 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 -50 -40 -30 -20 -10 0 10 20 30 P a/ T h  Longitude (c) Sedi Pa/Th @ 1500-1750m 13Sv 26Sv 39Sv  155   Figure 4.27: Changes in the sediment Pa/Th vertical gradient between deep and intermediated depth as a function of latitude and PMOC strength. However, the relationship between the vertical gradients in sediment Pa/Th and the rate of PMOC is complex. An increase in the rate of PMOC from 26 to 39Sv would decrease the vertical gradients by ~ 0.01 between 40°S and 20°N, but a similar decrease between 30°S and the equator could be generated by a decrease in PMOC from 26Sv to 13Sv. These two situations could be distinguished from the vertical gradient at 10 – 20°N. High rates of PMOC would result in lower vertical gradients in this region than low rates of PMOC (Fig. 4.27). However, the predicted changes in gradients are rather small and it may be difficult to precisely measure them to constrain the rate of PMOC. While the available data capture the decrease in sediment Pa/Th with depth (Fig. 4.23), they are not precise enough to show the expected latitudinal trends (Fig. 4.26). This database has been obtained over the course of several decades using different techniques 0.00 0.01 0.02 0.03 0.04 0.05 0.06 -50 -40 -30 -20 -10 0 10 20 30 Δ P a/ T h  Longitude Sedi Pa/Th @ 3000m - Sedi Pa/Th @ 4750m 39 Sv 26 Sv 13 Sv  156  (radiometric and mass spectrometric) and without inter-calibration. The latter is particularly crucial to obtain a precise data base (Anderson et al., 2012). Establishing whether the latitudinal trends can be recognized in the sediments of the Pacific Ocean will require careful collection of surface sediment with appropriate coring techniques, assessing the impact of bioturbation, particularly in low sedimentation rate cores, and a systematic coring program based on our understanding of the main flow path of AABW and LCDW.  Figure 4.28: Difference in sediment Pa/Th generated by the 2D scavenging model with PMOC = 26Sv and 13Sv. Faster PMOC results in a substantial increase in sediment Pa/Th in the southern ocean. The 2D model also suggests that lower rates of PMOC would result in a substantially smaller 231 Pa export from the Pacific into the Southern Ocean, where it is effectively  157  scavenged by opal (Fig. 4.28). Sediment Pa/Th in the Pacific sector of the Southern ocean appears to have been significantly lower during the last glacial maximum than during the Holocene (Chase et al., 2003), which would be consistent with a significant decrease in the rate of PMOC (Jaccard et al., 2010). Such a trend could be further confirmed by a simultaneous increase in Pa/Th at mid-depth in mid-northern latitudes at 30 – 40 °N (Fig. 4.28). 4.3.6 Relative sensitivity of sediment Pa/Th in low productivity regions to changes in PMOC and “boundary scavenging” – What is the most promising approach to constrain PMOC from sediment Pa/Th? In the Pacific, the sediment Pa/Th in low productivity regions could also be significantly affected by changes in boundary scavenging. It is therefore important to establish how the signal generated by PMOC could be modified by possible changes in Pa scavenging at the margins. If we keep PMOC constant but multiply the effective removal rate constant to the margins (R) by a constant, sediment Pa/Th decreases with increasing R, as expected (Fig. 4.29). The changes in sediment Pa/Th are more pronounced at intermediate depths, resulting in large changes in the vertical gradients (Fig. 4.30), but the latitudinal gradients change little (Fig. 4.29). Latitudinal gradients in sediment Pa/Th in deep waters are thus sensitive mainly to changes in the rate of PMOC and largely unaffected by changes in boundary scavenging and could potentially be used to reconstruct past changes in the rate of the  158  overturning circulation in the Pacific Ocean. In contrast, changes in sediment Pa/Th at intermediate depth and changes in the vertical gradient in sediment Pa/Th are more sensitive to changes in boundary scavenging than they are to changes in PMOC.    Figure 4.29: Variations in sediment Pa/Th as a function of latitude in low productivity regions predicted by the 2D scavenging model with varying boundary scavenging (R = effective removal rate constant to the margins) 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 -50 -40 -30 -20 -10 0 10 20 30 P a/ T h  Longitude Sedi Pa/Th @ 4500-4750m 0.5R 1R 2R Data 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 -50 -40 -30 -20 -10 0 10 20 30 P a/ T h  Longitude Sedi Pa/Th @ 2750-3000m 0.5R 1R 2R Data 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 -50 -40 -30 -20 -10 0 10 20 30 P a/ T h  Longitude Sedi Pa/Th @ 1500-1750m 0.5R 1R 2R  159   Figure 4.30: changes in the difference in sediment Pa/Th between 3000 and 4750m as a function of latitude for a fixed PMOC (26Sv) and varying effective removal rate constant to the margins.  4.4. Conclusions The 2D scavenging model indicates that the Pacific meridional overturning circulation can explain the curvature in the seawater dissolved 231 Pa profiles often observed in the North Pacific. The model also indicates that changes in the rate of PMOC can influence the distribution of sediment Pa/Th, particularly in low productivity regions. The latter raises the prospect for developing a paleocirculation proxy for the Pacific Ocean. The results obtained in this study suggest that past changes in the rate of the PMOC would be more easily estimated from changes in the longitudinal gradient of sediment Pa/Th at depth, following the main flow path of deep water. On the other hand, changes in 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 -50 -40 -30 -20 -10 0 10 20 30 40 Δ P a/ T h  Longitude Sedi Pa/Th @ 3000m - Sedi Pa/Th @ 4750m 3000m-4750m 0.5R 3000m-4750m 1R 3000m-4750m 2R  160  sediment Pa/Th at intermediate depths and vertical gradients in sediment Pa/Th appear to be more sensitive to the intensity of boundary scavenging. Although PMOC seems essential to generate a vertical sediment Pa/Th gradient in low productivity regions, this gradient is strongly modulated by boundary scavenging. Sediment Pa/Th in the equatorial upwelling regions is affected both by particle flux and water depth, indicating that changes in circulation could also affect sediment Pa/Th in this region, and interpretation of changes in this ratio in terms of particle scavenging alone maybe unwarranted. The existing sediment Pa/Th database, although good enough to document the relatively large vertical gradient in low productivity areas, is not sufficiently precise and accurate to discern the expected latitudinal gradients. Documenting these relatively subtle changes to document past changes in PMOC will require a concerted effort involving careful planning of a targeted coring program and inter-calibration of measurements.       161   Chapter 5 A comparison of POC fluxes recorded by sediment traps and 234Th:238U disequilibrium in a coastal region (Saanich Inlet, British Columbia)  5.1 Introduction The role played by the coastal zone in the marine carbon cycle is disproportionate to its surface area. While accounting for only 8% of the surface area of the whole ocean, it contributes up to 30% of oceanic primary production and 80% of the organic carbon burial in marine sediments (Liu et al., 2000, Middelburg et al., 1997, Holt et al., 2009). However, notwithstanding the clear quantitative importance of the coastal zone for the carbon budget of the ocean and its potential role for sequestering anthropogenic CO2, there is not, as yet, a consensus on whether this oceanic region is a net source or a net sink for atmospheric CO2, and what fraction of the organic carbon produced in the coastal zone or added from nearby landmasses is exported to the deep sea, buried in margin sediments, or remineralized in coastal waters. The reason for the relatively poor characterization of the carbon cycle at ocean margins is  162  that coastal zones are complex and the processes regulating the fate of carbon are highly variable both in time and space. This complicates the interpretation and the integration of instantaneous or localized measurements of quantities such as primary production, export flux and burial rates. Although satellites provide a powerful tool to start addressing both spatial and temporal variability, they do not provide direct estimates of flux and are prone to biases resulting from non-biotic influences on the optical properties of surface waters (Carr et al., 2006). Clearly, there is a need to develop means of estimating carbon fluxes (uptake, export, burial) in coastal waters using methods that provide some level of integration to generate flux estimates that could be more meaningfully parameterized in global carbon models. According to Tsunogai et al. (1999), the biological pump operating in the coastal zone (i.e., the “continental shelf pump”) is distinct from the more classical biological pump of the ocean interior in that it requires lateral transport to offshore waters to remove carbon from the atmosphere on centennial to millennial timescales. Therefore, even though coastal production is high, its efficacy at sequestering atmospheric CO2 depends on physical factors which control the transport of carbon to the deep sea. The “continental shelf pump” is primarily driven by isopycnal transport of dissolved inorganic carbon (DIC) from shelf bottom waters to greater depths offshore. Dissolved inorganic carbon (DIC) concentration in bottom shelf waters is largely controlled by the remineralization of particulate organic carbon (POC) settling from surface coastal waters. It is also possible that a fraction of the settling POC is laterally exported to the continental slopes. Measuring the sinking flux of carbon in the coastal zone is thus an essential part of better characterizing the continental  163  shelf pump and its variability under different environmental conditions (high vs low latitude, wide vs narrow shelves; proximity to major rivers and canyons, circulation, etc.). In this chapter, two methods of estimating the sinking flux of organic carbon are compared in a coastal setting: bottom tethered sediment traps and 234 Th/ 238 U disequilibrium. Both approaches provide means of integrating the sinking flux of carbon over several weeks and documenting seasonal variations. However, both methods are also associated with significant uncertainties, particularly when applied in coastal waters. By comparing directly the two approaches in this setting, a reasonable agreement, while not proving their accuracy, would nonetheless bring more credence to the validity of the fluxes measured by either method.  5.2 Measuring the sinking flux of carbon with sediment traps Sediment traps of various designs have been widely used to measure the flux of sinking particles in the ocean. They range from free-drifting floating traps, which are deployed from surface buoys for a few days at a time (a more recent design consists of neutrally buoyant sediment traps; Buesseler et al., 2000), to bottom-tethered time-series traps that are typically deployed in the deep ocean for extended periods of time (Honjo et al., 2008). In coastal waters, bottom-tethered traps of simple designs, as used in this study, are often deployed for extended periods of time and turned-over on a regular basis to collect time-series samples (e.g., Timothy et al., 2003).  164  While conceptually straightforward, the measurement of particle flux with sediment traps is subject to significant uncertainties. When deployed in dynamic regions with strong currents, the flux measured by sediment traps often seems to exceed or underestimate the true vertical flux (e.g., Buesseler, 1991). This is mostly because of hydrodynamic biases resulting from the formation of eddies near the mouth of the trap (Butman et al., 1986). Another difficulty with sediment traps is the presence of “swimmers”, living organisms that swim into the trap and die if the trap has been poisoned (thereby increasing the measured carbon flux), or consume the organic matter accumulating in the trap in the absence of poison (thereby decreasing the measured carbon flux). These biases are more severe in shallow waters (Yu et al., 2001) and therefore particularly problematic in coastal waters, which are relatively shallow and dynamic environments.  5.3 Measuring the sinking flux of carbon using 234 Th/ 238 U disequilibrium in surface waters In addition to sediment traps, the 234 Th: 238 U disequilibrium method has also been applied extensively to estimate the sinking flux of carbon in the ocean (Buesseler et al., 2006 and references therein). While sediment traps can provide flux estimates at any depths, the 234 Th/ 238 U method is mostly restricted to the upper water column, where particle flux is high enough to produce a measurable deficit in 234 Th. Thorium-234 is a relatively short-lived (half-life: 24.1 days) isotope of thorium produced  165  by decay of dissolved 238 U in seawater. Uranium is soluble in seawater and has a long residence time in the ocean (200,000-400,000 years; Ku et al., 1977; Dunk et al., 2002). As a result, its concentration is virtually proportional to salinity (Owens et al., 2011; Chen et al., 1986), and the 234 Th formation rate is thus fairly uniform (i.e., proportional to salinity) throughout the ocean. The production rate of 234 Th (atoms/m 3 ) is the rate at which its parent 238 U decays, which is also known as the activity of 238 U:  A238 = 238 x N238      (5.1) where 238, N238 and A238 are the decay constant (min -1 or d -1 ), concentration (atoms/m 3 ) and activity (dpm/m 3 ) of 238 U. When sinking particles remove 234 Th from seawater in a closed system at steady-state, the rate of production of 234 Th must be equal to the sum of its rate of decay and removal by scavenging: 238 x N238 = 234 x N234  + r234    (5.2) or A238 = A234 + r234     (5.3) where 234, N234, A234 and r234 are the decay constant (d -1 ), total (dissolved + particulate)  166  concentration (atoms/m 3 ), activity (dpm/m 3 ) and removal rate by scavenging (atoms/m 3 .d) of 234 Th. Multiplying each term by 234 and rearranging equation 5.3 expresses the removal flux of 234 Th in dpm/m 3 .d (R234): R234 = 234 x (A238 - A234)    (5.4) When scavenging removes 234 Th at a rate which is similar to or greater than its rate of decay, as is the case in most oceanic surface waters, the activity of 234 Th in seawater drops below that of 238 U, and this measurable deficit can be used to estimate R234. Integrating over the depth of the euphotic zone (or the depth where a measureable 234 Th deficit is found) gives the removal rate of 234 Th from the upper water column, expressed in dpm/m 2 .d Flux Th z= 234∫ (A238 − A234) 𝑧 0  dz    (5.5) where z is the depth of integration. While A234 is measured at discrete depths, A238 is often calculated assuming conservative behavior of U in seawater (Chen et al., 1986): A238 (dpm l -1 ) = 0.0704dpm l -1 x salinity   (5.6) The 234 Th flux obtained from equation 5.5 is then multiplied by the POC/ 234 Th ratio of the particles that sink from the surface and scavenge 234 Th to calculate the export flux of carbon:  167  Flux POC z= Flux Th z *(C POC z/C Th z)     (5.7) where Flux POC z and Flux Th z are the sinking fluxes of POC and 234 Th at depth z, C POC z and C Th z are the POC and 234 Th concentrations of the sinking particles at depth z. Complications arise, however, when the system is open, and 234 Th is added or removed from the area of study by advection or mixing (Buesseler et al., 1995; Bacon et al., 1996; Cochran et al., 1995), or when the system is not at steady state (e.g., at the onset or at the end of a bloom). In these situations, equation 5.5 must be modified (e.g., Savoye et al., 2006): Flux Th z= 234∫ (A238 −  A234]) 𝑧 0  dz + V – A234/t (5.8) where V is the sum of the advective and diffusive fluxes of 234 Th (dpm/m 2 /d) to or from the study site and A234/t is the change in 234 Th activity with time at the study site. Another complication arises from uncertainties regarding the (C POC z/C Th z) ratio that should be used to convert 234 Th fluxes into carbon fluxes (Buesseler et al., 2006). This ratio often increases with particle size but not always. Generally, the ratio obtained from large particles (>53um) sampled with large volume pumps is used, based on the assumption that these are the particles that actually remove 234 Th from surface waters, but very different estimates of carbon flux can be obtained depending on the method used to collect particles for measuring C POC z/C Th z.  168  Application of the 234 Th: 238 U method in coastal regions may be further complicated by other factors such as possible deviations from the 238 U-salinity relationship caused by fresh water input with variable U content (Owens et al., 2011; Rutgers van der Loeff et al., 2006; Palmer and Edmond, 1993). In situations where anoxia develops in the water column, as is the case for Saanich Inlet, possible removal of U from the water column is also a concern. There are also potential calibration problems caused by high and variable particle load on the filters used for particulate 234 Th measurements and variable recovery of dissolved 234 Th by co-precipitation due to complexation by dissolved or colloidal organic matter. It is therefore recommended to directly measure the uranium concentration of anoxic or low salinity waters, to carefully control the particle load on filters and to use a yield monitor during the co-precipitation of 234 Th from seawater samples (Rutgers van der Leoff et al., 2006).  5.4 Materials and methods 5.4.1 Study Site Saanich inlet is located in southern Vancouver Island, British Columbia, Canada (Fig. 5.1). It has a maximum depth of 230 m and is connected to Georgia Strait through Satellite Channel and a relatively broad and shallow (80 m) sill which restricts deep water renewal. River runoff at the head of the inlet is very small (Stucchi and Whitney, 1997) and the largest sources of fresh water are the Cowichan and Fraser Rivers (Fig. 5.1) adding  169  freshwater through Satellite Channel (Herlinveaux, 1962). The tidal-induced overflow of the water from Georgia Strait results in the continuous and vigorous renewal of the waters above the sill depth, but deep water renewal only takes place in late summer or fall. Wind-driven upwelling along the coast of the northeast Pacific Ocean and enhanced estuarine circulation from the freshet of the Fraser River draw dense and oxygenated waters through Juan de Fuca Strait and into the Strait of Georgia (Masson & Cummins, 1999). These waters are then able to flush the anoxic water that develops through the course of the year in the deep Saanich basin (Anderson and Devol, 1973).  Figure 5.1: Map of Saanich inlet and coastal southwestern British Columbia. The star indicates the position of the sediment trap mooring for this study. SN-9 and SN0.8 are the sediment trap mooring locations in Timothy et al. (2003).  170  Previous investigations have shown that primary production in Saanich Inlet is highly seasonal and generally higher than in neighboring Georgia Strait (Timothy and Soon, 2001). Diatoms contribute most of the yearly primary production, with occasionally significant growth of dinoflagellates and nanoflagellates (Timothy et al., 2003). 5.4.2. Sample collection, preparation and analyses 5.4.2.1 Hydrography Monthly temperature, salinity and oxygen profiles were obtained during most of the cruises (Table. 5.1) using a CTD deployed over the entire water column at the study site. Salinity and oxygen were calibrated on a regular basis with measurements from bottle samples collected at several depths during each cruise. Table 5.1: Sampling schedule for sediment traps, 234 Th deficit measurements and hydrography.  