Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Characterization and investigation of submarine groundwater discharge from a coastal aquifer into the… Caulkins, Joshua Lee 2004

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

Item Metadata

Download

Media
831-ubc_2004-0184.pdf [ 11.58MB ]
Metadata
JSON: 831-1.0052755.json
JSON-LD: 831-1.0052755-ld.json
RDF/XML (Pretty): 831-1.0052755-rdf.xml
RDF/JSON: 831-1.0052755-rdf.json
Turtle: 831-1.0052755-turtle.txt
N-Triples: 831-1.0052755-rdf-ntriples.txt
Original Record: 831-1.0052755-source.json
Full Text
831-1.0052755-fulltext.txt
Citation
831-1.0052755.ris

Full Text

CHARACTERIZATION A N D INVESTIGATION OF S U B M A R I N E GROUNDWATER DISCHARGE F R O M A C O A S T A L AQUIFER INTO THE NEARSHORE ENVIRONMENT BY JOSHUA LEE CAULKINS B.Sc , University of California, Santa Cruz, 1998 A THESIS SUBMITTED IN PARTIAL F U L F I L L M E N T OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE F A C U L T Y OF G R A D U A T E STUDIES Department of Earth and Ocean Sciences We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH C O L U M B I A April 2004 © Joshua Lee Caulkins, 2004 Library Authorization In presenting this thesis in partial fulfillment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Joshua Lee Caulkins 22/04/2004 Name of Author (please print) Date (dd/mm/yyyy) Title of Thesis: CHARACTERIZATION AND INVESTIGATION OF SUBMARINE G R O U N D W A T E R D I S C H A R G E F R O M A C O A S T A L AQUIFER INTO THE N E A R S H O R E ENVIRONMENT Degree: Master's of Science Year: 2004 Department of Earth and Ocean Sciences The University of British Columbia Vancouver, B C Canada ABSTRACT The current trend o f g lobal populations m o v i n g increasingly to h igh density, coastal cit ies places a greater emphasis upon the water qual i ty o f aquifers supply ing those cit ies. P rob lems that affect coastal aquifers (e.g. saltwater intrusion, non-point source po l lu t ion) w i l l be ampl i f i ed as this trend increases. The goal o f this research project is to understand the processes that control freshwater/saltwater interactions i n the coastal zone, spec i f ica l ly i n areas o f h igh submarine groundwater discharge ( S G D ) . A site i n N W F l o r i d a a long the G u l f Coast is a wel l -documented S G D locale and provides an excellent opportunity to examine h o w t idal fluctuations, differential pressure i n the seabed and groundwater seepage rates are interrelated. Exper iments at the site focus o n characterizat ion o f the nearshore aquifer, submarine groundwater discharge at the seabed and numer ica l mode l i ng o f the system. N e w onshore wel l s at the f ie ld site show that hydraul ic conduct ivi t ies i n onshore region are s imi la r to those i n the offshore region. S l u g tests and water l eve l moni tor ing o f the onshore wel l s are used to measure a seaward, hor izonta l hydraul ic gradient. Di rec t measurements o f discharge are conducted w i t h an automated seepage meter, w h i c h shows that peak discharge rates tend to occur at the t ransi t ion between h igh and l o w tides. A new apparatus ca l led a differential piezometer system ( D P S ) is designed and used to measure differential hydraul ic head i n the seabed created by seepage and t idal interactions. Th i s system fai led to accurately portray differential head fluctuations i n the seabed as a result o f cal ibrat ion error. Sa l in i ty samples are col lected f rom on and offshore wel l s and from a newly instal led mul t i - l eve l w e l l , the data o f w h i c h are made into sal ini ty profiles. These profiles define the i i boundaries o f a saline wedge and m i x i n g zone i n the nearshore region. F R A C 3 D V S is f l o w and transport groundwater mode l that is used to design and run a 1-dimensional numer ica l mode l . The mode l results conf i rm the temporal effects o f t ida l e levat ion o n discharge rates observed i n the seepage meter. i i i Table of Contents A B S T R A C T i i Table o f Contents i v Table o f Tables v Table o f Figures v A C K N O W L E D G E M E N T S v i i i 1. P r o b l e m B a c k g r o u n d 1 2. H y d r o g e o l o g i c Setting and Scope o f Research 7 2.1 H y d r o g e o l o g i c Setting 7 2.2 P r io r Submarine Groundwater Discharge Studies at F S U M L 8 2.3 Research O v e r v i e w 9 2.3.1 Nearshore A q u i f e r Character izat ion 10 2.3.2 Submarine Groundwater Discharge Exper iments 11 2.3.3 N u m e r i c a l mode l i ng o f the F l o w R e g i m e at F S U M L 13 3. F i e l d Measurements o f the Coas ta l Aqu i fe r , S G D and Pressure Changes i n the Seabed 21 3.1 Nearshore A q u i f e r Character izat ion 21 3.1.1 Determina t ion o f G e o l o g i c Contacts at D e p t h 21 3.1.2 Onshore Hor i zon t a l H y d r a u l i c Gradient and T i d a l S igna l i n Onshore W e l l s . 21 3.1.3 S l u g Test ing 24 3.1.4 Sa l in i ty p ro f i l ing 25 3.2 Submarine Groundwater Discharge Exper iments 28 3.2.1 T i d a l Measurements 28 3.2.2 Seepage Measurements 29 3.2.3 Different ia l Piezometer Sys tem 32 3.3 Resul ts and D i s c u s s i o n 36 3.3.1 Nearshore A q u i f e r Character izat ion 36 3.3.2 Submarine Groundwater Discharge 45 4. N u m e r i c a l M o d e l o f the F S U M L Site 95 4.1 O n e - D i m e n s i o n a l Compute r S imula t ion 95 4.1.1 Rat ionale and F R A C 3 D V S Descr ip t ion 95 4.1.2 Cons t ruc t ion Deta i l s 96 4.2 Resul ts and D i s c u s s i o n 98 5. C o n c l u s i o n ; 112 A p p e n d i x A - S l u g Test Recove ry Da ta 115 A p p e n d i x B - S l u g Test Recove ry Data 118 A p p e n d i x C - S l u g Tests - N o r m a l i z e d d rawdown versus t ime 119 A p p e n d i x D - Sa l in i ty and Ch lo r ide Ana lyses 126 References 129 B i o g r a p h i c a l Sketch 132 i v Table of Tables Table 1-1. Estimates o f S G D from Recent Studies 2 Table 3-1. Character iza t ion o f N e w Onshore W e l l s 23 Table 3-2. Independent Different ia l H e a d Measurements 35 Table 3-3. H y d r a u l i c Conduc t i v i t y o f W e l l s at F S U M L 39 Table 3-4. Seepage Da ta P S D results 55 Table 4-1 . M a t e r i a l properties o f Sur f ic ia l A q u i f e r and Intracoastal Fo rma t ion 97 Table of Figures Note: Figures are found at the end of each chapter. Figure 1-1. Processes affecting submarine groundwater discharge (adapted f rom L i et a l . 1999) 6 F igure 2-1 . M a p o f structural features o f F l o r i d a w i t h loca t ion o f Turkey Po in t f i e ld site as w e l l as the A p a l a c h i c o l a Embayment (Scott, 1992) 15 F igure 2-2. Site M a p 16 F igure 2-3. Resul ts f rom Burnett et a l . 2002. N o t e that highest discharge rates are observed between h i g h and l o w tides 17 F igure 2-4. Cross-sec t ion o f M o n i t o r i n g W e l l Sampl ing Depths 18 F igure 2-5. D i a g r a m o f Different ia l Piezometer System w i t h Seepage meter 19 F igure 2-6. N E X R A D Precipi ta t ion data 20 F igure 3-1. Onshore Cross-sect ion showing locat ion o f P , N and A nests (and the water table pos i t ion at 12:10pm o n Augus t 25th, 2002) 57 F igure 3-2. Water Table E leva t ion ( N A V D 8 8 ) and Ra in fa l l Da ta 58 F igure 3-3. D e p t h to Water Table and Ra in fa l l Da ta 59 F igure 3-4. W e l l P I Water Table E l eva t i on and Ra in fa l l Da ta 60 F igure 3-5. W e l l N I Water Table E l eva t i on and R a i n f a l l Da ta 61 F igure 3-6. W e l l s Nes t P and N w i t h water table "snapshots" 62 F igure 3-7. Wate r loo Prof i le r sampl ing at offshore site (site 1). N o t e : W e l l s beh ind the scaffolding are those o f B C - n e s t 63 Figure 3-8. Sa l in i ty Prof i le w i t h Dep th - August /September 2002 64 F igure 3-9. Sa l in i ty Prof i le as Percent Freshwater - August /September 2002 65 F igure 3-10. Sa l in i ty Prof i le w i t h Dep th - August /September 2000 66 F igure 3-11. T i d e M e t e r 67 F igure 3-12. T i d a l Osc i l l a t ions at F S U M L 68 F igure 3-13. Seepage Rate for R u n 1 at Site 1 69 v Figure 3-14. Seepage Rate (cm/day) for R u n 1 at Site 2 70 F igure 3-15. Seepage Rate (cm/day) for R u n 2 at Site 1 71 F igure 3-16. Photograph o f D P S at Site 2 '.. 72 F igure 3-17. Inside the D P S B o x 73 F igure 3-18. T i d a l Fluctuat ions at W e l l P I 74 Figure 3-19. T i d a l Fluctuat ions at W e l l N I 75 F igure 3-20. Seepage Rate, T i d a l E l eva t ion and Ra in fa l l Rate dur ing S 1 R 1 76 Figure 3-21. T i d a l E l eva t i on and Seepage Rate for S 1 R 2 . N o significant ra infal l was recorded 77 F igure 3-22. Seepage Rate, T i d a l E l eva t ion and Ra in fa l l Rate dur ing S 2 R 1 78 F igure 3-23. D P S Ca l ib ra t ion Exper iments for Sensor 1 79 F igure 3-24. D P S Ca l ib ra t ion Exper iments for Sensor 2 80 F igure 3-25. S 1 R 2 - D P S Sensors 1 and 2 - Processed w i t h a l l ca l ibra t ion data and Post f i e ld w o r k ca l ibra t ion data. Da ta processed w i t h post f ie ld w o r k ca l ibra t ion is plotted w i t h th icker l ines 81 F igure 3-26. Temperature Range i n D P S B o x for S 1 R 2 82 F igure 3-27. S 1 R 1 - D P S Sensor 1 and T i d a l E l eva t ion 83 F igure 3-28. S 1 R 1 - D P S Sensor 2 and T i d a l E l eva t ion 84 F igure 3-29. S 2 R 1 - D P S Sensor 1 and T i d a l E l eva t i on 85 F igure 3-30. S 2 R 1 - D P S Sensor 2 and T i d a l E l eva t i on 86 F igure 3-31. S 1 R 2 - D P S Sensors 1 and 2 and T i d a l E l eva t i on 87 F igure 3-32. Per iodogram for T i d a l Da ta w i t h 10 minute sampl ing frequency 88 F igure 3-33. Per iodogram for T i d a l Da ta w i t h 5 minute sampl ing frequency 89 F igure 3-34. Per iodogram for Seepage Da ta o f S1R1 90 F igure 3-35. Per iodogram for Seepage Data o f S 2 R 1 91 F igure 3-36. Pe r iodogram for Seepage Data o f R u n 2 at Site 1 92 F igure 3-37. Pe r iodogram for W e l l P I Water Table E leva t ion Data . . . 93 F igure 3-38. Per iodogram for W e l l N I Water Table E leva t ion Data 94 F igure 4 -1 . Conceptua l M o d e l for 1-Dimensional S imula t ion 103 Figure 4-2. S imula ted Different ia l H e a d and Seepage Rate f rom I D C o l u m n M o d e l us ing T i d a l Da ta col lec ted dur ing S 1 R 2 104 Figure 4-3 . M o d e l results f rom 1-dimensional mode l us ing t ida l data col lec ted dur ing S 1 R 2 105 Figure 4-4. Measu red Seepage Rate f rom S 1 R 2 plotted w i t h S imula ted Seepage Rate and Different ia l H e a d f rom I D C o l u m n M o d e l . No te that the s imulated seepage rate is i n units o f m/day, w h i l e the measured rate is i n cm/day 106 F igure 4-5. Measu red Seepage Rate f rom S 1 R 1 plotted w i t h S imula ted Seepage Rate and Dif ferent ia l H e a d f rom I D C o l u m n M o d e l . N o t e that the s imulated seepage rate is i n units o f m/day, w h i l e the measured rate is i n cm/day 107 Figure 4-6. Measu red Seepage Rate f rom S 2 R 1 plotted w i t h S imula ted Seepage Rate and Dif ferent ia l H e a d f rom I D C o l u m n M o d e l . N o t e that the s imulated seepage rate is i n units o f m/day, w h i l e the measured rate is i n cm/day 108 Figure 4-7. S imula ted Seepage Rate based on t idal data f rom S 1 R 2 per iod plotted against T i d e He igh t above Seabed. N o t e the circular nature o f the plot 109 v i Figure 4-8. S imula ted Seepage Rate based on t idal data f rom S 1 R 1 per iod plotted against T i d e He igh t above Seabed 110 Figure 4-9. S imula ted Seepage Rate based o n t ida l data f rom S 2 R 1 per iod plotted against T i d e He igh t above Seabed I l l v i i ACKNOWLEDGEMENTS T h i s project was funded f rom a Natura l Science and Engineer ing Research C o u n c i l grant. Spec ia l thanks are extended to the staff o f F l o r i d a State Un ive r s i t y M a r i n e L a b i n St. Teresa for their f l ex ib i l i t y and assistance i n f ie ld experiments: B o b b y Henderson, D r . John H i t r o n , N e l s o n Wheeler , Pa t r ic ia Jackson, Dennis T ins l ey and B u t c h . A d d i t i o n a l l y , thanks go to the F l o r i d a State Un ive r s i t y Department o f Oceanography graduate students and professors w h o were generous i n their equipment loans, advice and assistance: H a r m o n Harden, Henr ie ta D u l a i o v a , Chr i s t ina Springer, D r . B i l l Burnet t and D r . Jeffrey Chanton. I a m grateful to C r a i g T h o m p s o n for his assistance dur ing f ie ldwork . H i s insight dur ing the computer s imulat ions was also invaluable. I w o u l d l i ke to thank m y parents, m y sister and brother for their words o f encouragement. M y deepest gratitude goes to m y partner, Reg ina , for her constant encouragement and support. Wi thou t her commitment to m y endeavor, this thesis w o u l d not have been possible. F i n a l l y , thanks to m y supervisor, D r . Les l i e Smi th , and the members o f m y commit tee , D r . Roger B e c k i e and D r . U l r i c h M a y e r for their teaching and guidance. v m 1. Problem Background The definition of submarine groundwater discharge ( S G D ) has been refined over the last two decades as the driving processes controlling discharge have become more widely understood. Zekster et al. (1983), in exploring S G D from a global water budget perspective, equated S G D to the net groundwater discharge to the ocean. Church (1996) defined S G D as the direct groundwater outflow across the ocean-land interface into the ocean. L i et al. (1999) provided a more complete description, defining S G D as the combination of net groundwater discharge, the outflow due to wave-setup-induced groundwater circulation, and that due to tidally driven oscillating flow. They present a conceptual model as follows, DSGD=Dn+Dw+D, where D S G D is the total combined discharge across the seabed, D N is the net groundwater component, D W is the wave-induced groundwater circulation component, and D T is the tidally driven component of flow (Figure 1-1). The current research w i l l focus on the net groundwater discharge and the tidally driven components of the model presented above. Historically, submarine groundwater seeps and springs have been viewed as hydrologic "curiosities" rather than as a phenomenon worthy o f scientific research. (Kohout 1966). The number of publications produced over the last two decades investigating submarine groundwater discharge into the nearshore environment shows that this is no longer the case (Bokuniewicz 1980; Taniguchi and Fukuo 1993; Moore 1996; Cable et al. 1996; Burnett et al. 2001; Taniguchi 2002; Langevin 2003). A few recent examples of S G D quantification studies are listed in Table 1-1. S G D has been recognized as a potentially significant driver of nutrients to the coastal oceans, l ikely 1 impac t ing nearshore coastal ecosystems (Oberdorfer et a l . 1990; G i b l i n and Gaines 1990; M o o r e 1996). The in i t ia t ion o f S G D is freshwater ( D n ) f l o w i n g through unconf ined or confined aquifers that extend into the nearshore coastal zone. The processes d r i v i n g S G D i n the coastal zone are not w e l l understood and current research seeks to expose h o w onshore hydrau l ic gradients, t ida l p u m p i n g o f coastal aquifers and leakage across conf in ing units at depth interrelate to produce S G D at the seabed. Table 1-1. Estimates of SGD from Recent Studies Study Area Methods Estimated Discharge (cm/day) Reference Cape C o d , M A M a n u a l Seepage Meters 2.4 to 7.2 G i b l i n and Gaines , 1990 L a k e B i w a , Japan Automated and M a n u a l Seepage Mete rs 0 to 24 (manual) 3 to 23 (automated) T a n i g u c h i and F u k u o , 1993 N E coastal G u l f o f M e x i c o , F L Tracers: 2 2 2 R n 2 to 10 Cab le , e t a l . 1996 N E coastal G u l f o f M e x i c o , F L M a n u a l Seepage Mete rs 3 to 22 Cab le , et a l . 1997a N E coastal G u l f o f M e x i c o , F L M a n u a l Seepage Meters 0.1 to 1 Rasmussen , 1998 O s a k a B a y , Japan Au toma ted Seepage M e t e r 0.9 to 43* T a n i g u c h i , 2002 N E coastal G u l f o f M e x i c o , F L Tracers: 2 2 2 R n 8.6 to 13 Lamber t and Burnett , 2003 N E coastal G u l f o f M e x i c o , F L Tracers: 2 2 ( 5 R a 10.8 M o o r e , 2003 N E coastal G u l f o f M e x i c o , F L Au toma ted and M a n u a l Seepage Meters 2 to 50* (manual) 1 to 77* (automated) T a n i g u c h i , et a l . 2003 Waquo i t B a y , Cape C o d , M A M a n u a l Seepage Mete rs 3 to 37* M i c h a e l , et a l . 2003 Shelter Island, L o n g Island, N Y Ul t rason ic Seepage M e t e r 40 to 200* Paulsen , et a l . i n press * Es t imated from figures i n paper Despi te significant advancements i n the f ie ld o f submarine groundwater discharge, quant i fying S G D and determining its effects on the nearshore environment is 2 p r o v i n g to be a serious challenge for hydrogeologis ts , oceanographers and coastal zone managers. D i rec t methods o f quant i fying groundwater discharge r ema in subject to large errors and current best practices i nvo lve us ing mul t ip le approaches. These approaches have most recently focused on direct measurement w i t h manual or automated seepage meters, natural tracer concentrations ( R a and R n ) o f onshore and offshore samples and applications o f numer ica l mode l i ng (Burnett et a l . 