2009 2010 2011  M A M J J A S O N D J F M A M J J A S O N D J F Trap 234 Th -  x x    x     x x -   x   x x x x CTD  Grey indicates that data were obtained. White indicates no measurement for a given month. For 234 Th; (x) indicates that the data are used in the following discussion. In two instances (-) [March-09 and May-10], the data are not considered. The first set of total 234 Th samples was not spiked with 230 Th so  171  that the recovery during co-precipitation could not be estimated. The May2010 samples were damaged during shipping between UBC and IOS and the accuracy of their counts is suspect. Taken together, the 234 Th measurements yield a nearly complete monthly coverage of a full year if we take: Mar (2010), Apr (2010), May (2009), Jun (2009), Jul (gap), Aug (2010), Sep (gap), Oct (2009), Nov (2010), Dec (2010), Jan (2011), Feb (2011). This is the sequence of samples that was used to produce the time-series reported in Figs. 5.11, 5.16, 5.21 and 5.25. 5.4.2.2 Sediment traps The deployment of sediment traps began on February 10, 2009 and continued uninterrupted until February 9, 2011 (Table D1, Appendix D). The sediment traps were made from PVC tubing (Timothy and Soon, 2001) with an inside diameter of 14 cm and a height of 50 cm (aspect ratio = 3.6). In order to decrease mixing (Gardner, 1980), a baffle grid (1.5 cm squares) was placed at the opening of each trap and another approximately 10 cm above the base. The sediment traps were deployed in pairs at three depths (50 m, 115 m and 180 m) on a mooring located at 48.59°N, 123.50°W (water depth: 230 m; Fig. 5.1), and serviced on a monthly basis (Table 5.1). Concentrated sodium chloride (300 g/l) was added at the bottom of all traps before deployment and sodium azide was added to one trap of each pair to retard bacterial degradation (Knauer & Asper, 1989). Upon recovery, the samples were sieved using a Nylon screen with 200μm mesh size to remove the swimmers, centrifuged, freeze-dried and ground with a pestle and mortar for subsequent analysis. Total carbon and nitrogen were measured by gas chromatography on a model 1106 Carlo Erba CHN analyser with a precision of ±1.3% for carbon and ±2% for  172  nitrogen (Verardo et al., 1990). Inorganic carbon was measured with a Coulometrics Inc. CO2 coulometer (precision ±2%, derived from 2SD% of standards determination) and organic carbon calculated by difference. Biogenic silica was measured following the method described by Mortlock and Froelich (1989) or Müller and Schneider (1993). Samples were weighed (~ 20 mg) in centrifuge tubes. Hydrogen peroxide and HCl were added to remove organic and inorganic carbon, respectively. After removing the dissolved phase by centrifugation, opal was extracted with Na2CO3 at 85ºC in a water bath. Dissolved silica was measured colorimetrically at 812 nm on a LKB spectrophotometer. This method has a precision of about 10%. 5.4.2.3 234 Th method Th-234 disequilibrium was measured on multiple occasions during the sediment trap deployment period (Table 5.1). Sea water samples were collected with GO-FLO bottles (General Oceanics) attached to a cable and closed with a messenger. Total 234 Th was obtained by MnO2 co-precipitation on 2L samples (Buesseler et al., 2001) and filtration onto a 25mm diameter Tissuequartz filter (25mm TQ samples hereafter), while particulate 234 Th samples were obtained by filtering 4-8L of seawater onto 25mm diameter Tissuequartz filters. The samples were then counted for beta emission using a Risø low level multi-counter (DTU Nutech).    173  5.4.2.3.1 Total 234 Th: Upon return to the laboratory at the end of the day of sampling, a graduated cylinder was used to measure 2L of sea water (precision: ±10 ml) and transferred into a Nalgene bottle. The samples were acidified to pH2 with 2N HCl and spiked with 1g of a 10 dpm.g -1  solution of 230 Th in 8N HNO3. After 12 hours to allow isotopic equilibration, the pH was raised back to 8 with concentrated NH4OH before adding 100μl KMnO4 (3g/l) and 100μl MnCl2 (8g/l) to generate a MnO2 precipitate. The Nalgene bottles were then shaken vigorously and left to equilibrate for 1hr. They were subsequently heated in a water bath at 80ºC for an additional hour to reduce filtration time, following Cai et al. (2006). After cooling in an ice bath, the MnO2 precipitates were filtered onto tissue-quartz filters (25 mm diameter, 1μm nominal pore size), oven dried at ~50ºC, and covered with Polyethylene wrap (1.11mg/cm 2 ) for beta counting. Due to the relatively low counts generated by our 2 L samples, we did not cover our samples with Al foil to shield possible beta emissions from contaminants, and thus, some of our 234 Th counts may have been overestimated by 0-0.4 dpm (L. Miller, pers. comm.). The samples were counted three times in the course of 234 Th decay to verify that the beta emissions were primarily due to 234 Th. Background emissions from possible longer lived beta emitting contaminants (e.g., 40 K) were measured after ten 234 Th half-lives. Counting efficiencies were established for each counter by processing 2 L samples of acidified deep water (> 400 m, details in Chapter 6) collected in the NE Pacific at station P26 and stored in the laboratory for an extended period of time so that 234 Th was in equilibrium with 238 U. After the last count,  174  the samples were spiked with 200 mg of a solution containing 690 dpm/g of 229 Th in 8 N HNO3 and digested on a hot plate with 20ml concentrated HNO3, 5 ml concentrated HClO4 and 2 ml concentrated HF. The Th was then purified by anion-exchange to measure 230 Th (and therefore 234 Th) recovery during the Mn oxide co-precipitation, following Pike et al. (2005). This small volume method provides a fast and convenient way to measure total 234 Th and minimizes possible procedural errors caused by self-absorption when larger volumes, necessitating larger amount of MnO2, are used (Cai et al., 2006). 5.4.2.3.2 Particulate 234 Th: Two different methods were used to collect marine particles for beta counting: Approximately 6-8L of seawater were measured with a graduated cylinder (±10 ml) and filtered onto a pre-combusted (350ºC, 4hrs) TQ filter (25 mm diameter, 1 μm pore size). The samples were then oven-dried at ~50ºC. On several occasions, samples were also obtained using large volume in-situ pumps deployed at the same depths where the water samples were taken. Particles were collected by filtering relatively large volumes (generally > 100 L) of seawater onto 142 mm diameter Tissue Quartz filters with 1μm pore size. The filters were dried at ~50ºC in an oven and a 25mm diameter punch was taken for 234Th counting.  175  Both the 25mm filters and 25mm punches were covered with Polyethylene wrap and counted three times. Counting efficiencies for the particulate samples were determined using a standard 238 U/FeO filter made according to Rutgers van der Loeff and Moore (1999). Our particulate 234 Th measurements were consistent with those of other laboratories during the GEOTRACES inter-calibration exercise (Maiti et al., 2012). Because of the relatively large particle load on the filters, contribution from dissolved 234Th adsorbed on the filter during filtration was considered negligible. Several “dip” blanks were collected during the deployment of the large volume pumps and were found to have similar activities as the blank TQ filters (< 0.1 cpm). Following completion of counting, at least 8 months after the initial sample collection, POC and PON were measured by gas chromatography on a model 1106 Carlo Erba CHN analyser.  5.5 The U-salinity relationship in Saanich Inlet waters As indicated in section 5.3, when applying the 234 Th/ 238 U disequilibrium method to coastal waters, in addition to the usual precautions regarding blanks, background corrections, yield during co-precipitation, and calibration, special attention must be given to the U/salinity correlation used to estimate seawater U activity from salinity measurements. Deviations from the correlation observed in open ocean waters could arise as a result of local freshwater input with relatively high U content or, in the case of Saanich Inlet, removal of U from anoxic seawater.  176  Although the deeper waters of Saanich Inlet are anoxic, their frequent renewal suggests that their residence time in the inlet is too short to produce a substantial decrease in uranium concentration. This is corroborated by estimates of the removal rate of authigenic uranium in the anoxic sediments of Saanich Inlet (~5,000 dpm.m -2 .y -1 ; Anderson et al., 1989). If the authigenic U removed in Saanich sediment is uniformly removed from a 100m layer of anoxic water which gets renewed every year, the resulting maximum decrease in U concentration at the end of the year would be 50 dpm/m 3 , i.e., 0.05 dpm/l. If deep water renewal happens only every two years, then the expected maximum drop in concentration at the end of two years would be 0.1 dpm/l. This is relatively small compared to total U activity in Saanich water. Moreover, there is no evidence from the analysis of material collected with sediment traps that U is scavenged uniformly from the anoxic zone (Anderson et al., 1989). Instead, it seems that a large fraction of U reaching the sediment is included in particles formed in surface waters or added by an unknown mechanism at the sediment water interface. In coastal waters with relatively low salinity, as found in Saanich Inlet, deviations from the conventional U-salinity relationship could be potentially important if there is a local source of U-enriched freshwater. Dissolved uranium concentrations measured in the water column of Saanich Inlet in April 2009 (Amini and Holmden, pers comm) are significantly higher than expected from salinity and equation 5.6 (Table 5.2), suggesting that Saanich Inlet waters are influenced by a freshwater end-member with a high U concentration. Assuming conservative mixing with a seawater end-member with a salinity of 35 and a U  177  concentration of 2.46 dpm/l (35 x 0.0704), we can calculate the U concentration of the freshwater end-member needed to explain the observed concentrations (Table 5.2): A[U]rw = (35 x (A[U]meas – (0.0704dpm l -1  x S))) / (35 – S)  (5.9) where S is the salinity of the sample. Table 5.2: Dissolved U concentrations in Saanich Inlet (123°30.2’N; 48°34.6’W) measured on 10 April, 2009 (M. Amini and C. Holmden, pers. comm), and estimated based on two different U – salinity relationships (see text for explanations). Depth Salinity [O2] [U]meas  A[U]meas  A[U]s  A[U]rw  A[U] m psu mM ppb* dpm.L -1 * dpm.L -1 ** nmol.kg -1#  dpm.L -1##  10 29.959 382.5 3.01 2.20 2.11 3.5 2.34 115 30.906 10.3 3.25 2.37 2.18 9.3 2.37 200 31.329 1.2 3.13 2.28 2.21 - 2.38 (*) Measured (**) Calculated from equation 5.6 (#) [U] in the freshwater end-member calculated by assuming conservative mixing between a seawater end-member with a salinity of 35‰ and a uranium concentration of 2.46 dpm.l -1  (equation 5.9) (##) Calculated from equation 5.10 The calculated freshwater end-member value is lower for the surface (10m) sample, which could reflect a local source (e.g., the Cowichan River) mostly affecting the salinity of surface water, while a more regional source is affecting the sample taken at 115 m. For the latter, the Fraser River is the most obvious potential source. However, the calculated value is much higher than obtained in preliminary measurements of Fraser River water (0.5-1 nmol.kg -1 ; Peuker-Ehrenbrink and Voss, pers comm), suggesting another U source,  178  possibly associated with groundwater. Interestingly, the U concentration at 200 m is ~ 0.1 dpm/l lower than at 115m, possibly reflecting removal of U from bottom waters by diffusion into anoxic sediments. Considering the very limited number of uranium concentration measurements from Saanich Inlet available to date and our ignorance regarding the source of the U-enriched freshwater end-member, the 234 Th deficit was calculated using two different approaches to provide an estimate of the error that could result from uncertainties in the exact U-salinity relationship. In the first approach, we calculate the 234 Th deficit using equation 5.5. As shown in Table 5.2, this calculation yields U concentrations that are 0.1 – 0.2 lower than observations. Alternatively, we derived a regional U-salinity relationship based on the U concentration and salinity measured at 115m, which yields a freshwater end member with a U concentration of 2.20 ppb or 1.66 dpm/l (or 9.3 nmoles.kg -1 ; Table 5.2): A[U] (dpm L -1 ) = (1.66dpm l -1  x ((35-S)/35)) + (2.46dpm l -1  x S/ 35) (5.10) where S is the salinity of the sample. This calculation slightly overestimates the U content of bottom waters, possibly depleted in U by diffusion in sediments, and in surface water, possibly diluted by a more local source of freshwater with lower U concentration.    179  5.6 Results and discussion 5.6.1 Salinity, temperature, density and O2 The lowest surface salinities in Saanich Inlet occur in winter when local precipitation is highest (Fig. 5.2a). The temperature profiles reflect seasonal heating (Fig. 5.2b) with maximum surface temperatures between June and August and minimum between December and February. The entire water column at the study site is stratified year round with very shallow mixing layers (Fig. 5.2c; Timothy and Soon, 2001). The temperature of the deep water in the basin varies little from month to month. Small increases in salinity and density in late summer or early fall are the results of deep water renewal events (Anderson and Devol, 1973).  180   2 3 2 3 2 3 2 4 2 4 2 4 2 5 2 5 26 2 6 2 6 2 7 2 7 2 7 2 8 2 8 2 8 2 9 2 9 2 9 3 0 30 3 0 3 0 2 3 2 3 2 3 2 4 2 4 2 4 2 5 2 5 2 5 2 3 2 3 2 3 2 3 2 3 2 6 2 6 2 6 7 7 7 2 4 2 4 2 4 2 4 2 4 2 4 2 8 2 8 2 8 2 5 2 5 2 5 2 5 2 5 2 5 2 6 2 6 2 6 2 6 2 6 2 6 2 9 2 9 2 7 2 7 2 7 2 7 7 2 7 2 8 2 8 2 8 2 8 2 8 2 8 2 9 2 9 2 9 2 9 2 9 2 9 31 31 3 1 3 0 3 0 30 3 0 3 0 3 0 3 0 3 0 3 0 31 31 3 1 3 1 3 1 3 1 Cruise Month D e p th  ( m ) Salinity SI   Apr/2009 Jun/2009 Aug/2009 Oct/2009 Dec/2009 Feb/2010 Apr/2010 Jun/2010 Aug/2010 Oct/2010 Dec/2011 Feb/2011 20 40 60 80 100 120 140 160 180 200 23 24 25 26 27 28 29 30 31 6 6 6 8 8 8 6 6 6 6 6 6 6 8 8 8 8 8 8 8 8 8 8 8 10 10 10 1 0 1 0 10 12 1 212 14 1 0 16 12 Cruise Month D e p th  ( m ) Temperature SI   Apr/2009 Jun/2009 Aug/2009 Oct/2009 Dec/2009 Feb/2010 Apr/2010 Jun/2010 Aug/2010 Oct/2010 Dec/2011 Feb/2011 20 40 60 80 100 120 140 160 180 200 5 6 7 8 9 10 11 12 13 14 15 16  181    Figure 5.2: Time-series hydrographic data from surface to 200m during the 2 year period: (a) Salinity (psu), (b) temperature (°C), (c) density () and (d) oxygen (ml/L). Anoxia develops in the water column below the sill depth (Fig. 5.2d) and has been attributed mainly to high primary production, which results in high export flux of carbon, 1 9 1 9 20 2 0 2 0 21 2 1 2 1 2 2 2 2 2 2 2 2 23 2 3 2 3 2 3 1 9 1 9 1 9 2 0 2 0 2 0 1 9 1 9 1 9 1 9 1 9 1 9 1 2 1 2 1 2 0 2 0 2 0 2 0 2 0 2 0 2 1 2 1 2 121 2 1 2 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 22 2 3 2 3 23 2 3 2 3 232 3 2 3 2 3 2424 2 4 2 4 2 4 2 4 24 2 4 2 4 Cruise Month D e p th  ( m ) Density SI   Apr/2009 Jun/2009 Aug/2009 Oct/2009 Dec/2009 Feb/2010 Apr/2010 Jun/2010 Aug/2010 Oct/2010 Dec/2011 Feb/2011 20 40 60 80 100 120 140 160 180 200 18.5 19 19.5 20 20.5 21 21.5 22 22.5 23 23.5 24 1 1 1 2 2 2 3 3 3 3 3 1 1 1 1 1 2 2 2 2 2 3 3 3 3 3 4 4 4 4 4 4 4 4 4 5 5 5 5 6 66 7 7 4 5 8 5 9 8 3 4 4 6 9 610 6987 10 65 Cruise Month D e p th  ( m ) Oxygen SI   Apr/2009 Jun/2009 Aug/2009 Oct/2009 Dec/2009 Feb/2010 Apr/2010 Jun/2010 Aug/2010 Oct/2010 Dec/2011 Feb/2011 20 40 60 80 100 120 140 160 180 200 1 2 3 4 5 6 7 8 9 10 11  182  and weak estuarine circulation, which limits lateral export of phytoplankton biomass from the inlet (Timothy and Soon, 2001). High primary productivity in Saanich Inlet appears to be mostly sustained by lateral transport of nutrients from Georgia Strait (Timothy and Soon, 2001). The boundary of the anoxic zone ([O2] < 10 μmol/kg), defined by Kamykowski and Zentara, 1990) is generally between 120m and 130m in the winter seasons but rises to shallower depths in late summer and early fall, as anoxic water is pushed upward by denser water overflowing from the sill. The oxygen concentration of deep water does not increase substantially during these events because of the presence of reducing chemical species (H2S, CH4, Mn 2+ , etc.) that react with the incoming oxygen. 5.6.2 Sediment traps 5.6.2.1: Sample mass and concentrations The weight of material collected for each deployment and the concentration of organic carbon (POC), total nitrogen (PON), inorganic carbon (PIC) and biogenic silica (BSi) are reported in Table D1, Appendix D. Only POC and BSi are discussed in this chapter. Calcium carbonate (CaCO3) contributed less than 2% of the mass flux on average as foraminifera and coccolithophorids are rare. The percentage of PIC was only used to estimate organic carbon from total carbon measured with the CN Analyzer. Nitrogen (PON) was used to calculate the C/N ratio of the settling material. As indicated above, sediment traps were deployed in pairs at each depth. A concentrated NaCl solution was added to the bottom of the two traps and NaN3 was added to one of  183  them to retard bacterial degradation (Knauer & Asper, 1989). However, there is no significant difference in the % POC measured by the two sets of sediment traps (%POC NaN3 -%POC NaCl  = 0 ± 1.5 %), corroborating the findings of Timothy et al. (2003). From January 2010, the samples from the non-NaN3 treated traps were used to measure oxygen isotopes in biogenic silica (De Baere et al., in prep). The data plotted in the figures of this chapter are therefore just from the NaN3 treated traps. Seasonal variations in % opal and % organic matter are shown in Fig. 5.3. Percent organic matter was estimated from %POC x 1.85, the conversion factor used in Saanich Inlet by Timothy et al. (2003). Biogenic silica concentrations show a clear seasonal pattern with high concentrations in spring, a gradual decrease through summer and minimum values in winter. The settling material intercepted by the trap deployed at 50 m tends to have higher opal concentration, reflecting either dissolution or dilution by laterally transported material in the deeper traps. On the other hand, organic matter concentrations reach a maximum in late summer, which is observed at all depths in 2009 but only in the shallow trap in 2010. The relatively low concentration of organic matter collected during the spring bloom is likely due to dilution by biogenic silica. The higher concentrations in late summer suggest a shift in plankton community toward species other than diatoms.  184   Figure 5.3: Seasonal changes in (a) Opal% and (b), OM% in the material collected by sediment traps at 3 depths in Saanich Inlet. Previous studies conducted in Saanich Inlet have revealed relatively high OC/N ratio in the organic matter collected with sediment traps (Timothy and Soon, 2001; Timothy et al., 2003). Our results (Fig. 5.4) confirm this observation, yielding a molar OC/N ratio of 8.7 which is consistent with value of 8.5 reported at the head of Saanich Inlet by Timothy et al., 2003. This relatively high ratio must reflect addition of terrigenous organic matter (with a high C/N ratio), as would be expected for coastal waters. Terrigenous organic matter contributions are expected to be higher in winter when productivity is low and 0.00% 10.00% 20.00% 30.00% 40.00% 50.00% 60.00% Feb-09 May-09 Aug-09 Dec-09 Mar-10 Jun-10 Sep-10 Jan-11 O p al %  Seasonal compositional characteristics of settling particles in Saanich Shallow Trap @ 50m Middle Trap @ 115m Deep Trap @ 180m 0.00% 10.00% 20.00% 30.00% 40.00% Feb-09 May-09 Aug-09 Dec-09 Mar-10 Jun-10 Sep-10 Jan-11 O M %  Shallow Trap @ 50m Middle Trap @ 115m Deep Trap @ 180m  185  continental run-off is highest and are expected to be lowest during the spring blooms. Lower C/N ratios are found in spring (Fig. 5.4a). Samples collected in summer 2009 had higher C/N than those collected in summer 2010, suggesting higher input of terrigenous organic matter in 2009. 