2002). M o o r e and C h u r c h (1996) chal lenged hydrologis ts to develop S G D models that w i l l inc lude t ida l p u m p i n g due to diurnal , month ly , seasonal or longer changes o f sea l eve l ; saltwater in t rus ion and changes i n G W usage; and m i x i n g and chemica l reactions w i t h i n coastal aquifers. One aspect o f this research seeks to expand the current k n o w l e d g e o f S G D b y examin ing onshore groundwater f l ow i n a coastal aquifer through the measurement o f nearshore water table elevations i n onshore we l l s and through the creat ion o f sa l in i ty profi les f rom samples o f onshore and offshore we l l s . One hypothesis is that b y determining aquifer parameters from onshore we l l s , a rough estimate o f the discharge at the shoreline can be obtained. T h i s estimate is equivalent to the net groundwater discharge, or D„ from the conceptual mode l above, w h i c h can be used to obtain a clearer understanding o f freshwater input to the system. Sa l in i ty data f rom bo th o n and offshore we l l s should prov ide a general locat ion o f the saltwater/freshwater contact, g i v i n g a spatial understanding o f the nearshore m i x i n g zone. Recent investigations have shown that there is a need for further explora t ion into pressure fluctuations w i t h i n the seabed, w h i c h w o u l d expose the nature o f ver t ica l hydraul ic gradient changes (Smi th and Z a w a d z k i 2003). E x p l o r i n g h o w ver t ica l hydrau l ic gradients fluctuate w i t h discharge and tides should provide insight into h o w the m i x i n g o f 3 groundwater and seawater occur w i t h i n the seabed. W o r k o f this nature has been attempted w i t h some success i n streams ( K e l l y and M u r d o c h 2002) yet these systems are i n pseudo-steady-state and are not subject to the d iurna l and semi-d iurna l head fluctuations affecting coastal aquifers. F a n g et a l . (1993) l ooked at pore pressure fluctuations i n abyssal p l a i n sediments near the A t l a n t i c mid-ocean ic r idge but their differential pressure transducer system ( P U P P I ) was designed for deep sea w o r k and is not necessari ly appl icable to the present study. N o study to date has attempted to investigate nearshore ver t ical hydraul ic head fluctuations b y c o u p l i n g direct measurements o f discharge w i t h pressure measurements i n the seabed i n a t i da l l y dominated setting. Thus , it was not k n o w n pr ior to this study h o w tides affected fluctuations o f differential head i n sha l low, nearshore sediments or h o w those fluctuations related to S G D . T h e hypothesis is that i f correlations can be made between temporal fluctuations o f differential head i n sha l low sediments and submarine groundwater discharge rates, a greater understanding o f the nature o f S G D w i l l emerge. T o test this hypothesis, two dual-port piezometers hyd rau l i ca l l y connected to differential pressure transducers were inserted para l le l to shore, one on each side o f an automated seepage meter. Successful ca l ibra t ion and deployment o f a dual-port system should increase the ab i l i ty o f scientists and coastal zone managers to understand rates and variations i n discharge across the seabed. N u m e r i c a l models capable o f reso lv ing density-dependent f l o w have been used to understand the groundwater/seawater interactions i n numerous recent publ ica t ions (Smi th and Z a w a d z k i 2003; L a n g e v i n 2003; U c h i y a m a et a l . 2000; L i et a l . 1999; A t a i e - A s h t i a n i et a l . 1999). A 1-dimensional, non-density-dependent f l ow m o d e l was created to predict 4 S G D rates and the ver t ica l hydrau l ic gradients w i t h i n the nearshore sha l low seabed. T i d e data f rom the f ie ld experiments were used as input for the m o d e l and the tempora l effects o f tides o n discharge rates and differential pressure i n the seabed were examined . The hypothesis is that accurate m o d e l predict ions, w h e n compared w i t h f i e ld measurements o f S G D and differential pressure w i t h i n the seabed, w o u l d indicate that the m o d e l has correct ly portrayed the system. Fa i lure to accurately portray the system shou ld point to areas where increased f ie ld or lab w o r k is required. The current study seeks to characterize the nature o f processes affecting the discharge o f groundwater to the nearshore environment. The exper imental focus is o n t ida l and onshore hydrau l ic properties and does not explore w a v e induced re-c i rcula t ion as proposed i n the conceptual mode l . The f o l l o w i n g three chapters explore these questions i n detai l . Chapter 2 presents the hydrogeologic setting o f the f i e ld area and describes the scope o f the research conducted. Chapter 3 is an inves t igat ion o f the nearshore aquifer characterizat ion w o r k and submarine groundwater discharge experiments conducted at the site, where each experiment is out l ined and the results discussed. Chapter 4 demonstrates the construct ion o f a 1-dimensional f l o w m o d e l designed to predict seepage rates and seabed pressure fluctuations observed i n the f ie ld . 5 land ocean Figure 1-1. Processes affecting submarine groundwater discharge (adapted from L i et al. 1999). 6 2. Hydrogeologic Setting and Scope of Research 2.1 Hydro geologic Setting The field area is situated in the NE Gulf of Mexico, along the Florida coast (Figure 2-1). Three major geologic units occur in the region immediately surrounding the field site. The Bruce Creek Limestone, a porous aquifer, is the uppermost unit of the Floridan Aquifer System. Overlying the Bruce Creek Limestone is the Intracoastal Formation, a very sandy, highly microfossiliferous, poorly consolidated, argillaceous, calcarenitic limestone (Schmidt 1984). The Intracoastal Formation produces artesian conditions in the Bruce Creek Limestone and is estimated to be 1 to 5 meters thick in the region immediately surrounding FSUML. Overlying the Intracoastal Formation are the Pleistocene-to-recent sands that make up the Surficial Aquifer System. These sediments are between 6 and 7 meters thick in the nearshore region, based on Waterloo Profiler data collected in 2002 (discussed below). The Intracoastal Formation and Bruce Creek Limestone dip locally to the south and regionally to the southwestward at angles generally less than 0.05°. The Intracoastal Formation is exposed north and east of the field site. The Floridan Aquifer becomes unconfined near where the Intracoastal Formation is exposed. The current research focuses on flow within the Surficial Aquifer yet recognizes that leakage from the Floridan Aquifer across the Intracoastal Formation may contribute to submarine groundwater discharge in the field area (Rasmussen 1998; Smith and Zawadzki 2003). 7 2.2 P r i o r Submar ine Groundwater Discharge Studies at F S U M L A number o f previous studies have explored processes that affect S G D at the study site. Rasmussen (1998) conducted extensive f i e ldwork at the site between 1995 and 1997 where groundwater discharge was measured d i rec t ly v i a manua l seepage meters and a groundwater f l ow m o d e l was used to simulate the hydro log ic system. F i v e piezometer nests w i t h a total o f 18 we l l s were instal led i n the Sur f ic ia l A q u i f e r and used to measure sal ini ty and hydrau l ic head. E a c h nest has 2 to 4 we l l s , inserted to var ious depths. One nest (A-nest) is located onshore, another nest (AB-nes t ) is located i n the intert idal zone and the remain ing nests ( B , B C , C and D ) are located offshore ( F i g u r e 2-2). In 1997, the Scient i f ic Commi t tee o n Oceanic Research ( S C O R ) created W o r k i n g Group 112 ( S C O R webpage, 2003), whose mandate was to investigate the " M a g n i t u d e o f Submar ine Groundwater Discharge and its Influence o n Coas ta l Oceanographic Processes." In A u g u s t o f 2000, W G 1 1 2 conducted an "Intercomparison Expe r imen t , " where different methods o f quant i fying S G D were implemented and evaluated. The team o f researchers estimated S G D us ing automated and manual seepage meters, groundwater and seawater chemis t ry and water l eve l measurements i n the offshore and onshore we l l s . The a i m o f that study was to compare S G D measurement techniques ( F i g u r e 2-3). Wate r l eve l measurements were not used because o f the coarse temporal resolu t ion o f the data. A d d i t i o n a l l y , surface seawater and pore water chemist ry analyses were conducted us ing radioact ive isotopes ( R a and R n ) . Three o f the research groups i n v o l v e d i n that study, Lambert /Burnet t ; T a n i g u c h i et a l . and M o o r e , produced independent estimates o f S G D rates us ing R a d i u m isotopes, seepage meters and R a d o n isotopes, respect ively. Lamber t and Burnett , us ing R a d i u m isotope ratios, estimated a S G D rate o f 8.6 to 13 cm/day w h i l e 8 M o o r e predicted a rate o f 10.8 cm/day us ing the R a d o n isotope ratios. T a n i g u c h i et a l . measured seepage rates that ranged from 2 to 50 cm/day for manua l seepage meters and 1 to 77 cm/day for automated seepage meters. A l l in tercomparison experiment results have been publ i shed i n a special edi t ion o f Biogeochemistry (66, 2003) . A project that was part o f the A u g u s t 2000 in tercompar ison experiment was an attempt to numer i ca l ly m o d e l the hydro log ic regime o f the f ie ld site. S m i t h and Z a w a d z k i (2003) used F E F L O W , a c o m m e r c i a l l y avai lable density-dependent f l o w m o d e l , to produce a two-d imens iona l f l ow and transport m o d e l based o n data co l lec ted at the site. The i r research was an attempt to b u i l d a hydrogeologic m o d e l for the site f rom w h i c h estimations o f S G D c o u l d be made. Offshore f l ow and rec i rcula t ion w i t h i n the saltwater wedge were inc luded i n their mode l . Th i s m o d e l under predicted the f ie ld measurements o f S G D at the site, p roduc ing a number approximate ly 1 to 2 orders o f magni tude l ower than those observed i n the f ie ld . F indla ter (2001) discusses the hydraul ic conduc t iv i ty o f sediments at the site and an adapted M O D F L O W s imula t ion where her m o d e l produces the same l o w discharge values that were observed b y S m i t h and Z a w a d z k i (2003). 2.3 Research O v e r v i e w A m a i n thrust o f this study is to investigate h o w discharge at the seabed is affected b y tides and differential pressure i n the seabed. A l l f i e ldwork for this research was completed over a 1-month per iod from A u g u s t 9 t h to September 9 t h , 2002 at the F S U M L site. Equ ipmen t calibrations were conducted between M a y and A u g u s t o f 2002, 9 at the University of British Columbia, Vancouver, Canada. A 1-dimensional, density-independent flow model is constructed and used to predict discharge rates at the site. 2.3.1 Nearshore Aquifer Characterization In order to properly identify the magnitude of groundwater flow to the nearshore area, a number of parameters required definition, including horizontal hydraulic gradients toward the ocean, hydraulic conductivity of the system and depth of geologic units. Prior to the current field work, five new onshore wells were installed in two nests, one with three wells ~80m inland of A-nest and the other with two wells ~20m from A-nest (Figure 2-2). These nests are labeled "P-nest" and "N-nest," respectively. The wells were surveyed by Roddenberry & Assoc., Inc. for latitude, longitude and elevation relative to North American Vertical Datum, 1988 (NAVD88). An average onshore horizontal hydraulic gradient was calculated from continuous measurements made by pressure transducers in wells PI and NI, the deepest wells of each nest noted above. This data can be used to estimate a volumetric flow of groundwater within the Surficial Aquifer that is delivered to the nearshore. An accurate horizontal hydraulic gradient also allows for appropriate boundary conditions to be applied in computer simulations. In order to constrain the hydraulic conductivity of the system slug tests were used to assess of sediments surrounding new onshore wells. Slug tests were attempted on all new wells and successful in wells PI, P2 and NI. The screened intervals of wells P3 and N2 did not extend below the water table and were not tested. 10 Sa l in i ty data was col lec ted at the site i n order to ident ify h o w sal in i ty is distributed w i t h depth i n the Sur f i c ia l Aqu i f e r . Sa l in i ty samples were co l lec ted from exis t ing on and offshore we l l s and us ing a Wate r loo P r o f i l e r ™ ( W P ) , w h i c h samples porewater at any depth i n the same borehole. The W P was also used to insert a m u l t i l e v e l w e l l at offshore site 1 ( labeled " M L " on Figure 2-2), approximate ly 2 6 m offshore o f the l o w tide l ine . T h i s method is advantageous because ver t ica l prof i les w i t h numerous, discrete sampl ing points can be made w i t h relative ease. T w o such prof i les were conducted dur ing the course o f the f ie ld study, one onshore and one offshore. O n e reason for us ing the Water loo Prof i le r was to determine i f penetration o f the Intracoastal Fo rma t ion was possible . A s stated previous ly , the Intracoastal F o r m a t i o n is " p o o r l y consol ida ted" and thus m a y have been accessible w i t h the W P , w h i c h is designed for use on ly i n saturated sediments. The combined sampl ing points (wel ls and Waterloo profi les) were plot ted onto a ver t ica l cross-section (Figure 2-4). 2.3.2 Submarine Groundwater Discharge Experiments T w o sites were located offshore at w h i c h simultaneous measurements o f seepage and differential pressure i n the seabed were made. Dif ferent ia l pressure ( A P ) is defined as the difference i n pressure between two ports separated by a k n o w n ver t ica l distance, Az. AP is measured b y a differential pressure transducer, w h i c h produces an electr ical output representing that differential pressure. The differential pressure is d i rec t ly related to differential hydrau l ic head (Ah) w i t h a standing manometer w h i c h was used to calibrate the output. Ca l ib ra t i on experiments were conducted both before and after the f i e ldwork i n order to ensure accurate data processing. Different ia l hydrau l ic head, defined b y the 11 equation Ah = (AP/pg)+ Az is d i rect ly propor t ional to differential pressure because a l l other variables are k n o w n . The density o f water i n the seabed is assumed to be constant to a depth o f at least l m , the m a x i m u m depth o f the D P S piezometers , based o n sa l in i ty profiles o f we l l s immedia te ly adjacent to the sites chosen and f rom samples col lec ted from sediments at depths o f 0 .7m be low the seabed. H o w differential pressure changes are affected b y t ida l osci l la t ions and h o w they relate to discharge rates has not yet been explored b y any o f the current literature invest igat ing submarine groundwater discharge. A differential piezometer system ( D P S ) was used to measure pressure fluctuations i n the seabed. T w o ports i n a single piezometer are hyd rau l i ca l l y connected to a differential pressure transducer located i n a secure box above the h i g h tide l ine . T w o , dual-port piezometers were inserted o n each side o f the automated seepage meter. A diagram o f a DPS/Seepage meter experiment is shown i n Figure 2-5. T i d a l p u m p i n g o f a coastal aquifer is defined as the osc i l l a t ing f l o w i n the aquifer induced b y t ida l fluctuations, contr ibut ing to the water exchange between the seabed and the ocean ( L i et a l . 1999). L i et a l . (1999) predict that the major i ty o f discharge across the seabed m a y be caused i n large part b y these t idal forcings and wave setup, as presented i n their conceptual m o d e l that was inspi red b y M o o r e (1996). D i rec t seepage measurements, described be low , are used to investigate h o w seepage rates are inf luenced through t ime b y t ida l osci l la t ions . A Taniguchi -s ty le automated seepage meter was used to di rect ly measure groundwater discharge (DSGD) across the seabed. Seepage rates are measured w i t h a heat pulse/thermistor system, described i n detail i n Chapter 2, w h i c h attach di rect ly to a d r u m s imi la r to those used for manual seepage meters. Discre te measurements were made at 5 minute intervals. 