5.6.2.2 Fluxes Fluxes of total mass, biogenic silica, and organic carbon recorded at the three depths between February 2009 and February 2011 are shown in Figs. 5.5, 5.6 and 5.7, respectively, and reported in Table D2; Appendix D. Opal and organic carbon fluxes are clearly highest in spring at all depths, as a result of diatom blooms. In comparison, seasonal changes in total mass fluxes are muted, particularly at 50 m (Fig. 5.5a). Total mass fluxes are also relatively high between November 2009 and January 2010. Since the increase in opal and organic carbon fluxes was less pronounced during that period, the winter increase in total mass fluxes can be attributed to an increase in the flux of lithogenic material, possibly associated with higher run-off or resuspension of sediments from the nearby sill. The main component of biogenic Si in Saanich is diatom frustules (Sancetta & Calvert, 1988). Particulate organic carbon fluxes at 50 m showed a seasonal pattern similar to that of Biogenic Si (Fig. 5.6a, 5.7a), despite the fact that terrigenous and other marine OM (e.g., flagellates and zooplankton) were also caught by the sediment traps. Both BSi and POC fluxes roughly reflected the yearly cycle of primary production (Timothy et al., 2003), increasing in March–April and decreasing in September–October. Although the flux of BSi exhibits an rapid decrease in June-July, the POC flux still  186  remained at high levels until after September, supporting the possibility that flagellates play an increasingly important role after the spring diatom bloom (Hobson and Mcquoid, 2001; Timothy and Soon, 2001).  Figure 5.4: (a) Seasonal variations in the OC/N ratio of sediment trap material collected at the three depths. (b) OC versus N and OC/N values for all the sediment-trap samples. Weight percent N is multiplied by 12/14, so the slope of the regression line provides the estimated OC/N molar ratio. 3 4 5 6 7 8 9 10 11 12 13 Dec-08 Mar-09 Jul-09 Oct-09 Jan-10 May-10 Aug-10 Nov-10 Feb-11 M o la r O C /N  Shallow Trap @ 50m Middle Trap @ 115m Deep Trap @ 180m OC = 8.7124 N + 0.2462 R² = 0.9378 0 5 10 15 0 0.5 1 1.5 O C %  N% * (12/14)  187   Figure 5.5: Mass fluxes recorded by sediment traps deployed at three different depths over the 2-year time series. The data from shallow, middle and deep traps are highlighted in 5.5(a), 5.5(b) and 5.5(c), respectively. 0 1 2 3 4 5 6 7 8 9 Feb-09 May-09 Aug-09 Dec-09 Mar-10 Jun-10 Sep-10 Jan-11 M as s F lu x  ( g /m 2 /d ay ) Mass fluxes recorded by sediment traps deployed at 3 different depths Shallow Trap @ 50m Middle Trap @ 115m Deep Trap @ 180m 0 1 2 3 4 5 6 7 8 9 Feb-09 May-09 Aug-09 Dec-09 Mar-10 Jun-10 Sep-10 Jan-11 M as s F lu x  ( g /m 2 /d ay ) Shallow Trap @ 50m Middle Trap @ 115m Deep Trap @ 180m 0 1 2 3 4 5 6 7 8 9 Feb-09 May-09 Aug-09 Dec-09 Mar-10 Jun-10 Sep-10 Jan-11 M as s F lu x  ( g /m 2 /d ay ) Shallow Trap @ 50m Middle Trap @ 115m Deep Trap @ 180m  188   Figure 5.6: Opal fluxes recorded by sediment traps deployed at three different depths over the 2-year time series. The data from shallow, middle and deep traps are highlighted in fig. 5.6(a), 5.6(b) and 5.6(c), respectively. 0 0.5 1 1.5 2 2.5 3 3.5 Feb-09 May-09 Aug-09 Dec-09 Mar-10 Jun-10 Sep-10 Jan-11 O p al  F lu x  ( g /m 2 /d ay ) Opal fluxes recorded by sediment traps deployed at 3 different depths Shallow Trap @ 50m Middle Trap @ 115m Deep Trap @ 180m 0 0.5 1 1.5 2 2.5 3 3.5 Feb-09 May-09 Aug-09 Dec-09 Mar-10 Jun-10 Sep-10 Jan-11 O p al  F lu x  ( g /m 2 /d ay ) Shallow Trap @ 50m Middle Trap @ 115m Deep Trap @ 180m 0 0.5 1 1.5 2 2.5 3 3.5 Feb-09 May-09 Aug-09 Dec-09 Mar-10 Jun-10 Sep-10 Jan-11 O p al  F lu x  ( g /m 2 /d ay ) Shallow Trap @ 50m Middle Trap @ 115m Deep Trap @ 180m  189   Figure 5.7: POC fluxes recorded by sediment traps deployed at three different depths over the 2-year time series. The data from shallow, middle and deep traps are highlighted in fig. 5.7(a), 5.7(b) and 5.7(c), respectively. 0 0.1 0.2 0.3 0.4 0.5 0.6 Feb-09 May-09 Aug-09 Dec-09 Mar-10 Jun-10 Sep-10 Jan-11 P O C  F lu x  ( g /m 2 /d ay ) POC fluxes recorded by sediment traps deployed at 3 different depths Shallow Trap @ 50m Middle Trap @ 115m Deep Trap @ 180m 0 0.1 0.2 0.3 0.4 0.5 0.6 Feb-09 May-09 Aug-09 Dec-09 Mar-10 Jun-10 Sep-10 Jan-11 P O C  F lu x  ( g /m 2 /d ay ) Shallow Trap @ 50m Middle Trap @ 115m Deep Trap @ 180m 0 0.1 0.2 0.3 0.4 0.5 0.6 Feb-09 May-09 Aug-09 Dec-09 Mar-10 Jun-10 Sep-10 Jan-11 P O C  F lu x  ( g /m 2 /d ay ) Shallow Trap @ 50m Middle Trap @ 115m Deep Trap @ 180m  190  Fluxes are generally highest at 115m depth and the contrast in flux with water depth is larger for total mass than for biogenic silica and organic carbon. A similar trend was observed by Timothy et al. (2003) and attributed to the re-suspension of sediment from the sill and lateral transport at mid-depth into Saanich Inlet. Table 5.3: The comparison of fluxes for mass (a), biogenic opal (b) and organic carbon (c) between this study and those reported by Timothy et al. (2003), in which the depth deployments of the traps were 50m (shallow), 135m (mid) and 180m (deep) at station SN-0.8 (48.55° N, 123.55° W) and 45m (shallow), 110m (mid) and 150m (deep) at station SN-9 (48.67° N, 123.51° W). 5.3a: Annual mass fluxes (g/m 2 /d) Stations This study (48.59°N; 123.50°W) SN-0.8 SN-9 Sampling period 03/2009-02/2010 03/2010-02/2011 01/1984-12/1989 08/1983-12/1989 Shallow Trap 3.09 2.63 1.63 5.67 Mid Trap 4.91 4.31 2.33 11.1 Deep Trap 3.87 3.83 2.40 12.7  5.3b: Annual biogenic opal fluxes (g/m 2 /d) Stations This study (48.59°N; 123.50°W) SN-0.8 SN-9 Sampling period 03/2009-02/2010 03/2010-02/2011 01/1984-12/1989 08/1983-12/1989 Shallow Trap 1.02 0.76 0.669 1.84 Mid Trap 1.41 1.09 0.750 2.70 Deep Trap 1.06 0.95 0.680 2.80   191   5.3c: Annual OC fluxes (g/m 2 /d) Stations This study (48.59°N; 123.50°W) SN-0.8 SN-9 Sampling period 03/2009-02/2010 03/2010-02/2011 01/1984-12/1989 08/1983-12/1989 Shallow Trap 0.229 0.21 0.168 0.351 Mid Trap 0.294 0.232 0.178 0.543 Deep Trap 0.214 0.206 0.177 0.498  The mass fluxes between March/2010 and February/2011 are slightly lower than fluxes measured between March/2009 and February/2010 (Table 5.3). Since the fluxes of biogenic silica and organic carbon are also slightly lower during the second year, this difference is at least partly due to slightly lower export productivity. This difference is however small compared to the difference in annual fluxes between this study and the earlier sediment trap study conducted in Saanich Inlet between January 1984 and December 1989 (Timothy et al., 2003). During this study, two sediment trap moorings were deployed in the deep basin of Saanich Inlet (Fig. 5.1), one near the sill (SN-9) and one near the head of the inlet (SN 0.8), while our sediment trap mooring was located between these two stations at the center of the inlet (Fig. 5.1). Fluxes for total mass, biogenic opal and organic carbon at our station fall between the fluxes measured at SN-0.8 and at SN-9. This trend coincides with the location of the stations which suggests that fluxes increase towards the sill at the mouth of the inlet. This increase has been  192  attributed both to higher primary production (Timothy and Soon, 2001) and closer proximity to the source of sediment from the sill (Timothy et al., 2003). 5.6.3 234 Th deficits Th-234 deficits in the water column were measured several times between March 2009 and February 2011 (Table 5.1; Table D3, Appendix D). Samples from May/2010 were damaged during shipment while the Mar/2009 samples were not spiked with 230 Th. These two sets of samples are not considered in the following discussion. By combining the measurements conducted on Mar/2010, Apr/2010, May/2009, Jun/2009, Aug/2010, Oct/2009, Nov/2010, Dec/2010, Jan/2011 and Feb/2011 we can obtain a nearly complete composite picture of the annual cycle of 234 Th deficit and partitioning between dissolved and particulate phases. The samples were counted long enough to obtain counting errors within 3-8% for all the samples. The errors associated with counting efficiency and decay curve regression resulted in a final propagated error of about 8% for total 234 Th and 15% for particulate 234 Th. Because dissolved 234 Th was calculated by difference and the fraction of 234 Th dissolved is generally smaller than the fraction adsorbed on particles, the relative error on dissolved 234 Th is large and variable. The average dissolved 234 Th calculated by difference in this study is 0.26 ± 0.21 dpm/l (1 standard deviation, n = 97) and the absolute error is 0.13 ± 0.05 dpm/l. The average relative error on dissolved 234 Th is thus 56%.  193  Particulate samples were also collected by Large Volume Pump (LVP) in Mar/2009, Dec/2010 and Jan/2011 on 150mm Tissuequartz filter (Table D4, Appendix D). Subsamples (LVP samples hereafter) were punched from the 150 mm filters for 234 Th analysis, which provided a comparison to the results from the particulate samples obtained by small volume filtration onto 25 mm diameter TQ filters (25mm TQ samples). 5.6.3.1 Total 234 Th profiles Total 234 Th activities generally show large deficits in the water column (Fig. 5.8; Table D3, Appendix D) compared to the 238 U activities derived from equations 5.6 and 5.10, indicating that the particle flux in Saanich Inlet is nearly always high enough to scavenge a large fraction of the 234 Th before it decays.   194   Figure 5.8: Total 234 Th activities measured in Mar/2010, Apr/2010, May/2009, Jun/2009, Aug/2010, Oct/2009, Nov/2010, Dec/2010, Jan/2011 and Feb/2011. The 238 U activities are estimated using equation 5.6 [U-238 (a)] and 5.10 [U-238 (b)]. The averages of all data measured at a given depth +/– one standard deviation are shown in Figure 5.9. In general, total 234 Th activities are lower in the upper 100 m of the water column (~ 0.5 – 1 dpm/l) and higher but more variable towards the bottom. 0 20 40 60 80 100 120 140 160 180 200 0 0.5 1 1.5 2 2.5 3 D ep th  ( m ) Total 234Th activity  (dpm/L) Dec Nov Oct Aug Jun May Apr Mar Feb Jan U-238 (a) U-238 (b)  195   Figure 5.9: Average total 234 Th measured at a given depth (dpm/l). The two black lines show the one standard deviation envelope. The 238 U activities are estimated using equation 5.6 [U-238 (a)] and 5.10 [U-238 (b)]. Total 234 Th activities in the upper 100m tend to be somewhat higher in late winter (Fig. 5.10a). The deficit is however still pronounced, indicating continued scavenging by particles. In early spring, total 234 Th in the upper 100 m decreases, indicating enhanced scavenging during the spring bloom. This trend is clearly illustrated by the May-09 and April-10 profiles (Fig. 5.10b). The latter, however, show significantly higher total 234 Th in deeper water. In late summer or early fall, total 234 Th profiles tend to show intermediate values in the upper 100 m but high values in deeper water (Fig. 5.10c). The intermediate total 234 Th activity in the upper 100 m is consistent with the intermediate fluxes generally 0 20 40 60 80 100 120 140 160 180 200 0 0.5 1 1.5 2 2.5 3 D ep th  ( m ) Total 234Th (dpm/L) U-238 (a) U-238 (b)  196  recorded at that time by the sediment traps (Fig. 5.5), while the high 234 Th measured at depth could be attributed to resuspension of recently deposited sediment as a result of deep water renewal in the deep basin of the inlet.   0 20 40 60 80 100 120 140 160 180 200 0.00 0.50 1.00 1.50 2.00 2.50 D ep th  ( m ) Total 234Th (dpm/L) Jan-11 Feb-11 0 20 40 60 80 100 120 140 160 180 200 0.00 0.50 1.00 1.50 2.00 2.50 D ep th  ( m ) Total 234Th (dpm/L) May-09 Apr-10  197   Figure 5.10: Total 234 Th during (a) late winter, (b) early spring, (c) late summer or early fall. The thin black lines are the extrema of total 234 Th from Figure 5.9.  Figure 5.11: Monthly composite time-series of total 234 Th. Numbers on the horizontal axis correspond to the months of the year. The figure was drawn by combining the measurements conducted on Mar/2010, Apr/2010, May/2009, Jun/2009, Aug/2010, Oct/2009, Nov/2010, Dec/2010, Jan/2011 and Feb/2011. 0 20 40 60 80 100 120 140 160 180 200 0.00 0.50 1.00 1.50 2.00 2.50 D ep th  ( m ) Total 234Th (dpm/L) Aug-10 Oct-09 0.4 0 .6 0.6 0. 8 0 .8 0 .8 0.8 0.8 0.8 0 .8 0.8 1 1 1 1 1.2 1.2 1.2 1. 4 1.40 .8 1. 6 1 0 .6 0.6 1.2 0 .6 0.6 1.4 1.8 1 1.6 0 .8 1 1.8 2 1 2 Cruise Month D e p th  ( m ) Total Th234 SI (dpm/L)   3 4 5 6 7 8 9 10 11 12 1 2 20 40 70 100 115 130 160 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2  198  These general trends can also be seen in the composite picture of the monthly changes in total 234 Th over an entire year (Fig. 5.11). We clearly see the lower total 234 Th activities in spring and higher values in deep water over most of the year, particularly during the spring bloom and the late summer deep water renewal. 5.6.3.2 Particulate 234 Th profiles Total 234Th activity is partitioned between dissolved (< 1 μm) and particulate (> 1 μm) phases. In Saanich Inlet, the particulate 234 Th activity is generally higher than the dissolved 234 Th activity. Averaging all available data at a given depth (Fig. 5.12) indicates that the fraction of particulate 234 Th increases from 48% at the surface to 85% near the oxic/anoxic interface and decreases back to 60% towards the bottom.  Figure 5.12: Average fraction of particulate 234 Th (% of total 234 Th) as a function of water depth. Particulate 234 Th activities vary from ~ 0.2–1.7 dpm/l (Fig. 5.13) and generally increase with depth (Fig. 5.14). Temporal changes in particulate 234 Th generally follow those observed for total 234 Th (Fig. 5.15; Fig. 5.16). The increasing trend with depth is observed 0 20 40 60 80 100 120 140 160 180 200 0 20 40 60 80 100 D ep th  ( m ) % particulate 234Th  199  through the year and particulate 234 Th dominates the deep total 234 Th maxima observed in spring and late summer.  Figure 5.13: Particulate 234 Th activities over an annual cycle. The 238 U activities are estimated using equation 5.6 [U-238 (a)] and 5.10 [U-238 (b)].  Figure 5.14: Average particulate 234 Th measured at a given depth (dpm/l). The two black lines show the averages ±one standard deviation. 0 20 40 60 80 100 120 140 160 180 200 0 0.5 1 1.5 2 2.5 3 D ep th  ( m ) Particulate 234Th activity  (dpm/L) Dec Nov Oct Aug Jun May Apr Mar Feb Jan U-238 (a) U-238 (b) 0 20 40 60 80 100 120 140 160 180 200 0.0 0.5 1.0 1.5 2.0 2.5 D ep th  ( m ) Particulate 234Th (dpm/L)  200     Figure 5.15: Particulate 234 Th during (a) late winter, (b) early spring, (c) late summer or early fall. The thin black lines are the extrema of particulate 234 Th from Fig. 5.14. 0 20 40 60 80 100 120 140 160 180 200 0.0 0.5 1.0 1.5 2.0 2.5 D ep th  ( m ) Particulate 234Th (dpm/L) Jan-11 Feb-11 0 20 40 60 80 100 120 140 160 180 200 0.0 0.5 1.0 1.5 2.0 2.5 D ep th  ( m ) Particulate 234Th (dpm/L) May-09 Apr-10 0 20 40 60 80 100 120 140 160 180 200 0.0 0.5 1.0 1.5 2.0 2.5 D ep th  ( m ) Particulate 234Th (dpm/L) Aug-10 Oct-09  201   Figure 5.16: Monthly composite time-series of particulate 234 Th. Numbers on the horizontal axis correspond to the months of the year. The figure was drawn by combining the measurements conducted on Mar/2010, Apr/2010, May/2009, Jun/2009, Aug/2010, Oct/2009, Nov/2010, Dec/2010, Jan/2011 and Feb/2011.  Figure 5.17: Average fraction of dissolved 234 Th (% of total 234 Th) as a function of water depth. 0. 4 0.4 0. 4 0 .4 0 .4 0.6 0. 6 0.6 0.6 0 .6 0. 6 0 .8 0. 8 0 .8 0.8 0 .8 1 1 1 .2 1 .2 0. 8 1 1 .4 1 0. 8 1.2 1 .6 1.4 Cruise Month D e p th  ( m ) Particulate Th234 SI (dpm/L)   3 4 5 6 7 8 9 10 11 12 1 2 20 40 70 100 115 130 160 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 0 20 40 60 80 100 120 140 160 180 200 0 10 20 30 40 50 60 D ep th  ( m ) % dissolved 234Th  202   Figure 5.18: Dissolved 234 Th activities over an annual cycle. The 238 U activities are estimated using equation 5.6 [U-238 (a)] and 5.10 [U-238 (b)].  Figure 5.19: Average dissolved 234 Th profile (dpm/l). The two black lines show the average ± one standard deviation. 0 20 40 60 80 100 120 140 160 180 200 0 0.5 1 1.5 2 2.5 3 D ep th  ( m ) Dissolved 234Th (dpm/L) Dec Nov Oct Aug Jun May Apr Mar Feb Jan U-238 (a) U-238 (b) 0 20 40 60 80 100 120 140 160 180 200 0.00 0.50 1.00 1.50 2.00 2.50 D ep th (m ) Dissolved 234Th (dpm/L)  203     Figure 5.20: Dissolved 234 Th during (a) late winter, (b) early spring, (c) late summer or early fall. The thin black lines are the extrema of dissolved 234 Th from Fig. 5.19. 0 20 40 60 80 100 120 140 160 180 200 0.0 0.5 1.0 1.5 2.0 2.5 D ep th  ( m ) Dissolved 234Th (dpm/L) Jan-11 Feb-11 0 20 40 60 80 100 120 140 160 180 200 0.0 0.5 1.0 1.5 2.0 2.5 D ep th  ( m ) Dissolved 234Th (dpm/L) Apr-10 May-09 0 20 40 60 80 100 120 140 160 180 200 0.0 0.5 1.0 1.5 2.0 2.5 D ep th  ( m ) Dissolved 234Th (dpm/L) Aug-10 Oct-10  204  5.6.3.3 Dissolved 234 Th profiles The average fraction of total 234 Th in the dissolved pool decreases from 53% in surface water to 15% near the oxic/anoxic interface and then increases to 38% towards the bottom (Fig. 5.17). Dissolved 234 Th activities vary from below detection limits to ~ 0.8 dpm/l (Fig. 5.18) and are generally lowest near the oxic/anoxic boundary (Fig. 5.19). Seasonal variations in dissolved 234 Th are not as clear as for total or particulate 234 Th, as a result of lower activities and larger errors on the measurement (Fig. 5.20). The clearest seasonal signal is found in the upper 50 m of the water column where dissolved 234 Th is lowest in spring and highest in winter, following primary production (Fig.5.21). The minimum dissolved 234 Th at mid-depth appears to be a year-round feature (Fig. 5.20), but attenuated during deep water renewal (Fig. 5.21).  Figure 5.21: Monthly composite time-series of dissolved 234 Th. Number on the horizontal axis correspond to the months of the year. The figure was obtained by combining the measurements 0 0. 1 0.1 0. 1 0.1 0.1 0 .1 0. 2 0.2 0. 2 0 .2 0.2 0 .20 .2 0 .3 0.3 0. 3 0.3 0. 3 0.3 0.3 0.2 0 .2 0. 2 0 .2 0.3 0.3 0.3 0 .3 0. 4 0 .4 0 0 0 0.2 -0.1 -0.1 0.5 0.3 0. 4 0 0 0 .4 -0.2 0.5 0.1 0.1 0.5 0.6 0.4 0 .4 0 .3 0.4 Cruise Month D e p th  ( m ) Dissolved Th-234 SI (dpm/L)   3 4 5 6 7 8 9 10 11 12 1 2 20 40 70 100 115 130 160 0 0.1 0.2 0.3 0.4 0.5 0.6  205  conducted on Mar/2010, Apr/2010, May/2009, Jun/2009, Aug/2010, Oct/2009, Nov/2010, Dec/2010, Jan/2011 and Feb/2011. 5.6.3.4 POC/ 234 Th ratio The POC/ 234Th ratio (μgC/dpm) of particles collected in Saanich Inlet generally varies between ~ 30 and 120 gC/dpm (Fig. 5.22), except for four abnormally high values (20m in Mar/2010, 80m in Apr/2010, 10m in Nov/2010 and 100m in Dec/2010), possibly due to the accidental inclusion of zooplankton in the samples. Neglecting these four samples, the average POC/ 234 Th ratio generally decreases with depth with a possible secondary maximum near the oxic/anoxic boundary (Fig. 5.23). The POC/ 234 Th ratios observed in Saanich Inlet are within the range reported elsewhere (Buesseler et al., 2006) and are not particularly high. While we might expect that particles from productive coastal waters would have relatively high POC/ 234 Th because of their high POC concentration, the ratios measured in Saanich Inlet are only slightly higher than those measured in spring in the surface waters of station Papa (Chapter 5). A possible explanation is the addition of lithogenic particles by rivers, which could lower the POC concentration of particles in Saanich Inlet, while keeping the adsorption of dissolved 234 Th (Buesseler et al., 2006). Seasonal changes in the POC/ 234 Th ratio are well defined above the oxic/anoxic interface with higher ratios during the spring bloom, lower ratios in the winter and intermediate ratios in late summer or early fall (Fig. 5.24; 5.25). Below 120 m, the seasonal variability  206  is less pronounced. Ratios of POC/ 234 Th are lower than in shallower water, but there is a hint of slightly higher ratios in late summer (Fig. 5.25). In particular, the samples from the deep basin collected in October-2009 (Fig. 5.24c) have higher ratios, which could possibly reflect sediment resuspension of high POC particles during deep water renewal.  Figure 5.22: POC/ 234 Th (gC/dpm) on particulate samples over an annual cycle.  Figure 5.23: Average of all the POC/ 234 Th measured in particles collected at a given depth (μg/dpm). The two black lines show the average ± one standard deviation. 0 20 40 60 80 100 120 140 160 180 200 0 100 200 300 400 500 600 D ep th  ( m ) POC/234Th on particulate samples (μg/dpm) Dec Nov Oct Aug Jun May Apr Mar Feb Jan 0 20 40 60 80 100 120 140 160 180 200 0 20 40 60 80 100 120 D ep th  ( m ) POC/234Th on particulate samples (μg/dpm)  207     Figure 5.24: POC/ 234 Th during (a) late winter, (b) early spring, (c) late summer or early fall. The thin black lines are the standard deviations of POC/ 234 Th from Fig. 5.23. 0 20 40 60 80 100 120 140 160 180 200 0.0 20.0 40.0 60.0 80.0 100.0 120.0 D ep th  ( m ) POC/234Th on particulate samples (μg/dpm) Jan-11 Feb-11 0 20 40 60 80 100 120 140 160 180 200 0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0 D ep th  ( m ) POC/234Th on particulate samples (μg/dpm)  May-09 Apr-10 0 20 40 60 80 100 120 140 160 180 200 0.0 20.0 40.0 60.0 80.0 100.0 120.0 D ep th  ( m ) POC/234Th on particulate samples (μg/dpm) Aug-10 Oct-09  208  High POC/ 234 Th ratios appear sporadically (e.g., Feb-11 at 115m; Fig. 5.24.a) and result from relatively high POC concentrations (Table D3; Appendix D). Such variability could be due to the inclusion of zooplankton or large aggregates in the samples and could be eliminated by pre-filtering the samples with a coarser mesh filter (e.g., Nitex: 53 m). The very high POC/ 234 Th ratio measured at 115 m in November 2010 (Fig. 5.25) could in part reflect such an occurrence. On the other hand, relatively high POC/ 234 Th between 70 and 120 m could also reflect slow upwelling of fine particles resuspended from the sediments during deep water renewal.  Figure 5.25: Monthly composite time-series of POC/ 234 Th. Numbers on the horizontal axis correspond to the months of the year. The figure was obtained by combining the measurements conducted on Mar/2010, Apr/2010, May/2009, Jun/2009, Aug/2010, Oct/2009, Nov/2010, Dec/2010, Jan/2011 and Feb/2011. Rarely sampled depths (10m and 180m) and samples suspected to be affected by capture of zooplankton during filtration are excluded from this figure). 30 30 30 40 40 4 0 50 5 0 5 0 50 5 0 50 60 60 60 60 60 60 70 70 7 0 7 0 70 70 7 0 80 8 0 80 8 0 6 060 6 0 6 0 5 0 5 0 7 0 7 0 80 8 0 9 0 90 50 50 9 0 90 90 80 50 1 0 0 70 70 80 6 0 9 0 110 8 0 70 6 0 4 0 8 0 3 0 80 8 0 70 6 0 100 120 70 Cruise Month D e p th  ( m ) POC/Th-234 SI (ug/dpm)   3 4 5 6 7 8 9 10 11 12 1 2 20 40 60 70 80 100 115 130 160 30 40 50 60 70 80 90 100 110 120  209  5.6.3.5 Comparison between LVP samples and 25mm TQ samples. POC and particulate 234 Th were measured on three occasions (March 2009, December 2010, and January 2011) using samples collected with large volume in-situ pumps (Table D4, Appendix D). The 234 Th and POC profiles are generally similar to those obtained by small volume filtration, but with some significant differences (Fig. 5.26).  Figure 5.26: Comparison between POC, particulate 234 Th and POC/ 234 Th obtained by collecting particles with Large Volume in-situ Pumps (LVP) and small volume filtration, as described in the method section.  210  With the exception of one sample collected at 100m during the March-2009 sampling, when an abnormally high POC value was obtained with the LVP, the agreement between the two methods for POC concentration is reasonable. On the other hand, particulate 234 Th activity is consistently higher on the LVP filters (by 22 ± 28% (1 SD)), resulting in somewhat lower POC/ 234 Th ratios. The reason for this difference may be due to adsorption of dissolved 234Th onto the filter. The ‘dip’ blanks used to correct for 234Th adsorption on the LVP filters were obtained by counting the filters loaded on LVPs which were deployed to the same depth but had 0 liter water passed through. Therefore they were very low and may underestimate dissolved 234 Th adsorption during filtration. Because the samples obtained with the in-situ pumps are few, the carbon fluxes discussed below are all based on the samples obtained by small volume filtration. 5.6.3.6 Fluxes derived from 234 Th: 238 U disequilibria 5.6.3.6.1 234 Th fluxes 234 Th fluxes were first calculated using equation 5.5, which assumes a closed system at steady state. The 234 Th: 238 U deficits were calculated by integrating the difference between measured total 234 Th activities and 238 U activities using the trapezoid approximation of Equation 5.5. U-238 activities were calculated from salinity using equation 5.6 and 5.10. The difference provides an estimate of the uncertainty stemming from not having measured 238 U activity in all samples.  211  The 234 Th fluxes increase nearly linearly with depth and are high year-round compared to the open ocean sites (e.g., station Papa in Chapter 6). Th-234 fluxes calculated with equation 5.6 are 15 % lower than those calculated with equation 5.10 (Fig. 5.27). Improving the accuracy of 234 Th fluxes to better that 15% would thus require measuring 238 U in seawater more extensively than has been done in this study.  0 20 40 60 80 100 120 140 160 180 200 0.0 0.2 0.4 0.6 0.8 1.0 D ep th  ( m ) 234Th fluxes in May-09 (dpm/cm2/day) 0 20 40 60 80 100 120 140 160 180 200 0.0 0.2 0.4 0.6 0.8 1.0 D ep th  ( m ) 234Th fluxes in Jun-09 (dpm/cm2/day) 0 20 40 60 80 100 120 140 160 180 200 0.0 0.2 0.4 0.6 0.8 D ep th  ( m ) 234Th fluxes in Oct-09 (dpm/cm2/day)  0 20 40 60 80 100 120 140 160 180 0.0 0.2 0.4 0.6 0.8 D ep th  ( m ) 234Th fluxes in Mar-10 (dpm/cm2/day)  0 20 40 60 80 100 120 140 160 180 0.0 0.2 0.4 0.6 0.8 D ep th  ( m ) 234Th fluxes in Apr-10 (dpm/cm2/day)  0 20 40 60 80 100 120 140 160 180 0.0 0.2 0.4 0.6 0.8 D ep th  ( m ) 234Th fluxes in May- 10(dpm/cm2/day)   212   Figure 5.27: 234 Th fluxes calculated from estimates of 238 U seawater concentration obtained from equation 5.6 (blue line) and 5.10 (red line). Seasonal changes in the fluxes of 234 Th are relatively small but well defined. Fig. 5.28 shows the average 234 Th flux profile (± 1 standard deviation). The relative standard deviation at different depths varies between 6 and 13%. Within this relatively small range, however, there is a clear seasonal signal. The 234 Th fluxes measured in late winter are at the lower range of measured fluxes (Fig. 5.29a), while those measured during the spring 0 20 40 60 80 100 120 140 160 180 0.0 0.2 0.4 0.6 0.8 D ep th  ( m ) 234Th fluxes in Aug-10 (dpm/cm2/day)  0 20 40 60 80 100 120 140 160 180 0.0 0.2 0.4 0.6 0.8 D ep th  ( m ) 234Th fluxes in Nov-10 (dpm/cm2/day)  0 20 40 60 80 100 120 140 160 180 0.0 0.2 0.4 0.6 0.8 D ep th  ( m ) 234Th fluxes in Dec-10 (dpm/cm2/day)  0 20 40 60 80 100 120 140 160 180 0.0 0.2 0.4 0.6 0.8 D ep th  ( m ) 234Th fluxes in Jan-11 (dpm/cm2/day)  0 20 40 60 80 100 120 140 160 180 0.0 0.2 0.4 0.6 0.8 D ep th  ( m ) 234Th fluxes in Feb-11 (dpm/cm2/day)   213  bloom are at the upper range (Fig. 5.29a). Fluxes measured in late summer are intermediate (Fig. 5.29b). Linear fits of the data shown in Fig. 5.29a yield: Jan-11: 234 Th flux = 34.1 10 -4  z  r 2  = 0.997 Feb-11: 234 Th flux = 33.8 10 -4  z  r 2  = 0.999 May-09: 234 Th flux = 44.7 10 -4  z  r 2  = 0.997 Apr-10: 234 Th flux = 42.3 10 -4  z  r 2  = 0.988 234 Th fluxes during spring blooms are thus on average 27% higher ([(44.7 + 42.3) / (34.1 + 33.8)] = 1.27) than in winter.  Figure 5.28: Average calculated 234 Th flux profile (red line) ± 1 standard deviation (black lines), assuming steady state. Fluxes were calculated from 238 U based on equation 5.5. The blue line represents the maximum 234 Th flux (234 A238) by assuming 234 Th is completely depleted. 0 20 40 60 80 100 120 140 160 180 200 0.0 0.2 0.4 0.6 0.8 D ep th  ( m ) dpm/cm2/d 234Th fluxes (average ± 1 SD)  214    Figure 5.29: (a) 234 Th fluxes in late winter and during the spring bloom, (b) 234 Th fluxes in late summer and early fall. The thin black lines are the standard deviations of the average 234 Th fluxes from Fig. 5.28. 0 20 40 60 80 100 120 140 160 180 200 0.0 0.2 0.4 0.6 0.8 D ep th  ( m ) 234Th fluxes (dpm/cm2/d) Jan-11 Feb-11 May-09 Apr-10 0 20 40 60 80 100 120 140 160 180 200 0.0 0.2 0.4 0.6 0.8 D ep th  ( m ) 234Th fluxes (dpm/cm2/d) Aug-10 Oct-09  215  Since 234 Th profiles were measured during consecutive months on several occasions (Table 5.1), the non-steady-state model (equation 5.8, assuming V = 0) can be used to refine our calculation of 234 Th flux in June-09, April-10, December-10, January-11 and February-11. We used the equation of Buesseler et al. (1992), also reported in Gustafsson et al. (2004): Flux Th z= 234 (zi –zi+1) ((A238 (1- e -t ) + 1 A234 e -t  – 2A234) / (1 – e -t )    (5.11) where zi and zi+1 are the depths of integration, 1 A234 and 2 A234 are the 234 Th activity measured at the start and end of the sampling period under consideration, and t is the length of that period (days). For each of these months, the 234 Th fluxes derived from equation 5.11 are very close to the fluxes calculated with the steady state model (Fig. 5.30). The fluxes calculated with the non-steady state model are within 15% of those calculated with the steady state model (Fig. 5.31). The main reason for the reasonable accuracy obtained with the steady-state model is the generally low total 234 Th activities in Saanich Inlet water, so that the relative changes in activity between consecutive months are small compared to the large 234 Th deficit.  216  Figure 5.30: Th-234 fluxes calculated with the steady state model vs the non-steady state model (equation 5.11). Blue diamonds and red squares are 234 Th fluxes obtained from 238 U seawater  217  concentration estimates obtained from equation 5.6 and 5.10, respectively. The black line is the 1:1 relationship.  Figure 5.31: Ratio of fluxes estimated with the non-steady state model to those estimated with the steady state model from 238 U seawater concentration estimates obtained from equation 5.5. In our calculation of 234 Th fluxes with equation 5.11, we assumed that exchanges of 234 Th between Saanich Inlet and surrounding areas are negligible. Timothy et al. (2003) indicated that primary production is higher within the inlet than outside. We would therefore expect some level of addition of total 234 Th from Georgia Strait into Saanich Inlet. We have not measured 234 Th outside Saanich Inlet and cannot ascertain whether this is the case. However, if there is addition of 234 Th into Saanich Inlet from neighboring regions, it should be small considering the low total 234 Th concentration typically found in these coastal waters.  0 20 40 60 80 100 120 140 160 180 200 0.4 0.6 0.8 1 1.2 1.4 D ep th  ( m ) 234Th fluxes - non steady state/steady state Jun-09 Apr-10 Dec-10 Jan-11 Feb-11  218  5.6.3.6.2 POC fluxes Particulate organic carbon (POC) fluxes were calculated by multiplying the 234 Th fluxes obtained with the steady state model at a given depth and the POC/ 234 Th ratio of the particles collected at the same depth (equation 5.7). The average POC fluxes increase with depth from the surface to the oxic/anoxic interface (Fig. 5.32). Below the oxic/anoxic interface, the POC fluxes decrease slightly but they increase again towards the bottom. In open ocean waters, the flux of organic carbon generally increases with depth within the euphotic zone, where the rate of photosynthesis exceeds the rate of respiration, and then decreases with depth as a result of organic matter remineralization (e.g., Bacon et al., 1996; Chapter 6). Since the depth of the euphotic zone in Saanich Inlet is always shallower than 20 m (Timothy and Soon, 2001), the increase in organic carbon flux with depth down to 120 m can only be explained by lateral transport of organic matter that has been previously deposited at shallower depths for a period of time long enough to allow decay of 234 Th that had been scavenged during sinking to the initial site of deposition.  219   Figure 5.32: Average POC fluxes derived at a given depth (μmol/cm 2 /d) with the steady state model from 238 U seawater concentration estimates obtained from equation 5.5. The two black lines show the average ±one standard deviation.  0 20 40 60 80 100 120 140 160 180 200 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 D ep th  ( m ) POC fluxes (μmol/cm2/d) 0 20 40 60 80 100 120 140 160 180 200 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 D ep th  ( m ) POC fluxes (μmol/cm2/d) Jan-11 Feb-11 May-09 Apr-10  220   Figure 5.33: (a) POC fluxes in late winter and during the spring bloom, (b) POC fluxes in late summer or early fall. The thin black lines are the standard deviations of POC fluxes from Fig. 5.32. There are clear seasonal variations in POC flux from the surface to the oxic/anoxic interface (Fig. 5.33, 5.34). Late winter and spring bloom POC fluxes are at the lower and higher ends of the POC flux range, respectively (Fig. 5.33a), while POC fluxes during late summer are intermediate (Fig. 5.33b). Seasonal variations in POC fluxes are less clear in deeper waters. The fluxes in deeper waters tend to be lower during winter and spring and higher in late summer/early fall. The latter may reflect resuspension of sediments by deep water renewal. 0 20 40 60 80 100 120 140 160 180 200 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 D ep th  ( m ) POC fluxes (μmol/cm2/d) Aug-10 Oct-09  221   Figure 5.34: Contour plot of the seasonal and depth variation in POC fluxes derived with the steady state model from 238 U seawater concentration estimates obtained from equation 5.5. Timothy et al. (2003) attempted to estimate the export ratio (e-ratio) of organic carbon (the ratio of the settling flux of POC to net production in the euphotic zone) in Saanich Inlet by dividing annually averaged primary production measurements obtained by 14 C incubations (Timothy and Soon, 2001) by the annually averaged flux of POC measured with a trap deployed at 50 m depth at stations SN-9 and SN-0.8 (Fig. 5.1). In doing so, they found surprisingly low e-ratios (0.12 at SN-9; 0.09 at SN0.8). Although they attempted to correct for the presence of terrigenous organic matter in the trap samples using 13C, they made the assumption that the flux of organic carbon intercepted by the 50 m trap originated entirely from the overlying surface water. However, the increase in 234 Th-based POC flux between 20 m and 50 m (i.e. below the euphotic zone) indicates that some organic matter intercepted by the 50 m traps may have been laterally 0. 5 0.5 1 1 1 1 1 1 1 .5 1. 5 1 .5 1.5 1 .5 1.5 1 .5 2 2 2 2 2 2.5 2.5 2.5 2 2 2.53 2.5 3 2.5 2 1 1.5 3 3 .5 Cruise Month D e p th  ( m ) POC Flux SI (umol/cm2/day)   3 4 5 6 7 8 9 10 11 12 1 2 20 40 60 70 80 100 115 130 160 0.5 1 1.5 2 2.5 3 3.5  222  transported instead of being exported from overlying surface water. This would yield even lower e-ratios. Primary production was not measured at our study site, but if we take the average between Timothy and Soon’s sites of study (SN-9: 13 mol/cm2/d; SN-0.8: 9 mol/cm2/d) and compare it to our 234Th-based POC flux at 20m (0.5 mol/cm2/d) we find an e-ratio of 0.5/11 = 0.045. This must be taken as a maximum since we did not account for contribution of terrigenous organic matter in the export flux. This suggests that there is extensive nutrient recycling in the surface waters of Saanich inlet (or that sediment traps and the 234 Th deficit method are missing an important part of the sinking flux (see Chap 6)).  Figure 5.35: Average POC fluxes derived from 234 Th deficit assuming steady state from 238 U seawater concentration estimates obtained from equation 5.6 and average POC fluxes measured by sediment traps. The two black lines show the average ±one standard deviation for the fluxes derived from 234 Th. The error bars on the average trap-recorded POC fluxes represent ± one standard deviation.  0 20 40 60 80 100 120 140 160 180 200 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 D ep th  ( m ) POC fluxes (μmol/cm2/d) Th-234 Trap  223   Figure 5.36: Two year time series records of POC fluxes measured with sediment traps and corresponding fluxes derived from 234 Th deficits, using the steady state model and 238 U seawater concentration estimates obtained from equation 5.5. Sediment trap fluxes at 50 m are compared to average 234 Th-based fluxes measured at 40m and 60/70m. Sediment trap fluxes at 180 m are compared to 234 Th-based fluxes measured at 180 m or 160m when no samples were obtained at 180m. 5.6.4 Comparing POC flux measured with sediment traps and 234 Th deficits. The POC fluxes derived from 234 Th assuming steady state are in reasonable agreement with the POC fluxes measured with sediment traps (Fig.5.35). The average POC flux 0 1 2 3 4 Feb-09 May-09 Aug-09 Dec-09 Mar-10 Jun-10 Sep-10 Jan-11 Apr-11 μ m o l/ cm 2 /d  Trap 50m Trap 115m Trap 180m Th-234 50m* 0 1 2 3 4 Feb-09 May-09 Aug-09 Dec-09 Mar-10 Jun-10 Sep-10 Jan-11 Apr-11 μ m o l/ cm 2 /d  Trap 50m Trap 115m Trap 180m Th-234 115m 0 1 2 3 4 Feb-09 May-09 Aug-09 Dec-09 Mar-10 Jun-10 Sep-10 Jan-11 Apr-11 μ m o l/ cm 2 /d  Trap 50m Trap 115m Trap 180m Th-234 180m*  224  measured at 50 m with 234 Th is lower than the flux intercepted by the shallow trap, but the two methods still agree within a factor of 2 and their ranges overlap. The difference between the two methods at 50 m is further illustrated by plotting the two data sets versus time (Fig. 5.36). The agreement between the two methods is better in deeper water, particularly at mid-depth, but on occasion 234 Th-based carbon fluxes are more than twice the corresponding sediment traps carbon fluxes (Fig. 5.36). These occurrences (115m in March-2010, November-2010, Feburary-2011 and 180m in October-2009) are the results of high POC/ 234 Th and could be attributed to the capture of zooplankton on the filters. If correct, a simple way to avoid this problem would be to screen the seawater samples before filtration (although this may also point to a more serious problem with the 234 Th method that will be discussed in Chapter 6). Fig 5.36 also confirms that at 50 m, the 234 Th method produces flux estimates that are significantly lower than those obtained with sediment traps. At this point, the reason for this difference is unclear. The difference could be explained if the 50 m trap intercepted more than the vertically settling flux (i.e. over trap) or if the 234 Th method underestimates the vertical POC flux. Lateral advection or mixing of 234 Th from Saanich Inlet into Satellite Channel or Georgia Strait could result in such an underestimation, but confirmation requires measurements of 234 Th in these regions. This explanation seems unlikely, considering the low dissolved 234 Th activities that are expected in these waters. Alternatively, the difference between the two methods could be explained by focusing of the POC flux between surface and 50 m depth. In fjords, the sinking flux of particles is increasingly focused with depth due to the gradual decrease in the width of the fjord (Wassmann, 1991; Timothy et al., 2003). Particle focusing thus  225  increases the flux of refractory particles with depth (by a factor inversely proportional to the decrease in the inlet’s width with depth) and mitigate (or reverse) the decrease in flux with depth of particles subjected to remineralization or dissolution. On the other hand, particle focusing is not affecting the POC/ 234 Th ratios of sinking particles but increases the activity of particulate 234 Th with depth. This would result in a decrease in POC flux estimated from 234 Th with depth, as long as sinking and focusing occurs rapidly compared to the half-life of 234 Th. These contrasting responses to focusing by the two methods could help explain the difference in fluxes recorded at 50 m. In addition to focusing, the increase in particle flux with depth in the upper 120 m of the water column could also be due to lateral transport of particles at depth. This process is different from sediment “focusing” discussed above. Focusing is syndepositional (i.e. it occurs as particles settle) but lateral transport is postdepositional (i.e. it occurs after initial deposition at another site). If particles were first deposited at shallower depth (e.g., on the sill) for a period of time long enough to allow decay of the excess 234 Th they scavenged from the water column, their subsequent resuspension, lateral transport and sinking would result in a gradual increase in particle flux with depth which would be recorded by both methods. The sediment traps would intercept both the vertical and lateral flux, while lateral transport of “aged” particles would not affect particulate 234Th concentration in seawater but would increase the POC/ 234 Th ratio of the particles. The better agreement between the POC fluxes estimated from 234 Th and measured by sediment traps deployed at mid depth (Fig. 5.35) suggests that the increase in flux with depth between 50 m and 120 m recorded by the 234 Th deficit method is mainly due to this mechanism. This stands  226  to reason since the depth of the sill, probably the main source of material supplied laterally to Saanich Inlet, is at 70 m depth. 