12 . T i d a l data was col lec ted so as to determine the magnitude o f t ida l inf luence o n S G D at the site. A tide meter was attached to a p i l i n g on the F S U M L dock, used to record the t idal fluctuations near the f ie ld site. A pressure sensor at the base o f a 2 m l o n g P V C pipe measured the height o f water. Later, this station was surveyed b y the surveying company ment ioned previous ly . Th i s e levat ion data was corrected to N A V D 8 8 . T h e loca t ion o f the tide meter was approximate ly 3 7 0 m west o f A-nes t . T w o experiment runs were conducted at each site, p roduc ing 4 sets o f seepage and D P S data o f va ry ing quali ty. These data sets were then super imposed over t idal data col lected w i t h the tide meter. Resul ts are discussed i n chapter 3. 2.3.3 Numerical modeling of the Flow Regime at FSUML A one-dimensional m o d e l was created us ing F R A C 3 D V S , a groundwater f l o w m o d e l produced b y researchers at the U n i v e r s i t y o f Water loo , Ontar io . W i t h this mode l , the author has attempted to simulate the D P S and seepage meter data. T h i s was done b y assigning estimated hydrau l ic conduct ivi t ies to the Sur f i c ia l A q u i f e r and the Intracoastal Fo rma t ion and app ly ing the observed t idal fluctuations to the seabed upper boundary o f the mode l . Non-dens i ty dependent f lows were used i n mul t ip le runs o f the m o d e l . R a i n f a l l rates were estimated from N E X R A D radar data because no weather station has been established at the site and none were instal led dur ing the course o f the experiments. The closest operating weather station was i n A p a l a c h i c o l a ( - 3 0 mi l e s to the west). In order to account for ra in events at F S U M L , N E X R A D data was used to estimate the amount o f precipi ta t ion at the site. Prec ip i ta t ion data was p rov ided b y the U S N a t i o n a l Oceanographic and Atmospher i c A g e n c y and J i m S m i t h o f Pr ince ton 13 2 Unive r s i t y , w h o processed the data for this project. N i n e , 1 k m cel ls conta in ing 15 minute averaged ra infa l l data were determined f rom D o p p l e r radar data co l lec ted at a station i n Tallahassee, F L . The center c e l l , w h i c h contained the f ie ld area, served as a p r o x y for a weather gauge. Unfortunately, a number o f periods exist where no data was col lected over the site. These periods are concentrated between A u g u s t 3 1 s t and September 6 t h , and are g iven a value o f " - 1 " i n F i g u r e 2-6 and i n a l l subsequent plot conta ining the ra infa l l data. F r o m conversations w i t h the night securi ty staff and f rom f ie ld notes, the o n l y ra infa l l to occur dur ing this per iod occurred o n A u g u s t 3 1 s t , i n the form o f l ight ra in i n the early m o r n i n g (3:00 to 6:00am) and l ight r a in sporad ica l ly throughout that day, c lear ing b y late afternoon. 14 Figure 2-1. Map of structural features of Florida with location of Turkey Point field site as well as the Apalachicola Embayment (Scott, 1992). 15 Legend A Seepage Meter/ DPS site (current study) • Piezometer • Waterloo profile o Multi-level well • Seepage meter (Rasmussen '98 and LOICZ "00) P3 PI P2 NI N2 X2 X3 DPS Site 1 A O ML B C 1 » • • » B C 4 BC2 BC3 C1 C2 A DPS Site 2 X4 X5 N Mean tide line tide line Y3 Y4 D1 D3 Meters X6I ?ure 2-2. Site M a p 16 CTQ C to e o W •t 0 • n & O o © CT 85 ST era' CT SZ1 ft cr » era a » » o cr a cr D cr era' cr » a a ST 3 is* a a Seepage Measurement (cm/day) {Q o 301 to cn cu Water Depth (cm) a Differential Piezometer Note: Not to Scale Figure 2-5. Diagram of Differential Piezometer System with Seepage meter 19 Rainfall (mm/hr) ora e n i ON Z 5 n n as O 3 CL » ss O o 0) oo o CO M CO ^ CO cn CO CO CO ro o ro ro ro 4^ go ro CO go w o CD ro o ro cn CO o CO cn o 2. £ ro SJ. < -w SL E~ ~ o 3 3 s< Q. O o ~" 0) CD (A 3" 3 o ro o ™ o. o-Si a 2 » </) ro_ tu a. 2 2. S"n ro c/) fi> c CO CO CD cn » ro ~ ° - • o 3 w ro CD =3 CD CD 3. Field Measurements of the Coastal Aquifer, SGD and Pressure Changes in the Seabed 3.1 Nearshore A q u i f e r Character izat ion 3.1.1 Determination of Geologic Contacts at Depth The Water loo Prof i le r was pushed to refusal at two locat ions at the F S U M L f ie ld site. The Prof i le r , a pneumatic hammer powered b y compressed air, was used to dr ive the steel rods into the Sur f i c ia l A q u i f e r w i t h a heat-treated steel t ip. The depth o f terminat ion was assumed to be the top o f the Intracoastal Fo rma t ion as the Prof i l e r is designed to penetrate o n l y non- l i th i f ied strata. A competent rock unit, as the Intracoastal Fo rma t ion appears to be, w o u l d not be penetrated b y the Prof i ler . Rasmussen (1998) had assumed that the water jet used to insert the offshore we l l s stopped at the top o f the Intracoastal Fo rma t ion but the Wate r loo Prof i le r was able to extend that estimate b y 2-3 meters i n both sampl ing locations. It should be noted, however , that a dense sediment or shel l layer -0 .1 to 0.2 m th ick was encountered at Rasmussen ' s terminal depth. U s i n g the depth o f terminat ion o f the profi ler , the top o f the Intracoastal Fo rma t ion near A-nes t is estimated to have an e levat ion o f - 5 . 5 m ( N A V D 8 8 ) and an elevat ion o f - 6 . 8 m ( N A V D 8 8 ) near B C -nest. T h i s is deeper than Rasmussen ' s we l l s b y 3 .5m and 2 .2m respect ively. 3.1.2 Onshore Horizontal Hydraulic Gradient and Tidal Signal in Onshore Wells The hor izonta l hydrau l ic gradient i n the onshore area was estimated i n order to constrain the estimate o f the f lux o f freshwater to the nearshore and p rov ide the locat ion o f the water table. Water levels i n we l l s P I and N I were moni to red con t inuous ly between Augus t 14 t h and September 7 t h , 2002. A Nor th -Sou th cross-section w i t h the locat ions and 21 depths o f the onshore w e l l nests is shown i n Figure 3-1. In this figure, a snapshot o f the water table was captured i n we l l s A l , N I and P I o n A u g u s t 2 5 t h at 12 :10pm because at that t ime an independent measurement o f the water table depth was made at w e l l A l . A l l distances i n this study are made relative to the l o w tide l ine, estimated for this study at the locat ion o f the B-nest. It should be noted at this point that w e l l A l had been damaged i n a fire sometime between 1998 and 2002. The standpipe had been mel ted to the ground surface and was sit t ing deformed and open. The w e l l was repaired w i t h a new standpipe and cap. A s there is some discrepancy between the depth o f the w e l l measured i n 2002 and the depth referenced i n the thesis b y Rasmussen (1998), it is l i k e l y that w i n d and storms had forced sand into the opening, decreasing the total depth o f that w e l l . Sol ins t L e v e l o g g e r s ™ were p laced immedia te ly above the screens i n each o f w e l l s P I and N I . The Leve logge r pressure transducers were used to measure the height o f the water c o l u m n above the sensor and can detect changes i n water l eve l to an accuracy o f + / - 1 m m and to a m a x i m u m water depth o f 5m. Readings were taken every 10 minutes and plotted against t ime. The depth o f each Leve logger be low the top o f cas ing is 4 . 5 m for w e l l P I and 4 . 4 m for w e l l N I . A l l measurements were converted from height o f water above sensor to e levat ion relat ive to N A V D 8 8 based o n survey data co l lec ted at each w e l l . E a c h logger was cal ibrated for 40 minutes at atmospheric pressure before be ing submerged i n a bucket conta in ing 0 .3m o f water for 90 minutes. These measurements were used dur ing data processing to calibrate the water levels i n each w e l l . A Sol ins t B a r o l o g g e r ™ was p laced at ground l eve l near w e l l N I and used to correct for atmospheric pressure fluctuations. Prec ip i ta t ion data is plotted w i t h the onshore water table data because it shows dis t inc t ly h o w the water table responds to large ra infa l l events (Figures 3-2). 22 Figure 3-3 is s imi la r to Figure 3-2 but the plot is o f water table depth b e l o w ground surface, not water table elevat ion as i n Figure 3-2. N o t e that w e l l P I has a ground surface elevat ion about 0 .3m higher than that o f w e l l N I . D e p t h and e levat ion data for a l l new onshore we l l s is presented i n Table 3-1. Table 3-1 1. Characterization of New Onshore Wells W e l l T o p o f C a s i n g E l e v a t i o n (m) G r o u n d Surface E l eva t i on (m) M i d - S c r e e n E l eva t i on (m) Screen L e n g t h (m) P I 5.311 3.871 -0.75 0.45 P 2 4.685 3.950 1.16 0.91 P3 5.003 3.953 2.716 0.45 N I 4.452 3.612 -0.245 0.41 N 2 4.100 3.700 1.917 0.91 Notab le i n the data col lec ted at each w e l l are t ida l signals and intense ra infa l l events that caused significant r i s ing o f the water table at each w e l l . T i d a l signals are stronger i n w e l l P I (amplitudes o f 1 to 1.5cm) than i n w e l l N I (ampli tudes o f 0.5 to 1cm) despite the fact that this w e l l is 5 6 m further onshore (Figures 3-4 and 3-5). T h i s is discussed i n greater detai l i n Sec t ion 3-3. T w o significant storm events (>30 m m / h r o f rainfal l) occur two-thirds o f the w a y through the data co l l ec t ion per iod on the evening o f A u g u s t 2 8 t h and the m o r n i n g o f A u g u s t 3 0 t h , ra is ing the water table at both we l l s (total increase o f 0 .28m at P I , 0 .16m at N I ) . These events effect ively bisect the data set, l eav ing two periods o f un i fo rm water table decl ine i n each w e l l w h i c h are used to determine the hydrau l i c gradient. There is one long per iod o f 11 days preceding the events and one shorter pe r iod o f 2 days f o l l o w i n g the events i n the Leve logger data. Three sma l l ra infa l l events occurred dur ing the first sequence but d i d not s ignif icant ly affect the water table e levat ion i n either w e l l . It should be noted that the ra is ing o f the water table at w e l l N I was m u c h s lower than i n 23 w e l l P I (Figures 3-4 and 3-5).,This is due to the fact that the gradient can o n l y be rea l i s t ica l ly calculated w h e n the water table i n each w e l l is d ropping (or r is ing) at approximate ly the same rate. The determination o f the hydrau l ic gradient is presented later i n this chapter. The two other we l l s i n P-nest and one other w e l l i n N-nes t were also investigated and characterized al though water table informat ion was not gathered outside o f a few isolated measurements. W e l l P 2 is the second deepest w e l l at the P-nest w i t h a depth b e l o w ground surface o f 2 .8m (ground surface to mid-screen). W e l l P 3 is the shallowest w e l l i n the nest w i t h a depth be low ground surface o f 1.24m (ground surface to m i d -screen). The water table never fu l ly rose above the screen o f this w e l l and was m u d d y whenever it was d ipped w i t h the water l eve l . See Figure 3-6 for an Eas t -Wes t cross-sectional d iagram o f P-nest. W e l l N 2 had been affected b y a brushfire and was bent over 90 degrees, po in t ing eastward. Figure 3-6 also has an Eas t -West cross-sect ion o f N-nest . The water table never reached higher than 0 .2m above the base o f w e l l N 2 . 3.1.3 Slug Testing W e l l s located at P-nest and N-nes t were tested to determine the hydrau l ic conduc t iv i ty o f the sediments surrounding the w e l l screen o f each w e l l . These were the o n l y we l l s i n the f ie ld area that had not been evaluated i n previous studies. The s lug was composed o f a sand-f i l led, P V C pipe 0.52 meters l ong and 0.032 meters i n diameter, p roduc ing a v o l u m e displacement o f 418 c m 3 . W e l l s N I , P I and P 2 were successful ly tested w h i l e water l eve l s . in we l l s N 2 and P3 were too l o w for effective measurement w i t h 24 the slug. Three tests were conducted i n each w e l l o n September 6 t h and 7 m , 2003 us ing the Sol ins t Leve loggers descr ibed previous ly . A s hydrau l ic conduct iv i ty estimates were not avai lable for the onshore sediments a p r io r i , equi l ibra t ion t ime for the s lug tests had to be estimated. The t ime allotted for each test was set to 5-7 minutes. T h i s p roved to be inadequate for complete recovery o f w e l l P I after test 1 and thus data f rom tests 2 and 3 for this w e l l were not used i n the analysis. A l l other tests recovered comple te ly w i t h i n this t ime interval i n we l l s P 2 and N I . The H v o r s l e v method (1951) was used to calculate the hydrau l ic conduct iv i t ies . In this method, an in i t i a l water level measurement is taken before the s lug is lowered into the w e l l w i t h an electronic interface meter (a w e l l "dipper") . The s lug is then instantaneously inserted into the w e l l and the decl ine i n the elevat ion o f the water table is measured w i t h t ime. S l u g test recovery plots are presented i n Appendix A, hydrau l ic conduct iv i ty calculat ions f rom slug tests are presented i n Appendix B and no rma l i zed displacement vs. t ime figures are presented i n Appendix C. 3.1.4 Salinity profiling Anothe r a i m o f the F S U M L research was to constrain estimates o f upward f lux o f fresh water f rom the Intracoastal Fo rma t ion and to examine the m i x i n g zone i n the Sur f ic ia l Aqu i f e r . B y determining the spatial var ia t ion o f sa l in i ty w i t h depth, an estimate can be made as to the locat ion and extent o f the saltwater/freshwater m i x i n g zone. W i t h these estimates, a f l ow m o d e l can be designed and run i n an attempt to simulate S G D data observed i n the f ie ld . A 1-dimensional f l ow m o d e l o f offshore site 1 was successfully designed and run, w h i c h is discussed i n detail i n Chapter 4. 25 Sa l in i ty p ro f i l i ng o f the Sur f i c ia l A q u i f e r was conducted b y sampl ing a l l exis t ing o n and offshore we l l s us ing a peristalt ic pump, and w i t h a direct-push sampl ing method i n two addi t ional locations. A l l exis t ing we l l s were purged and sampled once over the course o f the f ie ld study except for we l l s P 2 and P3 due to t ime constraints. These we l l s are expected to have s imi la r sa l in i ty values to w e l l P I . W e l l s were purged o f 3 w e l l vo lumes before a 5 0 m L sample was col lec ted for analysis. The Wate r loo Prof i ler" 1 1 is a direct-push sampl ing method that uses a pneumatic hammer to push steel A W rod into the surface sediments (see Figure 3-7). T w o sites i n the f ie ld area were prof i led w i t h the Wate r loo Prof i le r , one onshore approximate ly 5 m seaward o f w e l l nest A (labeled W P 1 i n Figure 2-3) and one offshore at site 1, approximate ly 4 m shoreward o f w e l l nest B C (labeled M L i n Figure 2-3). The offshore site is a permanent mu l t i l eve l w e l l instal led w i t h 6 ports spaced approximate ly l m apart, spanning the entire depth o f the Sur f i c ia l Aqu i f e r . Porewater samples were analyzed for sal ini ty, w h i c h were then mapped i n ver t ical cross-sections to produce sal ini ty profi les (Figure 3-8) and percent freshwater profi les o f the site (Figure 3-9). Percent freshwater is defined relative to seawater sampled f rom the water c o l u m n above site 1. Technic ians at F l o r i d a State U n i v e r s i t y analyzed the samples. T h e sa l in i ty values were determined us ing refractometry w h i l e chlor ide concentrations were determined v i a i o n chromatography. Tables containing sa l in i ty and chlor ide data are avai lable i n Appendix D. The 2002 sal ini ty p ro f i l i ng produced a s imi la r sa l in i ty cross-sect ion plot as the 2000 study (Figure 3-10). B o t h sa l ini ty profi les o f the f ie ld site co l lec ted dur ing the 26 current research (Figure 3-8) and for the Intercomparison Exper imen t i n A u g u s t 2000 (Smi th and Z a w a d z k i 2003) show distinct freshening o f the porewater w i t h depth i n the offshore region. A re la t ive ly diffuse area defines the m i x i n g zone i n 2002, bounded roughly between the onshore W P site and the B-nest data points . The sa l in i ty o f seawater col lec ted i n the water c o l u m n above site 1 was 31.5 ppt. T h e m u l t i l e v e l w e l l was sampled three t imes i n September 2002 and a fourth t ime i n September 2003 to determine i f salinit ies at depth had changed. T w o o f the s ix ports had ceased to funct ion proper ly and were not sampled. The 2003 samples