5.6.5 Can POC fluxes in coastal regions be estimated by only measuring particulate 234 Th and POC? The activity of dissolved 234 Th in Saanich Inlet is consistently low (Fig. 5.18, 5.19) and is almost completely depleted at mid depth (Fig. 5.21). Seasonal variability in dissolved 234 Th is thus also very small and difficult to precisely measure by difference (Fig. 5.20). This suggests that one might obtain as good an estimate of carbon sinking fluxes in coastal waters by only measuring 234 Th and POC on particles. If such is the case, the approach could provide a relatively simple and rapid method to survey and monitor the regional sinking flux of carbon or any other particle constituent in the coastal zone. To assess the potential of this approach, we calculated carbon fluxes from our data, assuming that dissolved 234 Th is zero or always equal to the average activity measured during this study (0.26 ± 0.21 dpm/l). In this case, equation 5.5 becomes: Flux Th z= 234∫ (𝐴238 − 𝑃𝐴234) 𝑧 0  dz     (5.12) or Flux Th z= 234∫ (𝐴238 − 𝑃𝐴234 − 0.26) 𝑧 0  dz    (5.13)  227  where p A234 is the activity of particulate 234 Th. In Saanich Inlet, average seawater activities are (Table D3; Appendix D): A238 = 2.26 ± 0.11 dpm/l A234 = 0.92 ± 0.34 dpm/l p A234 = 0.66 ± 0.32 dpm/l Therefore, if we assume that dissolved 234 Th = 0 (equation 5.12), we would overestimate the 234 Th flux (and the POC flux) by: [2.26 (±0.11) – 0.66 (±0.32)] ----------------------------------- = 1.19 ± 0.41 [2.26 (±0.11) – 0.92 (±0.34)] On the other hand, equation 5.13 would evidently give the right flux but with an uncertainty of 41%. [2.26 (±0.11) – 0.66 (±0.32) – 0.26] ------------------------------------------- = 1.00 ± 0.41      [2.26 (±0.11) – 0.92 (±0.34)] To further assess the potential of this approximation, equations 5.12 and 5.13 were applied to each data point reported in Table D3 in Appendix D to calculate organic carbon fluxes which were then be compared to the fluxes obtained with equation 5.5. The results confirm that equation 5.12 overestimates the flux of organic carbon reaching 180m by 20% (Fig. 5.37a). While equation 5.13 provides an accurate mean flux at 180m, it overestimates fluxes in the upper 50m by ~ 10% (Fig. 5.37b) because the dissolved 234 Th concentrations at these depths are higher than 0.26 dpm/l (Fig. 5.19). If we were to  228  substitute 0.26 in equation 5.13 by the mean dissolved 234 Th concentration in the upper 50 m, the equation would evidently provide a more accurate 234 Th flux estimate.   Figure 5.37: (a) Ratios of organic carbon fluxes obtained with equation 5.12 and 5.5; (b) ratios of organic carbon fluxes obtained with equation 5.13 and 5.5. The 234 Th fluxes are derived from 238 U seawater concentration estimates obtained from equation 5.6 for both cases. 0 20 40 60 80 100 120 140 160 180 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 D ep th  ( m ) POC fluxes - Simplified model (Equation 5.12) / Steady state model (Equation 5.5) AVG Dec Nov Oct Aug Jun May Apr Mar Feb Jan 0 20 40 60 80 100 120 140 160 180 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 D ep th  ( m ) POC fluxes - Simplified Model (Equation 5.13)/steady state model (Equation 5.5) AVG Dec Nov Oct Aug Jun May Apr Mar Feb Jan  229  The purpose of this exercise is to show that in productive coastal waters, measuring total 234 Th by the time consuming MnO2 co-precipitation method in every sample collected will not increase significantly the precision and accuracy of the calculated fluxes. Collecting particles by filtering 4-8 L of seawater to measure particulate 234 Th and POC concentration, complemented by a few measurements of total 234 Th in the seasons with highest and lowest primary productivities to constrain the mean value of total 234 Th would suffice. The time saving that this approach affords would provide a means to map and routinely monitor the POC fluxes over large swaths of coastal waters. To further illustrate the validity of this simplified approach, organic carbon fluxes measured with sediment traps, equation 5.5 and equation 5.13 are compared on figure 5.38 and 5.39. They show that the agreement between sediment traps and fluxes derived from equation 5.5 is similar to the agreement between sediment traps and fluxes derived from equation 5.13, confirming that simply measuring systematically 234 Th and POC on suspended particles only, complemented by a few measurements of total 234 Th to estimate its mean concentration, suffices to assess the seasonal amplitude of annual POC flux in the coastal zone.   230   Figure 5.38: Comparison of organic carbon fluxe time series obtained with sediment traps, equation 5.5 (colored dots) and equation 5.13 (black dots) from 238 U seawater concentration estimates obtained from equation 5.6 for all cases. 0 1 2 3 4 Feb-09 May-09 Aug-09 Dec-09 Mar-10 Jun-10 Sep-10 Jan-11 Apr-11 μ m o l/ cm 2 /d  Trap 50m Trap 115m Trap 180m Th-234 50m Th-234 50m * 0 1 2 3 4 Feb-09 May-09 Aug-09 Dec-09 Mar-10 Jun-10 Sep-10 Jan-11 Apr-11 μ m o l/ cm 2 /d  Trap 50m Trap 115m Trap 180m Th-234 115m Th-234 115m * 0 1 2 3 4 Feb-09 May-09 Aug-09 Dec-09 Mar-10 Jun-10 Sep-10 Jan-11 Apr-11 μ m o l/ cm 2 /d  Trap 50m Trap 115m Trap 180m Th-234 160/180m Th-234 160/180m *  231   Figure 5.39: (a) POC fluxes estimated from equation 5.5 and 238 U seawater concentration estimated from equation 5.6 versus POC fluxes from the sediment traps; (b) POC fluxes estimated from equation 5.13 and 238 U seawater concentration estimated from equation 5.6 versus POC fluxes from the sediment traps. The data points influenced by high POC/ 234 Th (115m in March-2010, November-2010, Feburary-2011 and 180m in October-2009) possibly reflect capture of zooplankton or other large particles on the filters are marked green. The surface data points (shown in Fig. 5.38), which show lower POC fluxes estimated from 234 Th than from sediment traps (excluding October-2009, March-2010, April-2010 and December-2010) are marked blue. Those two sets of data are not included in the linear regression. y = 1.0121x R² = 0.6166 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 P O C  f lu x es  f ro m  2 3 4 T h  ( μ m o l/ cm 2 /d ) POC fluxes from sediment traps (μmol/cm2/d) POC fluxes - derived from equation 5.5 vs from sediment traps y = 1.0004x R² = 0.4536 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 P O C  f lu x es  f ro m  2 3 4 T h  ( μ m o l/ cm 2 /d ) POC fluxes from sediment traps (μmol/cm2/d) POC fluxes - derived from equation 5.13 vs from sediment traps  232  5.7 Conclusions and future perspectives. A 2-year time series of monthly sediment trap sampling provides a continuous record of POC flux which can be compared to concurrent estimates obtained from 234 Th deficits in the water column. The organic carbon fluxes obtained by both methods are similar. In agreement with previous sediment trap experiments conducted in Saanich Inlet (Francois, 1988; Timothy et al., 2003), the present data indicate that POC fluxes vary in concert with surface water primary productivity, and particle fluxes are highest at mid-depth as a result of lateral transport of sediment winnowed from the sill by tidal currents. The present study also confirms a gradual increase in particle and POC flux from the head of the inlet towards the sill. The similarity of the fluxes measured by sediment traps and 234 Th deficits does not prove but at least supports the validity of both approaches for measuring POC flux in coastal waters. Compared more closely, the two approaches indicate that the agreement is better in deeper waters. At 50 m, 234 Th-based fluxes are lower than those measured by sediment traps, which could in part be due to particle focusing resulting from the narrowing of the fjord with depth. On the other hand, the greater POC fluxes and better agreement between the two methods at 115 m is more consistent with lateral transport of sediment from the sill into the deep basin. While these results are encouraging and supporting the use of either of these methods to document the sinking flux of POC or any other constituents of marine particles in the coastal zone, their applicability is still limited by the fact that they are both relatively  233  labor intensive and require substantial resources which can limit their application to large scale monitoring programs. The results from this study, however, suggest a possible alternative more readily applicable to larger surveys and monitoring programs, which might be necessary, for instance, to ground-truth satellite data. The simplified method that we propose stems from the observations that dissolved 234 Th activity in coastal waters is very low and varies little with season and location. Moreover, a large fraction of 234 Th in coastal waters is associated with particles. Because dissolved 234 Th activities are low, it is difficult to measure this quantity precisely by subtracting particulate 234 Th from total 234 Th. Consequently, there is little point in measuring the total 234 Th activities of coastal waters, which requires co-precipitation by manganese oxide. Instead, simply measuring particulate 234 Th by filtering a few liters of seawater and measuring the associated POC suffice to estimate of the sinking flux of POC (or any other particle constituents). Although a few measurements of dissolved 234 Th might still be needed to establish its approximate concentration in the region and/or season of interest, this simplification halves counting times and essentially eliminates the need for processing seawater samples. As such, the method becomes much more amenable to conducting larger scale monitoring or survey programs. In regions where sediment resuspension or lateral transport is important, export fluxes could be obtained by measuring samples collected in the mixed layer or the euphotic zone. The other advantage afforded by the high scavenging intensity in coastal waters is that the closed, steady state model is generally applicable because there is little 234 Th in seawater to be transported by advection and mixing and because 234 Th activities are generally low  234  year-round. In the present study, the residence time of 234 Th in the upper 100 m of the water column in Saanich Inlet varied between 7 and 14 days ([0.5 – 1 dpm/l] / [0.065 dpm/l.d-1]). The spacing between stations needed for a larger scale survey would depend on the lateral eddy diffusion coefficient in the region of study. The extent of lateral mixing varies widely in the coastal zone and it is beyond the scope of this chapter to address this question. However, a model tracking particles at 75 m depth in Georgia Strait suggests that particles would disperse over a radius of about 10 km over a 7 day period (E. Snauffer and S. Allen; pers. comm.). If this holds for Saanich Inlet, the 234 Th method may not be able to resolve the particle flux gradient between the head and the mouth of Saanich Inlet that has been clearly documented with sediment traps in Timothy et al. (2003). In this case, the good agreement between the fluxes estimated from sediment traps and from 234 Th in this study may be fortuitous and due to the location of the study site, where the flux measured by the sediment traps happens to be close to the average for the entire inlet. Measuring POC and 234 Th in particles collected along the main axis of the inlet would answer this question.      235   Chapter 6 Particle fluxes and dynamics in the northeast Pacific Ocean from paired water column measurements of Th-230 and Th-234 activity  6.1 Introduction Gaining a better understanding of the processes that control the sequestration of atmosphere CO2 to the deep sea is essential to predict the future evolution of atmospheric CO2 in response to human activities (combustion of fossil fuels, cement production, increasing land-use and deforestation). Large international research programs (e.g., JGOFS) have been established to address many aspects of this question and investigate the fluxes of particulate organic carbon (POC), carbonate, and biogenic silica (opal) to the bottom of the ocean. In particular, the deployment of sediment traps greatly improved our understanding of the geographic and vertical variations in particle flux and composition (Honjo et al., 2008). Nevertheless, many questions still remain, especially regarding the factors that control changes in particle flux, re-mineralization of organic matter, and dissolution of biominerals in the mesopelagic zone of the ocean (i.e.. the depth interval  236  between the bottom of the euphotic zone and ~ 1500 m). In this chapter, the activity of two thorium isotopes with widely different half-lives ( 234 Th, 24.1 d; 230 Th, 76.7 ky) are measured in seawater and particles of different sizes in the upper 1600 m of the water column at station Papa in an attempt to address some of these questions. Thorium-234 and Thorium-230 are produced uniformly in seawater by decay of their parent uranium isotopes ( 238 U and 234 U, respectively). Thorium is very insoluble in seawater, and the two isotopes are quickly scavenged from the water column by adsorption on the surface of settling particles, resulting in lower seawater Th activities compared to their parents. Because of its short half-life, 234 Th is mainly removed from seawater by radioactive decay. Measureable 234 Th deficits with respect to the activity of its parent 238 U require relatively high particle fluxes and 234 Th: 238 U disequilibria are more commonly observed in the euphotic zone (e.g., Buesseler et al., 2006 and references therein). In contrast, 230 Th, with its much longer half-life (76.7 ky), is almost entirely removed from seawater by scavenging and its seawater activity is always much lower than that of 234 U. Combining thorium isotopes with different half-lives is commonly used to study particle dynamics and scavenging in the water column (e.g., Bacon and Anderson, 1982; Bacon et al., 1985; Clegg et al., 1991; Cochran et al., 1993; Murname et al., 1994; Marchal and Lam, 2012). Typically, the activities of several isotopes among 230 Th, 228 Th, and 234 Th are measured in seawater and between two operationally defined classes of particle: small suspended particles (1 - 53 m), and large sinking particles (> 53 m; Fig. 6.1).  237   Figure 6.1: Model (A) used to describe thorium cycling in seawater. P = Th production in the water column from U decay; D = activity of dissolved Th; F = activity of thorium adsorbed on fine suspended particles; L = activity of thorium adsorbed on large sinking particles; λ = Th decay constant; K1 and K-1 are the adsorption and desorption rate constants between dissolved and fine particles; B1 and B-1 are the aggregation and disaggregation rate constants between fine and large particles; S = sinking rate of large particles. In this study, 234 Th and 230 Th were measured at several depths in the upper 1600m of the water column (Table 6.1) at station Papa (50°N, 145°W) to try to estimate changes in the flux of particle constituents (organic carbon, P, Ca, Al) as a function of depth in the mesopelagic zone. The particle dynamics model reported in Fig. 6.1 was also systematically modified to fit the observations. Once the particle dynamics model was established, it was used to refine our understanding of changes in particle flux with depth and to assess possible ways of estimating particle flux and remineralization in the mesopelagic zone from thorium isotope measurements. Station Papa was chosen for this  238  exploratory study on the ground that the profiles of dissolved 230 Th are nearly linear, indicating that the impact of deep water circulation on the 230 Th profiles is small (e.g., Francois, 2007).  6.2 Materials and methods 6.2.1 Sample collection and preparation The samples for this study were collected in June 2010 in the Northeast Pacific at Ocean Station Papa (OSP; 50ºN, 145ºW) in the Alaska Gyre. Thorium-234 and Thorium-230 activities were measured between 10 m and 1600 m in filtered seawater, fine (1-53μm) and large particles (>53μm) collected with large volume in-situ pumps, and a few “extra-large” particles collected with a 236 μm zooplankton MultiNet (Table 6.1). 6.2.1.1 234 Th samples Ten-liter sea water samples were collected at each depth (Table 6.1) using a rosette sampling system and drained into acid-cleaned containers.  239   240  Total 234 Th: The methods described by Pike et al. (2005) and modified by Cai et al. (2006) were used. Two liters of sea water from each container were measured with a graduated cylinder and transferred into a Nalgene bottle. After adjusting the pH to 2 with 2N HCl, a pre-weighed and calibrated 230 Th spike (~1g, ~10dpm/g) was added to each sample and left to equilibrate for 12 hours. Thereafter, the pH was raised back to 8 with concentrated NH4OH, and 100μl KMnO4 (3g/L) and 100μl MnCl2 (8g/L) were added to produce MnO2 for co-precipitation. The Nalgene bottles were shaken vigorously and left to equilibrate for 1hr, before heating in a water bath for an additional 1hr. The samples were then rapidly cooled in an ice bath and the MnO2 collected on tissue-quartz (TQ hereafter) filters (25mm diameter, 1μm pore size), oven-dried at ~50ºC and covered with LDPE film (1.11 mg/cm 2 ) for beta counting. Because of the relatively weak activity of the 2 L samples, the samples were not covered with Al foil to shield possible beta emissions from contaminants and our 234 Th counts may be slightly overestimated. Particulate 234 Th: Several methods were used to collect particles of different sizes for 234 Th analysis: Small volume filtration (>1μm): After 2L of sea water were transferred to a Nalgene bottle for total 234 Th analysis, the volume of the remaining sea water (~8L) was measured with a graduated cylinder and filtered on a pre-combusted (350ºC, 4hrs)  241  TQ filter (25mm diameter, 1μm pore size). The samples were then oven-dried at ~50º C and covered with LDPE film for beta counting. Large volume in-situ pump filtration: Large volume, in-situ pumps were used at several depths (Table 6.1) to collect particles on 142mm diameter, 1μm pore size Supor filters for 234 Th (and 230 Th) measurements. A Nylon mesh (53μm pore size) was placed on top of the Supor filters to collect the larger particles. Generally, over 100 L of sea water was filtered during each deployment. The Supor samples (1μm-53μm) were dried at ~50ºC and a 25mm diameter sub-sample was punched from each filter, covered with LDPE film and beta counted. The large particles collected on the Nylon mesh were recovered by sonication in distilled water and filtration on 25mm diameter Supor filters (1μm pore size). These Supor filters (marked as “mesh” in Table 6.2 and hereafter) were then mounted with LDPE film for 234 Th beta counting (> 53μm). MultiNet: Because large, rapidly sinking particles are very rare in the water column, it is very difficult to obtain a statistically meaningful sample even with large volume in-situ pumps. An opportunity arose from having on board a HydroBios MultiNet large plankton sampler, which was used to collect zooplankton samples and provided an opportunity to assess this sampling technique as a means of collecting large sinking particles from larger volumes of seawater. The HydroBios MultiNet was thus used in an attempt to collect “extra-large” (XL) sinking particles (>236 μm) at 5 depth intervals: 0-200 m, 200-500 m, 500-800 m, 800-1200 m and 1200-1600 m. This sampling method collected large, rapidly sinking particles from much larger volumes  242  of seawater (61m 3 , 84m 3 , 84m 3 , 111m 3  and 119m 3 , respectively). The rate of ascent of the nets was slowed down to 0.5m/s to allow live zooplankton, which are normally captured by the device, to escape. Visual inspection of the samples confirmed the absence of large living zooplankton in the samples which appeared to consist mostly of passively settling larger aggregates. Even if some smaller zooplankton were caught in the net, they would not significantly affect the Th activity of the sample. Upon recovery, the samples were preserved with 200 l of saturated HgCl2 and stored at 4º C before analysis. Back in the laboratory, the samples were filtered on HCl-cleaned (2N, Seastar) Supor filters (25mm diameter, 1μm pore size). The samples were then oven-dried at ~50ºC and covered with LDPE film for beta counting. 