had sal ini ty values that were w i t h i n l p p t o f the samples acquired one year earlier, ind ica t ing that the subsurface sa l in i ty prof i le had not changed at that locat ion. There appears to be a rather sharp freshwater/brackish water contact between the W P 1 prof i le and the A B - n e s t , a hor izonta l distance o f 12m. M o o r e (2003) recognized two i so top ica l ly distinct sources i n his r ad ium measurements o f the f ie ld area (and up to 2 8 k m offshore), l i k e l y ind ica t ing discharge f rom both the Sur f i c i a l A q u i f e r and the U p p e r F l o r i d a n Aqu i f e r . The idea o f two sources is d iscussed br ie f ly i n the next section. A s ment ioned i n the previous chapter, depths between 1 and 2 m b e l o w the seabed have sa l in i ty values s imi la r to that o f seawater, except those we l l s ve ry c lose to shore. It is apparent f rom the sa l in i ty profi les that complex interactions occur at depth w i t h i n the freshwater/brackish water m i x i n g zone, indicated b y lower sa l in i ty values at some nearshore, mid-depth we l l s . A d d i t i o n a l var iab i l i ty m a y be caused b y the fact that a l l we l l s have been projected onto a cross-section for presentation w h e n i n real i ty they are separated i n three dimensions . 27 3.2 Submarine Groundwater Discharge Exper iments 3.2.1 Tidal Measurements T i d a l osci l la t ions and waves cause pressure changes at the seabed, app ly ing or r emov ing pressure o n the fluids i n the sediment pores, in f luenc ing the rate o f submarine groundwater discharge. W a v e and w i n d act ion at the f ie ld site usua l ly fo l lows a set d a i l y pattern. The nearshore water surface is general ly c a l m i n the morn ing , w i t h w i n d increas ing steadily into the afternoon. W a v e act ion also increases, reaching a peak i n mid-af ternoon, usua l ly w i t h m i l d , choppy waves . C a l m water returns to the nearshore reg ion i n the late afternoon to evening as the w i n d and waves recede. The number o f major storms (>5cm o f ra infa l l per day) affecting the mar ine lab ranges between 2 and 6, annual ly . Seasonal ly , storms tend to occur i n late summer and early fa l l but they can occur at any t ime dur ing the year. N o major storms were recorded dur ing the experiment per iod but two signif icant events, described above, d i d occur . T i d a l osci l la t ions were measured us ing a submerged pressure transducer protected b y a 2 in . diameter, 2 m l o n g P V C pipe. The pipe was attached to a p i l i n g a long an aux i l i a ry w o o d e n dock, running para l le l to and immedia te ly nor th o f the m a i n concrete dock at F S U M L . The p i l i n g is approximately 3 5 0 m east o f w e l l A l . A photograph o f the apparatus is shown i n F i g u r e 3-11. Tape was appl ied at measured intervals a long the P V C pipe so that independent measurements o f t ida l e levat ion c o u l d be made and used to calibrate the sensor data. The transducer measured t idal fluctuations f rom A u g u s t 1 4 t h to September 7 t h , 2002, at a rate o f one measurement every 10 minutes un t i l A u g u s t 2 7 t h w h e n the rate was -28 increased to every 5 minutes. The sensor measures the height o f water above it instantaneously, occas iona l ly capturing "no ise" i n the fo rm o f anomalous ly h i g h or l o w water l eve l measurements. The P V C standpipe is not sensit ive to wave and w i n d act ion and noise is in t roduced f rom wakes or large waves . The data was corrected to N o r t h A m e r i c a n V e r t i c a l D a t u m 1988 ( N A V D 8 8 ) after the e levat ion o f the p i l i n g was determined f rom the survey conducted. The m a x i m u m t ida l range measured at the dock o f F S U M L over the pe r iod indicated above was 1.26m ( F i g u r e 3-12). The elevat ion ranged f rom - 0 . 4 7 m to 0 .79m, N A V D 8 8 , averaging 0 .21m. The t idal sequence over the experiment pe r iod began w i t h a spr ing tide (Augus t 14 t h ) , transitioned to neap tide (essentially equal tides o n A u g u s t 2 8 t h ) and back to spr ing tide again at the end (September 7 t h ) . The tides can be characterized as fo l lows : higher h i g h tides ( H H T ) ranged f rom 0 .472m to 0 .793m, l ower h i g h tides ( L H T ) ranged f rom 0 .298m to 0 .683m, higher l o w tides ( H L T ) ranged f rom -0 .169m to 0 .353m and lower l o w tides ( L L T ) ranged from -0 .466m to -0 .098m. T h e data presented i n this graph has been smoothed w i t h a 5-point running average. T h i s data is used as input for the numer ica l models and for statistical correlat ion analyses w i t h seepage and D P S data. 3.2.2 Seepage Measurements A n automated seepage meter system is preferred over a manua l seepage meter due to the increased eff ic iency i n data co l lec t ion i n temporal measurements and the ab i l i ty o f an automated system to cont inuously col lect data. W h i l e both systems are cha l leng ing to ins ta l l , manua l seepage meters have been noted for be ing dif f icul t to mon i to r and show significant a l ias ing effects i f bags are not pre- f i l led (Tan iguch i and F u k u o , 1993; C a b l e et 29 al . 1997). In l ight o f these drawbacks, automated seepage meters are b e c o m i n g more popular i n coastal zone studies. Character iza t ion o f seepage across the seabed was made us ing a Taniguchi - type automated seepage meter (Tan iguch i and F u k u o , 1993) and a differential p iezometer system ( D P S ) . The automated seepage meter employs a n ich rome w i r e heater and a series o f thermistors to determine the rate o f groundwater discharge through the seabed over a certain area. Water pass ing into the d rum from the sediments travels through a hose and past an electric heater. The temperature o f the water increases as it passes the heater. A series o f thermistors downstream o f the heater measure the temperature o f the water b y its ab i l i ty to conduct electr ici ty, w h i c h is output as a voltage. B a c k g r o u n d temperatures are r emoved w h e n the voltage measured at an upstream thermistor is subtracted f rom the downstream thermistors. The voltages are analyzed to back-out the seepage f l o w rate. Discharge rates as l o w as 1 x 10"7 m/sec (0.864 cm/day) can be detected b y the Taniguchi - type automated seepage meter ( M . Tan iguch i , 2004, personal communica t ion) . The seepage meter was set to col lect an instantaneous reading once every 5 minutes dur ing the t ime it was instal led. The data col lected can o n l y be examined after the system has been disconnected and removed f rom the seabed. T h i s can become a significant issue because problems w i t h the seepage data w i l l not be d i scovered un t i l the end o f the co l l ec t ion per iod. Th i s occurred dur ing the f inal run at site 2 and is discussed i n Sec t ion 3.3. T h e seepage meter was instal led four t imes at two locat ions w i t h i n the f ie ld area i n conjunct ion w i t h the differential piezometer system, w h i c h is descr ibed i n the next section. Site 1 is located approximate ly 2 5 m offshore o f the l o w tide l ine ( 4 m shoreward 30 o f BC-nes t ) and site 2 is located approximate ly 8 0 m offshore o f the l o w tide l ine ( 4 m seaward o f C-nest). T w o separate "runs" were conducted at each site, referred to here after as site 1, run 1 ( S 1 R 1 ) ; site 1, run 2 ( S 1 R 2 ) ; site 2, run 1 ( S 2 R 1 ) ; and site 2, run 2 (S2R2) . Tempora ry scaffolding was constructed at each site, p r o v i d i n g a p la t fo rm for equipment and where direct measurements o f seepage and differential pressure were made. Tan iguch i et a l . (2003) determined that submarine groundwater discharge rates var ied s ignif icant ly , bo th spat ial ly and temporal ly , at certain locat ions at the f ie ld site. The locat ions ment ioned above were chosen to help constrain this va r i ab i l i t y d iscovered i n the in tercomparison experiments o f 2000. Seepage output is presented i n F i g u r e s 3-13, 3-14 a n d 3-15. Three o f the four runs conducted at F S U M L had values that appeared to be realist ic but the f inal run at site 2 (S2R2) outputted values that were w e l l b e l o w the m i n i m u m detection l imi t for the seepage meter and are not presented. A drawback o f this system is that it can o n l y measure discharge out o f the seabed. The system is not current ly designed to capture both discharge and recharge as some seepage meters are (Paulsen, et a l . , 2003). The seepage meter was not cal ibrated immedia te ly p r io r to the 2002 study and thus, a reading o f zero seepage is not exact ly k n o w n nor is it k n o w n what voltage w i l l be output for recharge at the site, shou ld it occur. T h e ca l ibra t ion curve used to determine discharge rate was p r o v i d e d b y researchers at F l o r i d a State Un ive r s i ty , Department o f Oceanography. W h i l e recharge is u n l i k e l y and has not been measured at the site before, it is poss ible that it does occur. T h i s idea is explored i n Chapter 4. A diagram o f the seepage meter is s h o w n w i t h the differential piezometer system i n F i g u r e 2-5. 