6.2.1.2 POC The fine particulate 234 Th samples obtained by small volume filtration (1μm, on 25mm TQ filters) were used for POC/PON analysis after completion of the 234 Th counting. Particulate organic carbon on the LVP samples (1μm-53μm LVP Supor samples and >53 μm, Mesh samples) and XL samples (>236μm, Multinet) were estimated from the phosphorus content measured in the aliquots taken after the total digestion of the Supor filters (see below). 6.2.1.3 230 Th samples Dissolved 230 Th: Approximately 20L of sea water were collected at each depth (Table 6.1; 40 L were collected at depths shallower than 400m because of their low 230 Th activities) with a rosette system and drained into acid-cleaned, pre-weighed cubitainers. The samples  243  were immediately weighed and acidified (~pH 2) with 2N HCl before adding 500mg of a 229 Th spike (1.5 dpm/g) in a solution of iron chloride (200mg, 50mg/g). After equilibration (24 h), concentrated NH4OH was added to adjust the pH to 8–9 and co-precipitate 230 Th and 229 Th with iron hydroxide. The precipitates were redissolved in the laboratory with HCl and Th was separated by ion exchange chemistry as described by Choi et al. (2001). Particulate 230 Th: Fine particles (1μm-53μm) collected by the LVPs on Supor filters (142mm, 1μm) were completely digested in concentrated HF, HNO3 and HClO4 acids (Seastar) after addition of 200mg of a 229 Th solution (1.5 dpm/g) as a yield tracer. After digestion and evaporation to a small drop, the samples were transferred to centrifuge tubes and topped-up to 10ml with 2N seastar HCl. Aliquots were taken to measure 232 Th, P, Ca and Al, and the remaining of the samples passed through ion exchange resins for Th separation and subsequent measurement of 230 Th (Choi et al., 2001). After completion of 234 Th counting on the LVP mesh (>53 μm) and multiNet (>236 μm) samples, they were completely digested in concentrated acid (Seastar HF, HNO3 and HClO4) after addition of 200mg of a 229 Th solution (1.5 dpm/g) as a yield monitor. After digestion and evaporation to a small drop, the samples were treated as described above for the fine particles.    244  6.2.1.4 232 Th, P, Al and Ca samples After the total digestion of LVP Supor (1-53μm), LVP mesh (>53μm) and MultiNet samples (>236μm), the final solution was transferred to centrifuge tubes and topped-up to 10ml with 2N Seastar HCl. Aliquots (~1ml) were taken from each sample to measure P, Al and Ca. Separate aliquots (1ml) for 232 Th were taken from the LVP Supor samples and spiked with 229 Th to correct 230 Th activities for lithogenic contributions in fine particles (A( 230 Thlitho)/A( 232 Th) = 0.8 dpm/dpm). The Al in LVP mesh samples (>53μm) and MultiNet samples (>236μm) were used to correct 230Th activities for lithogenic contributions in large and extra-large particles (A( 230 Thlitho)/Al = 462dpm/mol). 6.2.2 Sample analysis 6.2.2.1 234 Th measurements The 234 Th samples were counted on a gas-flow proportional five-position beta counter manufactured by RISØ National Laboratories (Roskilde, Denmark). With approximately 10cm of low-radiation lead surrounding the counter, the background count rates were consistently low for the 5 positions (~0.05–0.09 cpm, Figure a, Appendix E) over the entire period of measurements (June/2010-May/2011). The detector efficiency for the Mn oxide precipitates was determined by measuring the samples collected below 400m and assuming secular equilibrium with 238 U estimated from salinity (Table a, Appendix E). The sample at 1000m appears to have been  245  contaminated by a beta emitter and it was thus not used for the calibration. The average obtained from all the samples was used to calculate the efficiency of the detectors and this calibration factor (0.558 ± 0.004) was used to calculate the total 234 Th activity of all samples. Detector efficiencies for the particulate samples were measured using a standard 238 U/FeO filter following Rutgers van der Loeff and Moore (1999) and tested during the GEOTRACES inter-calibration exercise (Maiti et al., 2012). All the samples were counted three times (1 st  counting shortly after sampling, 2 nd  counting after one month and 3 rd  counting after 8 months) for periods long enough to reach counting errors below 3%, and decay corrected to estimate the sample activity at the time of sampling (Buesseler et al., 2001). After the third counting, the filters for total 234 Th were digested with 229 Th as a yield tracer and 229 Th and 230 Th (the latter added during the co-precipitation) were measured by ICP-MS to estimate the 234 Th recovery, following Pike et al. (2005). The recoveries were generally above 90% (Table b, Appendix E). 6.2.2.2 POC measurements Following the completion of 234 Th counting, at least 8 months after the initial sample collection, POC and PON on 25mm TQ filters (fine particles, >1μm) were measured by gas chromatography on a model 1106 Carlo Erba CHN analyzer with a precision of ±1.3% for carbon and ±2% for nitrogen (Verardo et al., 1990). The POC on fine particles (collected on LVP Supor, 1-53μm), large particles (collected on LVP Mesh, >53μm) and XL particles (collected by multinet, >236μm) were estimated from the phosphorus measurements in the aliquots (1ml) taken after the complete digestion of the Supor filters.  246  6.2.2.3 230 Th measurements After separation of Th by anion-exchange, the samples were evaporated to a small drop and fumed with 5ml of concentrated HClO4 for 2 hours to remove organic matter that leached from the anion-exchange resin. After fuming, the samples were taken up in 2N Seastar HNO3 for ICP-MS measurements of 230 Th and 229 Th. 6.2.2.4 232 Th, P, Al, and Ca measurement The aliquots (1ml, in 2N Seastar HCl) for P, Al and Ca, taken from the final solution (topped-up to 10ml with 2N seastar HCl) after the total digestion of LVP Supor (1-53μm), LVP mesh (>53μm) and multinet samples (>236μm), were evaporated and taken up in 1ml 2N Seastar HNO3. They were then diluted by a factor of 10 to 50 (LVP Supor samples above 300m were diluted 50 times and all other samples were diluted 10 times) with 2N Seastar HNO3 and measured by ICP-MS. Standards for P, Al and Ca were prepared by diluting the 1000ppm standards to 1ppt - 1ppm with 2N Seastar HNO3. The aliquots for 232 Th from the LVP Supor samples were spiked with 100mg 229 Th solution (3950 dpm/g), transferred to 2N Seastar HNO3 and then measured on an ICP-MS by isotope dilution.  247    248     249   6.3 Results The 234 Th and 230 Th activities of all seawater and particle samples are reported in Table 6.2. 6.3.1 234 Th The vertical profiles of total 234 Th and 238 U (Figure 6.2) clearly exhibit a 234 Th deficit above 80m produced by scavenging. There is also an excess 234 Th below 80m, which is attributed to particle re-mineralization and release of 234 Th to seawater. 234 Th: 238 U equilibrium is assumed below 350m (within the error of measurements), except at 1000m where the sample appears to have been contaminated by a beta emitter. This sample is thus excluded from further discussion. Particulate 234 Th activities measured by small volume filtration on 25mm TQ filters (>1μm) decrease sharply between 60 m and 80 m and more gradually below that depth (Fig. 6.2). There are also minima at ~100m and 200-300m, which coincide with excesses in total 234 Th. Fine particulate 234 Th activities collected by LVPs on Supor filters (1-53 μm) generally agree with niskin samples but are systematically slightly lower (Figure 6.3). This could, in part, be explained by the collection of larger particles (> 53 μm) during small volume filtration. However, large particles are rare in small volumes, and adding the 234 Th activity measured on the Supor and mesh samples at each depth does not quite  250  account for the values measured when filtering small volumes of seawater on TQ filters. The difference is small and within the error bars of the measurements, but the consistent offset (Fig. 6.3) suggests a small but real difference between the two sampling techniques.  Figure 6.2: Total 234 Th, particulate 234 Th (obtained by small volume filtration) and 238 U profiles (calculated from salinity) at OSP (June/2010). The 234 Th activities of large particles collected by LVPs on mesh (>53 μm) increase between 60m and 100m in contrast to the sharp decrease observed with fine particulate 234 Th in this depth interval (Fig. 6.3, 6.4). Below 100m, their 234 Th activities generally decrease with depth. The 234 Th activities of extra-large (XL) particles collected by multiNet (>236 μm) are two orders of magnitude lower than the activities measured on the large particles collected with the LVPs (Fig. 6.4). They increase slightly from the 0-200m interval to the 200-500 interval and then decrease with depth. 0 200 400 600 800 1000 1200 1400 1600 0.00 0.50 1.00 1.50 2.00 2.50 D ep th  ( m ) 234Th at OSP June/2010 (dpm/l) U-238 Th-234 Total Th-234 Particulate (25mm TQ)  251   Figure 6.3: Profiles of fine particulate 234 Th obtained by small volume filtration on 25mm TQ filtration (>1μm, green triangles), by LVPs on Supor filters (1-53 μm, blue diamonds), and large (>53 μm) particulate 234 Th collected with LVPs on Nylon mesh (brown dots). The sum (>1 μm) of fine particulate 234 Th from LVP Supor filters (1-53 μm) and large particulate 234 Th from LVP Mesh (>53 μm) are marked as black dots.  Figure 6.4: Profiles of 234 Th in large particles (>53μm, collected on LVP Mesh) and “extra-large” (XL) particles (>236μm, collected by MultiNet). 0 200 400 600 800 1000 1200 1400 1600 0 0.2 0.4 0.6 0.8 1 1.2 D ep th  ( m ) Particulate 234Th at OSP (dpm/l) Th-234 (25mm TQ) Th-234 (LVP Supor) Th-234 (LVP Mesh) Th-234 (Supor+Mesh) 0 200 400 600 800 1000 1200 1400 1600 0 0.02 0.04 0.06 0.08 0.1 D ep th  ( m ) Particulate 234Th at OSP (dpm/l) Th-234 (LVP Mesh) 0 200 400 600 800 1000 1200 1400 1600 0 0.0001 0.0002 0.0003 0.0004 0.0005 D ep th  ( m ) Particulate 234Th at OSP (dpm/l) Th-234 XL MultiNet  252    253  6.3.2 POC The concentrations of POC in fine particles were measured directly on the 25mm TQ filters (>1μm). Particulate organic carbon on the LVP Supor filters (1-53 μm), the LVP mesh samples (> 53 m) and the multiNet samples (> 236 m) were estimated from their P content, assuming a C/P ratio of 106. POC concentrations obtained from the 25mm TQ samples decrease with depth and drop rapidly from 60m to 100m (Table 6.3, Fig. 6.5), which coincides with a total 234 Th excess (Fig. 6.3), further indicating intensive remineralization below the euphotic zone. Particulate organic carbon concentrations estimated from the P content of the LVP Supor samples and a Redfield ratio of 106 are similar to those measured on the 25mm TQ filters below 600m but significantly higher at shallower depths, especially above 100m (Fig. 6.5). If the two filtration methods sample the same pool of particles, this discrepancy suggests that the C/P of organic matter at shallow depth at station Papa is significantly lower than the canonical Redfield ratio of 106, and that C/P increases with depth. If we assume that the particles collected on the 25mm TQ filters and the LVP Supor have the same POC/ 234 Th, we can also estimate POC concentration in the LVP-Supor samples from their 234 Th activities and estimate the C/P ratio of the particles (Fig. 6.6). These estimates show the increasing trend in C/P with depth, which is consistent with the preferential remineralization of P over C. However, the C/P estimated for shallow waters appears unrealistically low.  254    Figure 6.5: POC profiles and POC/ 234 Th for the particles collected on 25mm TQ filters, the fine particles collected by LVP on Supor filters (1-53 μm), and Nitex mesh (>53μm). POC for LVP samples were estimated from the P concentrations multiplied by the Redfield ratio (106). The POC concentrations estimated from the P content of the large particles (>53 μm) collected on Nylon mesh (Table 6.4), assuming C/P = 106, are lower than the POC associated with fine particles but show a similar decreasing trend with depth (Fig. 6.5, Table 6.3). The POC concentrations estimated from the P content of the extra-large (XL) 0 200 400 600 800 1000 1200 1400 1600 1800 0 20 40 60 80 100 D ep th  ( m ) POC (μg/l) 25mm TQ LVP Supor LVP Mesh Multinet 0 200 400 600 800 1000 1200 1400 1600 1800 0.00 2.00 4.00 6.00 8.00 D ep th  ( m ) POC (μg/l) 25mm TQ LVP Supor LVP Mesh Multinet 0 200 400 600 800 1000 1200 1400 1600 1800 0 20 40 60 80 100 120 D ep th  ( m ) POC/234Th (μg/dpm) 25mm TQ LVP Supor LVP Mesh  255  particles (>236 μm) collected with the MultiNet (Table 6.4), assuming C/P = 106, decrease with depth gradually from 100m to 1000m (Fig.6.5). They are similar to the POC concentration associated with large particles (mesh samples) below 1000m but significantly higher above 1000m.  Figure 6.6: Redfield ratio derived from the 25mm TQ and LVP Supor data by assuming the POC/ 234 Th are the same on those two different types of fine particles 6.3.3 POC/ 234 Th The POC/ 234 Th ratios measured directly on the 25mm TQ filters decrease from a maximum of ~ 50 g C/dpm in surface water and quickly drop below 20 g C/dpm in the upper 100 m of the water column (Fig. 6.5, Table 6.3). These results are similar to those obtained by Charette et al. (1999). 0 200 400 600 800 1000 1200 1400 1600 1800 0 50 100 150 D ep th  ( m ) C/P (mol/mol)  256  To estimate the POC/ 234 Th ratios of the LVP-Supor samples, we can either assume that they are similar to the ratios measured directly on the 25mm TQ filters (this is the assumption used to estimate C/P in Fig. 6.6) or we can assume that C/P = 106 at all depth, in which case, we find a good agreement with the TQ filter measurements below 600 m but significantly higher values in shallower water (Fig. 6.5). The POC/ 234 Th ratios of the large particles (>53μm, LVP Mesh) estimated from their P content and C/P = 106 are similar to the values measured on the TQ samples in the upper 400 m and only slightly lower in deeper waters (Fig. 6.5). On the other hand, they are lower than the POC/ 234 Th on the LVP Supor samples estimated from P concentration and C/P = 106. The POC/ 234 Th ratios of the XL particles (>236μm, Multinet) estimated from their P content and C/P = 106 are hundreds of times higher than ratios of fine and large particles (Table 6.3). 6.3.4 230 Th Both dissolved and fine particulate (1-53μm) 230Th activities increase almost linearly with depth (Fig. 6.7). The fraction of total 230 Th associated with fine particles increases linearly from about 10% at the surface to over 20% at 1600m (Fig. 6.8), with a higher ratio at 60 m. The large particulate (>53μm) 230Th activities, however, don't follow this linear increase trend but display a maximum at 400m (Fig. 6.9). The XL particulate (>236 μm) 230 Th activities are three orders of magnitude lower and below detection limit for the  257  samples taken from the 0-200m and 200-500m depth intervals. Below 500 m, they become measureable and increase with depth (Fig. 6.9).  Figure 6.7: Profiles of dissolved 230 Th, 230 Th on fine particles (1-53 μm, collected on LVP Supor) and large particles (>53μm, collected on LVP mesh).  Figure 6.8: The fraction of total 230 Th associated with fine particles. 0 200 400 600 800 1000 1200 1400 1600 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 D ep th  ( m ) 230Th at OSP June/2010 (dpm/T) Th-230 Dissolved Th-230 Fine particulate (LVP Supor) Th-230 Large particulate (LVP Mesh) 0 200 400 600 800 1000 1200 1400 1600 1800 0.0% 20.0% 40.0% D ep th  ( m ) Particulate 230Th/Total 230Th  258   Figure 6.9: Profiles of 230 Th on fine (1-53 μm collected on LVP Supor filters), large (>53 μm, collected on LVP mesh) and XL particles (>236 μm collected on MultiNet).  Figure 6.10: POC/ 230 Th for particles collected by LVP on Supor filters (1-53 μm), and Nitex mesh (>53μm). POC for LVP samples were estimated from their P concentration and the Redfield ratio (106). 0 200 400 600 800 1000 1200 1400 1600 0 0.05 0.1 D ep th  ( m ) Particulate 230Th at OSP June/2010 (dpm/T) Fine (LVP Supor) Large (LVP Mesh) 0 200 400 600 800 1000 1200 1400 1600 0 0.00002 0.00004 D ep th  ( m ) Particulate 230Th at OSP June/2010 (dpm/T) XL (multinet) 0 200 400 600 800 1000 1200 1400 1600 0 500 1000 D ep th  ( m ) POC/230Th (mmol/dpm) LVP Supor LVP Mesh 0 200 400 600 800 1000 1200 1400 1600 0 10 20 30 40 D ep th  ( m ) POC/230Th (mmol/dpm) LVP Supor LVP Mesh  259    260  6.3.5 POC/ 230 Th The POC/ 230 Th ratios on the fine (LVP-Supor), large (LVP-Mesh) and extra-large (MultiNet) samples (Fig. 6.10) were calculated by assuming C/P = 106 at all depths (Table 6.5). The POC/ 230 Th ratios in fine particles decrease with depth from 1300mmol/dpm at 20 m to 1.8 mmol/dpm at 1600 m (Fig. 6.10), consistent with increasing POC degradation and 230 Th scavenging with depth. The POC/ 230 Th ratios on large particles increase sharply between 60 m and 100 m and then generally decrease with depth. They are substantially lower than the ratios measured in fine particles at shallower depth and then converge in deeper waters. This is in contrast to POC/ 234 Th ratios, which are more similar for the two particle sizes (Fig. 6.5). One way to explain this trend in POC/ 230 Th is by invoking formation of the large particles in the upper 1000 m by aggregation of fine particles and consumption of a larger fraction of their organic matter by heterotrophs before sinking, compared to the fine particles at the same depth. The size of the larger particles may be why they tend to be preferentially consumed. The similarity in the trend of POC/ 234 Th for the two particle pools could be explained by decomposition/consumption rates of POC in these large particles of the same order as the decay rate of 234 Th. On the other hand, the POC/ 230 Th ratios on the extra-large particles are several hundred times higher than POC/ 230 Th ratios on fine particles (Table 6.5). They also decrease sharply with depth, from ~1300 nmol/dpm at 600 m to ten times lower at 1400 m. Their much higher POC/ 230 Th ratios  261  compared to the smaller particles indicate that these extra-large particles are not formed by aggregation of smaller ones from the same depth horizon, even in the upper 200 m. The total amount of POC collected with the MultiNet between 0 – 200 m was ~ 10 mmoles (1.772 mgC/m 3  x 61 m 3 ). If these extra-large particles were formed by aggregation of small particles with a POC/ 230 Th ratio of ~ 200 nmoles/dpm (Table 6.5), their 230 Th activity should have been easily measureable, and not below detection limits, as we found. Instead, the extra-large particles we collected with the MultiNet must have been produced directly by marine organisms and their low POC/ 230 Th must reflect their low surface area to volume ratio (Buesseler et al., 2006). This conclusion is however in part based on the assumption that the slow rate of ascent of the MultiNet allowed all zooplankton to escape and the material recovered consisted only of passively sinking large particles. The technique was used without pre-planning and prompted by serendipity. Although a cursory examination of the samples seems to indicate that living zooplankton were largely absent, a much more systematic study of the technique is still required to fully evaluate this relatively simple sampling approach and the validity of this conclusion. 6.3.6 P and Ca The vertical profiles of fine (1- 53 m) particulate P and Ca are very similar. Both decrease sharply in the upper 100 m of the water column (Fig. 6.11, Table 6.4). While particulate P concentration decreases sharply between 20 m and 60 m, however, the largest decrease in particulate Ca occurs between 60 m and 100 m. Likewise, the  262  concentration of both elements decreases with depth in the large particles (> 53 m), particularly down to 400 m depth (Fig. 6.11). In contrast, the concentration profiles of the two elements in the XL particles (> 236 m) collected with the multiNet are very different. The P concentrations associated with XL particles are higher than the P concentrations associated with large particles, and decrease gradually down to 1000 m depth. On the other hand, the Ca concentrations associated with XL particles is much lower than the concentrations associated with the large particles (Fig. 6.11). 6.3.7 Al The Al concentrations associated with fine particles show a systematic trend with higher concentration towards the surface, a minimum at 300 m, and a slight increase toward 1600m. The Al concentrations associated with the XL particles are ten times lower (Fig. 6.11), and display a similar trend as for the fine particles, but with a deeper minimum at ~ 1000 m. In contrast, the Al concentrations in the large particles collected by LVP do not show a clear trend with depth, and they are generally slightly higher than for the XL particles (Fig. 6.11).  263     Figure 6.11: P, Ca, and Al concentrations associated with fine, large and XL particles. 0 200 400 600 800 1000 1200 1400 1600 1800 0 0.02 0.04 0.06 0.08 0.1 D ep th  ( m ) Fine particulate P (mmol/m3) LVP Supor 0 200 400 600 800 1000 1200 1400 1600 1800 0 0.0005 0.001 0.0015 D ep th  ( m ) L and XL particulate P (mmol/m3) LVP Mesh Multinet 0 200 400 600 800 1000 1200 1400 1600 1800 0 0.5 1 1.5 D ep th  ( m ) Fine particulate Ca (mmol/m3) LVP Supor 0 200 400 600 800 1000 1200 1400 1600 1800 0 0.02 0.04 0.06 0.08 0.1 D ep th  ( m ) L and XL particulate Ca (mmol/m3) LVP Mesh Multinet 0 200 400 600 800 1000 1200 1400 1600 1800 0 0.002 0.004 0.006 0.008 D ep th  ( m ) Fine particulate Al (mmol/m3) LVP Supor 0 200 400 600 800 1000 1200 1400 1600 1800 0 0.0001 0.0002 0.0003 0.0004 0.0005 0.0006 D ep th  ( m ) L and XL particulate Al (mmol/m3) LVP Mesh Multinet  264  6.4 Discussion 6.4.1 Th-234 fluxes Integrating the 234 Th deficit over the depth where it is measureable gives the removal rate of 234 Th from the upper water column, expressed in dpm/m 2 .d (see Chapter 5): Flux Th z=λ234∫ (A238 − A234 𝑥 0 ) dz    (6.1) where z is the depth of integration. Applying equation 6.1 to the 234 Th activities measured on the TQ samples at station P (Fig. 6.2) provides the flux of 234 Th as a function of depth (Fig. 6.12, Table 6.6). The 234 Th flux increases systematically in the upper 80m, where a deficit is found, then decreases gradually to a depth of 350 m where particle remineralization releases 234 Th back to the water column, and stays seemingly constant below. The latter observation stems from the fact that the beta counter was calibrated based on the assumption that the samples below 400m are in secular equilibrium. However, there is also a limit on how deep in the water column this calculation can be applied meaningfully. Given the precision of our 234 Th measurements, we cannot distinguish secular equilibrium from 234 Th excesses or deficits of ± 0.05 dpm/L. Therefore, below the depth of measurable deficit or excess (in the presence case, below 300 m), the uncertainty added to the flux of 234 Th to the seafloor is 50 dpm/m 3  x 0.0288 y -1  for every vertical meter, i.e. an added uncertainty of 144 dpm/m 2 .d per 100m. Since the 234 Th flux at 300 m is 500 dpm/m 2 .d ± 50%, the uncertainty on the flux scavenged from or released to  265  deeper waters very quickly overwhelms the 234 Th flux originating from the upper water column, as illustrated by the increasing error bars with depth reported on Fig. 6.12. Equation 6.1 also assumes that the surface water at station OSP is a closed system at steady-state. Since there is no distinct spring bloom at OSP (Harrison, 2002), this assumption should be reasonably valid but cannot be verified due to lack of 234 Th data from the previous month.  Figure 6.12: Th-234 fluxes based on 234 Th: 238 U deficit 0 200 400 600 800 1000 1200 1400 1600 1800 0 500 1000 D ep th  ( m ) Th-234 fluxes (dpm/m2/day)  266    267   6.4.2 Fluxes of POC, Ca and Al estimated from the 234 Th: 238 U deficit The fluxes of any particle constituent X can be estimated by multiplying the 234 Th fluxes obtained from equation 6.1 by the X/ 234 Th ratio of the particles that sink from the surface and scavenge 234 Th: Flux X z= Flux Th z *(C X z/C Th z)     (6.2) Where Flux X z and Flux Th z are the sinking fluxes of X and 234 Th at depth z, C X z and C Th z are the concentrations of X and the activities of 234 Th in the sinking particles at depth z. 6.4.2.1 POC fluxes Applying equation 6.2 to POC measured on the 25mm TQ samples (Fig. 6.5) provides the flux of POC as a function of depth (Fig. 6.13, Table 6.6). POC fluxes at station Papa gradually increase from the surface to 60 m and then decrease with depth due to organic matter remineralization, a pattern that has been observed before (e.g., Bacon et al., 1996). The flux estimates from this study are similar to those obtained at OSP in May 1996 by Charette et al. (1999).  268   Figure 6.13: POC fluxes derived from the 234 Th flux and POC/ 234 Th ratio on the 25mm TQ filters. The triangles show the POC fluxes from sediment traps reported by Timothy et al., (submitted) at the same location. The smaller black, yellow, purple and pink triangles stand for the winter (20/November-19/February), spring (20/February-20/May), summer (21/May-19/August) and fall (20/Augest-19/November) trap fluxes, respectively. The flux for each season is averaged over the corresponding sampling period. The annual flux (larger blue triangle) is the average of the flux over their entire sampling period (from May/1989 to June/2006 for the trap at 200m and from March/1983 to June/2006 for the trap at 1000m). POC fluxes were also measured at OSP with sediment traps deployed at 200m and 1000m (Timothy et al. submitted) between March/1986 and Jun/2006. The 234 Th-based POC fluxes agree reasonably well with these sediment trap results (Fig. 6.13). At 200 m, the 0 200 400 600 800 1000 1200 1400 1600 1800 0 10 20 30 40 D ep th  ( m ) POC fluxes (mg/m2/day) 25mm TQ Trap Winter Trap Spring Trap Summer Trap Fall Trap Annual  269  234 Th-derived flux is somewhat lower than the sediment trap flux collected in spring (20/Feb –20/May, Timothy et al., submitted). The POC flux estimated by the 234Th method at 1000 m is very close to the flux measured by the sediment trap in spring. This agreement is however fortuitous, considering the large uncertainties in the fluxes derived from 234 Th at this depth (Fig. 6.12). Nonetheless, the 234 Th method provides a maximum POC flux at 1000 m consistent with the sediment trap data. The POC fluxes estimated from the LVP-Supor 234 Th activities and P concentrations with a constant Redfield ratio of 106 are larger in the upper 100m than those obtained from direct POC and 234 Th measurements in 25mm TQ filters but converge at 200m (Fig. 6.14). This difference arises from uncertainties regarding the C/P ratio of particulate organic matter. Because the POC/ 234 Th ratios of the large particles collected by LVP on Nylon mesh (>53μm) are very similar to those measured on the TQ samples (Fig. 6.5), the large particles necessarily provide flux estimates that are also very similar to the estimation based on 25mm TQ (>1μm) (Fig. 6.14).  270   Figure 6.14: POC fluxes derived from the POC/ 234 Th ratio on all three samples. 6.4.2.2 Ca fluxes The flux of biogenic Ca is calculated based on equation 6.2 and the Cabio/ 234 Th ratio on both fine (LVP Supor) and large (LVP Mesh) samples (Tables 6.7, 6.8). Cabio was calculated by estimating lithogenic Ca from the Al content (Ca/Al = 0.3mol/mol) and subtracting it from the total particulate Ca. Biogenic Ca flux estimated from the fine particles shows a maximum at 60m, as for the POC flux. It quickly decreases from 60 m to 300m and stays almost constant (within large error bars) below 300m (Fig. 6.15). 0 50 100 150 200 250 300 350 400 0 10 20 30 40 50 60 70 D ep th  ( m ) POC fluxes (mg/m2/day) 25mm TQ LVP Supor LVP Mesh  271   Figure 6.15: Biogenic Ca fluxes derived from the biogenic Ca/ 234 Th ratio on LVP-Supor and mesh samples. Triangles show the fluxes measured with sediment trap (Timothy et al., submitted). The sharp decrease in biogenic Ca flux between 60 m and 300 m can be attributed to biologically-mediated carbonate dissolution resulting from CO2 released by organic matter remineralization in zooplankton guts or particles aggregates (e.g. Milliman et al., 1999). Biogenic Ca flux obtained from the large particles (LVP Mesh) shows a similar trend but with somewhat higher fluxes above 400m (Fig. 6.14). The fluxes of biogenic Ca derived from 234 Th are similar to fluxes measured with sediment traps (Timothy et al., submitted). The latter also indicate a nearly constant carbonate flux between 200m and 1000m. 6.4.2.3 Al fluxes The flux of Al is calculated based on equation 6.2 and the Al/ 234 Th ratio on both fine (LVP Supor) and large (LVP Mesh) samples (Tables 6.7, 6.8). The Al fluxes estimated from both 0 200 400 600 800 1000 1200 1400 1600 1800 0 0.5 1 1.5 2 2.5 3 D ep th  ( m ) Biogenic Ca flux (mmol/m2/d) LVP Supor LVP Mesh Trap Spring Trap Summer  272  fine and large particles are very low (Fig. 6.16) and consistent with aeolian dust flux estimated for this region (Mahowald et al., 1999; 0.1-0.2 g dust / m 2 .y; ~ 10% Al in dust; aeolian Al flux = ~ 0.001 – 0.002 mmol/m2.d). The Al fluxes associated with fine particles increase between 20 m and 100 m suggesting lateral transport of fine particles from the shelf (Lam and Bishop, 2008). Below 200 m, Al fluxes decrease and appear to remain constant within the large error bars of the method at these depths. This would suggest removal of fine Al-containing particles by aggregation below 100m. The Al fluxes obtained from the large particles are similar and uniformly low. If the decrease in Al fluxes associated with small particles below 100 m is due to aggregation, the aggregates involved must have been very large and not effectively sampled by the LVP.  The very low Al fluxes estimated here are also consistent with the sediment trap study at station Papa, which report “negligible” lithogenic fluxes (Timothy et al., submitted).  Figure 6.16: Al fluxes derived from the Al/ 234 Th ratio on LVP samples. 0 200 400 600 800 1000 1200 1400 1600 1800 0 0.002 0.004 0.006 0.008 0.01 0.012 D ep th  ( m ) Al flux (mmol/m2/d) LVP Supor LVP Mesh  273      274    275   276  6.4.3 230 Thxs fluxes The 234 Th: 238 U method provides an effective means for estimating POC or other elemental sinking fluxes in the water column, as long as the scavenging intensity is large enough to generate a 234 Th deficit that can be measured precisely. As discussed above, because 234 Th has a very short half-life, a measureable deficit is often restricted to the upper water column. Thorium-234 is thus a powerful tool to estimate export flux from surface water (Buesseler et al., 2006) but has very limited applicability to deeper water (with the exception of the bottom nepheloid layer, e.g. Rutgers van der Loeff and Bacon, 1989). With its much longer half-life (~76.6 ky) and short residence time in the water column (ca. 30 years), 230 Th decay in seawater is very limited and most of the 230 Th produced from the decay of 234 U must be removed by particle scavenging. As a result, there is always a very large deficit between the activity of total 230 Th in seawater and that of its parent 234 U. In fact, the deficit is so large that, to a very good approximation, the decay of 230 Th can be neglected and therefore the flux of 230 Th scavenged from the water column (Flux Thxs z) is very nearly equal to the rate of production in the water column: Flux Thxs z=λ230∫ (A234 − 0 𝑥 0 ) dz       (6.3) where A234 is the activity of 234 U in seawater and z is the depth of integration. Since the activity of 234 U is essentially constant with depth (2910 dpm/m 3 ), the flux of 230 Thxs is thus simply proportional to water depth: Flux Thxs z=λ230 A234 z = 0.0267 z dpm/m 2 .y    (6.4)  277  Some deviations from this simple equation could arise as a result of lateral transport of dissolved 230 Th by advection or turbulent mixing (e.g., Francois, 2007). However, observations and modeling studies indicate that equation 6.4 is accurate to within 30% over most of the ocean (Henderson et al., 1999; Yu et al., 2001, Luo et al., 2010). This equation has been widely used in sediments to estimate past changes in the rain rate of particles reaching the seafloor (e.g., Francois, 2004) but never applied systematically to water column particles. 6.4.4 Fluxes of POC, Ca and Al estimated by normalization to 230 Thxs Because the 230 Th deficit is always very large and easy to quantify, the principles applied to estimate particle flux in the upper water column from the 234 Th deficit could be similarly applied to 230 Th over the entire water column. Flux X z= Flux xsTh-230 z *(C X z/C xsTh-230 z)      (6.5) where Flux X z and Flux xsTh-230 z are the sinking fluxes of X (X=P, Ca, Al, etc.) and scavenged flux of 230 Th at depth z, C X z and C xsTh-230 z are the concentrations of X and the activities of “excess” 230Th in the sinking particles at depth z. “Excess” 230Th is the total activity of 230 Th in marine particles corrected for 230 Th in secular equilibrium with 234 U in the lithogenic fraction of the particles (Table 6.9). The main difficulty in applying this method is to measure the very low 230 Th concentrations associated with particles (C xsTh-230 z) particularly at shallow depths, which  278  requires the use of large volume in-situ pumps and, as for 234 Th, establishing the appropriate (C X z/C xsTh-230 z) to estimate the flux of X. 6.4.4.1 POC fluxes As the first step, the POC flux was calculated by multiplying the 230 Th fluxes by the POC/ 230 Thxs ratio estimated from the P content and the Redfield ratio (106), and the 230 Th activities measured on the fine particles collected by LVP on Supor filters (Fig. 6.17): Flux POC z= Flux Thxs z *(C P z/C Thxs z)*106     (6.6)  Figure 6.17: POC fluxes derived by 230 Th normalization on fine particles compared with the POC fluxes estimated by other means as in Figure 6.12. 0 200 400 600 800 1000 1200 1400 1600 1800 0 5 10 15 20 25 30 35 D ep th  ( m ) POC fluxes (mg/m2/day) Th-230 norm LVP Supor Th-230 norm LVP Mesh Th-234 25mm TQ Trap Winter Trap Spring Trap Summer Trap Fall Trap Annual  279  The POC fluxes derived by this approach are similar to the POC fluxes estimated from 234 Th (Fig. 6.5). The 230 Th-derived POC fluxes also display a maximum at 60m and a rapid decrease to 200m. The fluxes derived from this method also indicate a gradually decreasing POC flux from 300m to 1600m and at 200 m and 1000 m they are in reasonable agreement with the annually-average POC sediment trap flux measured at 1000 m. At 200 m depth, the residence time of 230 Th in seawater is approximately 2 years (0.05 dpm/m 3  / 0.0267 dpm/m 3 .y -1 ). The somewhat higher annually averaged fluxes measured by the sediment trap at 200 m may thus reflect inter-annual variability. At 1000 m, the residence time of 230 Th increases to 0.25 dpm/m 3  / 0.0267 dpm/m 3 .y -1  = ~ 10 years, consistent with the better agreement between trap flux (averaged between 1982 and 2006; Timothy et al., submitted) and 230 Th-normalization at this depth (Fig. 6.17). These results suggest that 230 Th normalization on fine particles may provide an effective means to estimate the POC fluxes in the deep water column where the 234 Th method cannot be applied. If the fluxes thus calculated are accurate, the extent of POC remineralization between sampling depths could also be readily calculated by difference, providing crucial data on the efficacy of the biological pump. However, particles with larger size are usually considered to be the major contributor of the POC flux to the deep sea because they have much faster sinking velocity compared to the fine particles. With our data, we can also test 230 Th normalization on the large (LVP Mesh, >53μm) and XL particles (MultiNet, >236μm) to estimate POC flux. The POC fluxes obtained from the large particles, however, are more than 5 times lower than that from the fine particles and are  280  evidently lower than the 234 Th-derived estimates or the sediment trap records (Table 6.9, Fig. 6.13). This is because their POC/ 230 Th ratios are lower than the POC/ 230 Th of fine particles (Fig. 6.10). On the other hand, the POC fluxes obtained from the XL particles are more than 200 times higher and much greater than the 234 Th estimation or the sediment trap records (Table 6.9). Clearly, 230 Th-normalization cannot be applied to the extra-large particles. In fact, the POC/ 230 Th ratio that must be applied to correctly estimate the flux of POC by 230 Th-normalization (and also the export fluxes estimated from the 234 Th deficit) is the flux-weighted average of all particle sizes that contribute to the sinking flux. Such an average could be obtained from the analysis of material collected by sediment traps that would work perfectly, i.e., traps that would intercept without biases the settling flux in all class sizes. The present observation that 230 Th-normalization on fine particles seems to work (or at least agree with sediment traps and the 234 Th method) is thus likely fortuitous and reflects the fact that the POC/ 230 Th of these fine particles is similar to the POC/ 230 Th flux-weighted average for the large (low POC/ 230 Th) and extra-large (high POC/ 230 Th) particles intercepted by the traps. 6.4.4.2 Ca fluxes The biogenic Ca fluxes were also calculated by multiplying the 230 Th fluxes by the biogenic Cabio/ 230 Thxs ratio on both fine (LVP Supor) and large (LVP Mesh) particles: Flux Cabio z= Flux Thxs z *(C Cabio z/C Thxs z)     (6.7)  281  The biogenic Ca fluxes calculated from the fine (LVP Supor) particles increase from 20m to reach the maximum at 60 to 100 m. They decrease quickly from 100m to 200m and then much more slowly from 200m to 1600m (Fig. 6.18). Taken at face value, these fluxes would indicate about 50% carbonate dissolution between 100 m and 200 m depth, consistent with other lines of evidence (Milliman et al., 1999, Feely et al., 2002, Antia et al., 2008). The 230 Th-normalized biogenic Ca fluxes are also reasonably consistent with biogenic Ca fluxes derived from 234 Th: 238 U (Table 6.8; although the flux estimated at 60 m is lower, possibly reflecting the difference in the integration time of the two proxies) and sediment traps (Timothy et al., submitted). The fluxes of biogenic Ca calculated from the large (LVP Mesh) particles are smaller. On the other hand, the Ca fluxes obtained by 230 Th-normalization on the extra-large particles are much larger, as was the case for POC fluxes. They are also very variable because of the very low carbonate concentration in these particles (Fig. 6.11). Again, at first sight, these results suggest that 230 Th normalization on fine particles may provide an effective means of estimating the biogenic Ca fluxes in the water column, and carbonate dissolution in the mesopelagic zone but, as for POC, this apparent success is likely fortuitous.  282   Figure 6.18: Biogenic Ca fluxes derived by 230 Thxs normalization compared to those derived by 234 Th: 238 U and sediment traps. 6.4.4.3 Al fluxes The Al fluxes calculated with equation 6.5 and the Al/ 230 Thxs ratio on fine particles (LVP Supor) are very similar to those estimated with 234 Th (Fig. 6.19) and also record a prominent maximum at 100m and a nearly constant flux below. These deeper fluxes can now be estimated with much smaller uncertainties than from 234 Th and indicate the Al fluxes are essentially constant with depth to 1,600m. Fluxes obtained from the large particles (LVP Mesh) are generally lower. 0 200 400 600 800 1000 1200 1400 1600 1800 0 0.2 0.4 0.6 0.8 1 1.2 1.4 D ep th  ( m ) Biogenic Ca flux (mmol/m2/d) 230Th (LVP Supor) 230Th (LVP Mesh) 234Th (LVP Supor) 234Th (LVP Mesh) Trap Spring Trap Summer Trap Annual  283   Figure 6.19: Al fluxes derived by 230 Thxs normalization compared to those derived by 234 Th: 238 U. 6.4.5 Particle dynamics The above observations suggest that 230 Thxs normalization on fine suspended particles could provide accurate flux estimates down the water column over the entire depth range of the mesopelagic zone and deeper. The fluxes derived from the present data set are consistent with the fluxes of POC, biogenic Ca and Al derived from 234 Th: 238 U disequilibrium in the upper water column and the fluxes measured with sediment traps at 200m and 1000 m. This is, however, surprising since small suspended particles are generally not viewed as the particles that are exported to deeper waters. To better evaluate the validity of this apparent success and establish whether this approach could be generally applicable, it is necessary to gain a better understanding of the influence of particle dynamics on flux estimates derived from 230 Th-normalization. 0 200 400 600 800 1000 1200 1400 1600 1800 0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 D ep th  ( m ) Al flux (mmol/m2/d) 230Th (LVP Supor) 230Th (LVP Mesh) 234Th (LVP Supor) 234Th (LVP Mesh)  284  The strategy followed to test 230 Th-normalization is to develop a particle dynamics model that would approximate the profiles of 234 Th, 230 Th and POC measured at OSP.  The goal is not to obtain a perfect fit and estimate the optimal particle dynamic parameters (adsorption and desorption rate constants; aggregation and disaggregation rate constants, sinking rates) that best describe particle dynamics at station Papa, but to develop a reasonable model which could then be used to compare the “true” sinking flux of POC (i.e., the sum of [POC] multiplied by sinking rates for each class of particles) and 230 Th-normalization on fine particles. Each parameter of the model can then be modified in sensitivity tests to better understand how they affect 230 Th-normalization and under what circumstances it could provide a good estimate of the “true” flux. 6.4.5.1 Model A As a first step, the parameters of the particle dynamic model (model A) depicted on Fig. 6.1 (K1, K-1, B1, B-1, S) were varied in an attempt to reproduce the profiles of 234 Th and 230 Th dissolved, and associated with fine and large particles measured at OSP. The following equations were used: P + K-1 F = (K1 + ) D   (6.8) K1 D + B-1 L = (K-1 + B1 +) F  (6.9) B1 F = (B-1 +) L + S L/z   (6.10)  285  where P = production rate of Th in seawater,  is its decay constant (negligible for 230Th); D = activity of dissolved Th; F = activity of thorium adsorbed on fine suspended particles; L = activity of thorium adsorbed on large sinking particles; λ = Th decay constant; K1 and K-1 are the adsorption and desorption rate constants between dissolved and fine particles; B1 and B-1 are the aggregation and disaggregation rate constants between fine and large particles; S = sinking rate of large particles. We started with K1 = 0.5/y and K-1 = 1.6/y (Bacon and Anderson, 1982), B1 = 3/y and B-1 = 150/y (Clegg et al., 1991), and S = 150m/day (Berelson et al., 2002). With these parametric values kept constant with depth, model A generates profiles of dissolved and fine particulate Th that are different from the observations but in the range of measured values. On the other hand, the model grossly underestimates the concentration of 234 Th and 230 Th associated with large particles (>53μm) (Fig. 6.20). The model is also unable to reproduce the low dissolved 234 Th and high fine particulate 234 Th measured in surface waters, where K1 must be >>0.5 y -1 , reflecting higher scavenging intensity due to higher particle concentrations and fluxes. Because we are mostly interested in measuring flux in the mesopelagic zone, only the data obtained between 60 m and 1600 m are considered in the following discussion and the sensitivity tests reported in Appendix E.  286   Figure 6.20: Output from model A with K1 = 0.5/y and K-1 = 1.6/y (Bacon and Anderson, 1982); B1 = 3/y and B-1 = 150/y (Clegg et al., 1991); and S = 150m/day (Berelson et al., 2002). Sensitivity test are reported in Appendix E. 0 0.1 0.2 0.3 0.4 0 500 1000 1500 2000 (dpm/m3) D e p th (m ) D230 0 0.05 0.1 0 500 1000 1500 2000 (dpm/m3) D e p th (m ) F230 0 0.01 0.02 0.03 0 500 1000 1500 2000 (dpm/m3) D e p th (m ) L230 0 500 1000 1500 2000 2500 0 500 1000 1500 2000 (dpm/m3) D e p th (m ) D234 0 200 400 600 800 0 500 1000 1500 2000 (dpm/m3) D e p th (m ) F234 0 50 100 0 500 1000 1500 2000 (dpm/m3) D e p th (m ) L234  287  In an attempt to improve the fit between observations and model A, each of the 5 parameters was systematically varied to assess its impact on the Th profiles. The results of these tests are shown in Appendix E. Taking Fig. 6.20 as the starting point, increasing the adsorption rate constant K1 decreases dissolved and increases the fine and large particulate 234 Th concentrations. It also decreases the dissolved 230 Th activity but does not affect the 230 Th associated with particles. While this improves the fit for 234 Th, it worsens the fit for 230 Th (Fig. c; Appendix E). Increasing the desorption rate constant K-1 has an opposite but smaller effect on dissolved and particulate 234 Th (Fig. d). It increases dissolved 230 Th activity but does not change particulate 230 Th. Changing K1 and K-1 has no impact on the activities of 230 Th associated with large particles, which always remain much too low (Figs c, d; Appendix E). Increasing the aggregation rate constant B1 decreases dissolved and fine particulate 230 Th activities but lowers much less the activities of 234 Th in these two pools. While increasing B1 increases the 234 Th activities associated with large particles, it fails to increase 230 Th activities significantly (Fig. e; appendix E). Increasing the disaggregation rate constant has the opposite effect. Decreasing the sinking rate of particles (S) can substantially increase the 230 Th activities of large sinking particles but fails to increase their 234 Th activities significantly. It also produces exceedingly high dissolved and fine particulate 230 Th (Fig. f; appendix E). Clearly, the distribution of 230 Th and 234 Th activities measured at station Papa cannot be explained by the model A described in Fig. 6.1. In particular, it is incompatible with the relatively large activities measured in the mesh samples.  288  The relatively high 230 Th activities associated with the large particles suggest that the mesh samples collected with large volume pumps consist mostly of relatively small particles that sink relatively slowly and this sampling technique fails to effectively collect the rarer and much larger particles that sink more rapidly. The next step is thus to use the data obtained with the multiNet and develop a model with three classes of particles: the small particles collected by LVP on Supor filters (1-53 m), the large particles collected by LVP on Nylon mesh (> 53 m) from a few 100 L of seawater, and the “extra-large” particles collected with the multiNet (> 236 m) from much larger volumes of seawater, and use a more complex model which considers exchanges between these three classes of particles (Fig 6.21). 6.4.5.2 Model B  Figure 6.21: In this model we consider 3 different classes of particle: fine particles (1 – 53 m) collected on Supor filters by large volume pumps; large particles (> 53 m) collected on Nylon mesh by large volume pumps; “extra-large” particles (>236 m) collected by MultiNet. P = Th production in the water column from U decay; D = activity of dissolved Th; F = activity of thorium adsorbed on fine suspended particles; L = activity of thorium adsorbed on large sinking particles; XL = activity of thorium associated with extra-large particles; λ = Th decay constant; K1 and K-1 = adsorption and  289  desorption rate constants; B1 and B-1 = aggregation and disaggregation rate constants between fine and large particles; C1 and C-1 = the aggregation and disaggregation rate constants between large and extra-large particles; SL = sinking rate of large particles; SXL = sinking rate of extra-large particles. The parameters of the particle dynamic model (model B) depicted on Fig. 6.21 (K1, K-1, B1, B-1, C1, C-1, SL, SxL) were varied to try reproducing the profiles of 234 Th and 230 Th measured at OSP. The following equations were used: P + K-1 F = (K1 + ) D    (6.11) K1 D + B-1 L = (K-1 + B1 +) F   (6.12) B1 F+C-1 XL = (B-1 + C1 +) L + SL L/z  (6.13) C1 L = (C-1 +) XL + SXL XL/z   (6.14) where P is the production rate of Th in seawater,  is its decay constant (negligible for 230 Th) and z is water depth. Sensitivity tests were conducted (Appendix E) to try optimizing the parameters of model B to generate Th profiles similar to observations. A reasonable fit (Fig. 6.22) was obtained with the following parameters: K1 = K1 50 *(Z/50) ^-0.4, (K1 50 =5/y) and K-1 = 5/y; B1 = B1 50 *(Z/50)^-0.4, (B1 50 =40) and B-1 = 150/y (Clegg et al., 1991); C1 = 3/y and C-1 = 150/y; SL = 10m/day and SxL =SxL 50 *(Z/50), SxL 50 =50m/d).  290   Figure 6.22: The optimal run of model B with K1 = K1 50 *(Z/50) ^-0.4, (K1 50 =5/y) and K-1 = 5/y; B1 = B1 50 *(Z/50)^-0.4, (B1 50 =40) and B-1 = 150/y (Clegg et al., 1991); C1 = 3/y and C-1 = 150/y; S = 10m/day and SL =SL 50 *(Z/50), SL 50 =50m/d). 0 0.2 0.4 0.6 0.8 0 1000 2000 (dpm/m3) D e p th (m ) D230 0 0.05 0.1 0.15 0.2 0 1000 2000 (dpm/m3) D e p th (m ) F230 0 0.01 0.02 0.03 0 1000 2000 (dpm/m3) D e p th (m ) L230 0 2 4 6 x 10 -5 0 1000 2000 (dpm/m3) D e p th (m ) XL230 0 500 1000 1500 2000 2500 0 1000 2000 (dpm/m3) D e p th (m ) D234 0 200 400 600 800 0 1000 2000 (dpm/m3) D e p th (m ) F234 0 50 100 0 1000 2000 (dpm/m3) D e p th (m ) L234 0 0.1 0.2 0.3 0.4 0 1000 2000 (dpm/m3) D e p th (m ) XL234  291  This model was then used as a starting point to evaluate the remineralization rates (Fig. 6.23) needed to reproduce the POC profiles reported in Fig. 6.5 (see also Appendix E).  Figure 6.23: Model B for POC dynamics. B1, B-1, C1, C-1, SL and SXL were from the optimal run of model B for Th (Fig. 6.22). PF, PL, PXL = POC production rate for each particle size at the surface; Rx F , Rx L  and Rx XL  are the remineralization rates for fine, large and extra-large particles, respectively. Particulate organic carbon is added at the surface (z = 0) in each class of particles at rates PF, PL, and PXL for fine, large and extra-large particles, respectively. Their values were chosen so as to produce a total sinking POC fluxes at 50m (275mgC/m 2 /d in the best run) close to the reported export production at OSP (net primary productivity in summer ~850 mgC/m 2 /d, f-ratio ~ 0.25, Harrison, 2002). Through a series of sensitivity tests (Appendix E), I found that to reproduce the POC profiles in Fig. 6.5, the POC remineralization rate from large particles needed to be higher than for fine particles, which is consistent with our earlier qualitative interpretation of the difference in POC/ 230 Th between the large and small particles (section 6.3.5). On the other hand, the remineralization rate of extra-large fast-sinking particles can be largely neglected. Because of their high sinking rates, the decomposition of organic matter from these particles while they settle is small. The decrease in extra-large POC concentration with depth (Fig. 6.5) is mostly due to  292  disaggregation. POC production at the surface in the extra-large particle pool has a large impact on the [POC] concentration profiles for all three particle sizes. The sensitivity tests reported in Appendix E indicate, however, that model B is unable to reproduce the POC profiles for the three classes of particles at the same time, if we keep the particle dynamics parameters established with the thorium isotopes (Fig. 6.22). One of the best fits (Fig. 6.24) was obtained with the following parameters for POC: PF = 106mmolC/m 3 .y -1 , PL = 0, PXL = 212mmolC/m 3 .y -1 ; Rx F = 0, Rx L = Rx L 50*(Z/50) ^-0.4 (Rx L 50 = 600/y); Rx XL  = Rx XL 50*(Z/50) ^-0.4 (Rx XL 50=5/y). While this set of parameters reproduce reasonably well the POC profiles for large and extra-large particles, they significantly underestimate the POC concentration associated with fine particles (Fig. 6.24). This problem likely arises from the relative simplicity of model B, which does not consider a suite of possible interactions such as the formation of fine particles directly from the extra-large particles. Adding such an additional source for fine particles could clearly increase the POC associated with fine particles without significantly affecting the POC concentration with the large and extra-large particles. The next step is thus to consider a more comprehensive model of particle dynamics that considers all the possible interactions between dissolved and the three size classes of particles (Fig. 6.25 and 6.26).   293   Figure 6.24: POC concentrations (mmolC/m 3 ) generated by Model B with the set of parameters reported in Fig. 6.22 and PF = 106mmolC/m 3 .y -1 , PL = 0, PXL = 212mmolC/m 3 .y -1 ; Rx F = 0, Rx L = Rx L 50*(Z/50) ^-0.4 (Rx L 50 = 600/y); Rx XL  = Rx XL 50*(Z/50) ^-0.4 (Rx XL 50=5/y).  0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 0 500 1000 1500 2000 (mmolC/m3) D e p th (m ) POCf 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 0 500 1000 1500 2000 (mmolC/m3) D e p th (m ) POCl 0 0.05 0.1 0.15 0.2 0.25 0.3 0 500 1000 1500 2000 (mmolC/m3) D e p th (m ) POCxl  294  6.4.5.3 Model C The parameters of the particle dynamic model C depicted on Fig. 6.25 were varied to try reproducing the profiles of 234 Th and 230 Th measured at OSP. The following equations were used: P + K-1 F + G-1 L + H-1 XL= (K1 + G1 + H1+ ) D    (6.15) K1 D + B-1 L +J-1 XL= (K-1 + B1 + J1 +) F + SF F/z              (6.16) G1 D + B1 F+C-1 XL = (B-1 + C1 + G-1 +) L + SL L/z              (6.17) H1 D+ J1 F+ C1 L = (C-1 + H-1 + J-1 +) XL + SXL XL/z   (6.18)  Figure 6.25: Model (C) used to describe the complex thorium cycling in seawater. P = Th production in the water column from U decay; D = activity of dissolved Th; F = activity of thorium adsorbed on fine suspended particles; L = activity of thorium adsorbed on large sinking particles; XL = activity of  295  thorium adsorbed on extra-large sinking particles; λ = Th decay; K1 and K-1 are the adsorption and desorption rate constants between dissolved and fine particles; G1 and G-1 are the adsorption and desorption rate constants between dissolved and large particles; H1 and H-1 are the adsorption and desorption rate constants between dissolved and extra-large particles; B1 and B-1 are the aggregation and disaggregation rate constants between fine and large particles; C1 and C-1 are the aggregation and disaggregation rate constants between large and extra-large particles; J1 and J-1 are the aggregation and disaggregation rate constants between fine and extra-large particles; SF = sinking rate of fine particles; SL = sinking rate of large particles; SXL = sinking rate of extra-large particles.  Figure 6.26: Model (C) used to describe the complex POC dynamics in seawater. RF, RL and RXL are the POC remineralization rates on fine, large and extra-large particles; PF, PL and PXL are the POC production at the surface. A reasonable fit can be obtained with the following settings of parameters (Figs. 6.27, 6.28): K1 = K1 50 *(Z/50)^-0.4, (K1 50 =8/y) and K-1 = 6/y; B1 = B1 50 *(Z/50)^-0.4, (B1 50 =40/y) and B-1 = 150/y (Clegg et al., 1991); C1 = 2/y and C-1 = 50/y; G1 = 1/y, G-1 = 50/y; H-1 = 400/y; J1 = 0.2/y and J-1 = 100/y; SF = SF 50 *(Z/50) ^-0.15, (SF 50 =100m/y), SL = 5m/d and SXL =(SXL 50 *Z/50)/2, SXL 50 =50m/d); PF = 106 mmolC/m 3 .y -1 , PL=0, PXL=318  296  mmolC/m 3 .y -1 ; Rx F  = Rx F 50*(Z/50) ^-0.8 (Rx F 50=5/y), Rx L  = Rx L 50*(Z/50) ^-0.8 (Rx L 50=1000/y), Rx XL  = Rx XL 50*(Z/50) ^-0.8 (Rx XL 50=5/y).  Figure 6.27: Th concentrations produced by model C. 0 0.1 0.2 0.3 0.4 0 1000 2000 (dpm/m3) D e p th (m ) D230 0 0.05 0.1 0 1000 2000 (dpm/m3) D e p th (m ) F230 0 0.01 0.02 0.03 0 1000 2000 (dpm/m3) D e p th (m ) L230 0 2 4 6 x 10 -5 0 1000 2000 (dpm/m3) D e p th (m ) XL230 0 500 1000 1500 2000 2500 0 1000 2000 (dpm/m3) D e p th (m ) D234 0 200 400 600 800 0 1000 2000 (dpm/m3) D e p th (m ) F234 0 50 100 0 1000 2000 (dpm/m3) D e p th (m ) L234 0 0.1 0.2 0.3 0.4 0 1000 2000 (dpm/m3) D e p th (m ) XL234  297   Figure 6.28: POC concentrations (mmolC/m 3 ) produced by model C. This is the model that was used to better understand the impact of particle dynamics on 230 Th-normalization flux calculations. 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 0 500 1000 1500 2000 (mmol/m3) D e p th (m ) POCf 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 0 500 1000 1500 2000 (mmol/m3) D e p th (m ) POCl 0 0.05 0.1 0.15 0.2 0.25 0.3 0 500 1000 1500 2000 (mmol/m3) D e p th (m ) POCxl  298  POC fluxes from Model C The POC fluxes generated by model C can be calculated by adding [POC]*S for all particle sizes (Fig. 6.29). The largest contribution to POC fluxes derived from Model C is associated with the extra-large particles, because of their much larger sinking rates (increasing linearly from 25m/d at 50 m to 800m/d at 1600 m). With their much lower sinking rates, large (5 m /d) and fine (100 – 60 m /y) particles contribute much less, notwithstanding the higher POC concentrations associated with these particle sizes. Within the model, these fluxes could be viewed as the “true” fluxes against which 230Th normalization could be compared.  Figure 6.29: POC fluxes calculated by [POC]*S from Model C 0 200 400 600 800 1000 1200 1400 1600 1800 0 5 10 15 20 25 30 35 40 45 50 D ep th  ( m ) POC fluxes (mg/m2/day) Flux[Total] Flux[F} Flux[L] Flux[XL]  299  Whether these fluxes, calculated by tuning the parameters of the particle dynamics model C to observations, provide a good estimate of the POC flux at station Papa in June, is still an open question, however, that needs to be further addressed. In particular: - Model C is a steady-state model and considering the short residence time of the extra-large particles in the water column, it is unlikely that the steady-state assumption holds for this particle pool. To address this question, the extra-large particles must be collected seasonally to better understand the variability of this particle pool. The data that should be used in the model should be averaged over at least one entire year. - Model C assumes that all extra-large particles are produced at the surface. It is however also conceivable that some of these large particles are produced by heterotrophs living in deeper water. To address this question, a closer and more systematic examination of the material collected by this method is required to try better establish the origin of these large particles and to ascertain that “swimmers” have not been included in the samples. - The parameters that yield Fig. 6.27 and 6.28 have not yet been optimized using a least-square method, and do not