31 3.2.3 Differential Piezometer System The differential piezometer system ( D P S ) is designed to measure the ver t ica l difference i n pressure between two points i n the seabed. T w o piezometers ins ta l led paral le l to shore o n each side o f an automated seepage meter are used to evaluate h o w pressure changes i n the seabed relate to seepage and t idal fluctuations. E a c h piezometer is constructed o f three to four, 3 l c m - l o n g , 4cm-outside diameter, steel, female-threaded A W rods attached w i t h steel, male-threaded A W connectors. T w o ports i n each piezometer (one located at the t ip, another located i n an A W connector 0.3 l m or 0 .62m above the tip) connect to plast ic tubing w h i c h runs up the ins ide o f the A W rod, attaching to each side o f a differential pressure transducer. The differential pressure transducer is thus hydrau l i ca l ly connected to the points located i n the seabed. A l s o , as ment ioned prev ious ly and shown i n F i g u r e 2-4, the densi ty o f porewater was near ly equivalent to seawater to depths o f 1.5 to 2 m . I f waters o f different densities exist at the two ports, the equation presented i n Chapter 2(Ah = (AP/pg)+ A z ) w i l l not be v a l i d as it assumes constant density. A ver t ica l distance o f 0 .31m was chosen in i t i a l l y for port separation ( lower port at a depth o f 0 .7m be low the seabed) based o n w o r k done b y F indla te r (2001). H e r thesis presented cross-sections through t ime o f average freshwater heads measured i n the various we l l s i n the offshore region. H e r cross-sections showed that ver t ica l hydrau l ic head differences d i d not exceed 0 .8m between any o f the we l l s and were usua l ly w i t h i n the range o f 0.05 to 0 .3m over ver t ica l distances o f 1 to 3 m . B a s e d o n the fact that the piezometers were to be sha l low and to have a ver t ica l distance o f o n l y 0 .31m, two Setra M 2 3 0 , b i -d i rec t ional differential pressure transducers were acquired that are sensitive to 32 pressure fluctuations w i t h i n the +/- 0.5psi range (+/- 3.45 x 10 3 Pa) . T h i s equates to a range approximate ly equal to +/- 0 .34m o f head difference ( m a x i m u m measurable head difference w o u l d be 0.68m) and accurate to w i t h i n +/- 0 .0017m (or +/- 1.7mm). Thus , a pos i t ive reading o n the differential transducer w o u l d indicate discharge and a negative reading w o u l d indicate recharge. The in i t ia l port separation o f 0 .31m proved to be inadequate as measured output dur ing S 1 R 1 and S 2 R 1 indicated that recharge was occur r ing (negative differential heads), an observat ion not supported b y the automated seepage meter. A photograph o f the external components o f the sys tem can be found i n F i g u r e 3-16. Internal components are i l lustrated i n F i g u r e 3-17. S t i l l i n g tubes f i l l ed w i t h sand were used to m i n i m i z e noise caused b y w a v e act ion i n the water c o l u m n between the seabed and the water surface. F l e x i b l e plas t ic tubing (4in. diameter) was used to m i n i m i z e noise caused b y w i n d between the s t i l l i ng tubes at the water surface and the differential pressure transducers located i n a plas t ic b o x above the h igh tide surface. These tubes m a y not have been adequately s t i f f to dampen out the significant onshore w inds occur r ing i n the late afternoons at the f ie ld site as observed i n no i sy sections o f the data and discussed i n the next section. T h e b o x houses the electronic components o f the system, w h i c h includes the two Setra differential pressure transducers, one temperature probe to col lect temperatures inside the D P S box , one 1 2 V battery used to power the system and one C R - 1 0 X C a m p b e l l data logger used to cont ro l the transducers and the temperature probe and record measurements f rom these devices. The two piezometers are independent o f each other and col lect separate data sets. T h i s a l lows for compar i son between data sets and ver i f ica t ion o f data qual i ty . 33 Installat ion o f the differential piezometer system is an i n v o l v e d process and takes 5 to 6 hours o n average, outside o f prep-work. One fundamental preparat ion was the process o f m a k i n g de-aired water. Th i s process is important because d i s so lved atmospheric gases, w h i c h were present i n the tap water used for the D P S , tend to come out o f solut ion, fo rming bubbles on the inside o f the plast ic tubing. B u b b l e s act as t iny cushions w h e n subjected to pressure changes and are m u c h more compress ib le than the water that surrounds them. Th i s means that the accuracy o f the pressure transducer readings m a y be th rown o f f b y air bubbles trapped i n the tubing o f the D P S . De-a i red water is made b y b o i l i n g tap water for 15 minutes, w h i c h strips out the d i s so lved air, poured into glass jars, sealed and a l l owed to c o o l . T h i s water is then p u m p e d through the tubing o f the system and out the ports dur ing instal la t ion to keep the ports f rom getting c logged w i t h sediment. Dep loymen t began immedia te ly after h igh tide, m a x i m i z i n g the t ime spent w o r k i n g i n l o w water. Piezometers and the seepage meter are assembled onshore. The piezometers are d r iven into the seabed w i t h a s ledgehammer w i t h water be ing pumped through the system and out the ports. The rate o f p u m p i n g is important because i f p u m p i n g is too s low, the ports can c l o g w i t h sediment. I f the ports become c logged , the process is aborted un t i l they can be cleaned. U n l i k e the seepage meter setup, however , the D P S data can be downloaded pe r iod ica l ly dur ing the experiment. It should be noted that on one occas ion , h igh tides forced the r emova l o f the system before a fu l l t ida l cyc le o f 24 hours c o u l d be recorded for fear o f the D P S b o x be ing submerged i n seawater. T h i s event occurred at site 2, run 2, where the highest tides dur ing the f ie ld s tudy were 34 observed. The combina t ion o f the h igh tide and 0 .25m waves led to the r e m o v a l o f the sensors and recording equipment at lower h igh tide (12noon) on September 7 t h . A t the end o f each D P S run, independent measurements o f differential head were made b y observ ing the water levels i n each tube after they had been disconnected f rom the D P S box . In the case o f S 2 R 2 , the measurements were made p r io r to the run, o n September 6 t h . The measurements were made as c lose to shutdown as poss ible so that transducer measurements c o u l d be compared w i t h the independent measurements. B o t h tubes f rom each piezometer were pu l l ed taut out o f the water and a l l o w e d to equil ibrate for a few minutes before be ing he ld together against a ruler. The height o f the water l eve l i n each tube was then measured relative to the sea surface and these numbers subtracted to give the head differential . I f waves were present dur ing the measurement, the sea surface was estimated and the heights o f the water inside the tubes measured against this temporary datum. T h i s data is presented i n Table 3-2. M u l t i p l e measurements were made o f the observed head difference i n the tubing and thus an averaged va lue was used. T h i s was compared to the last 30 seconds o f data from the transducers, w h i c h was averaged. Table 3-2. Independent Differential Head Measurements Site R u n Sensor Ave rage Observed Dif ferent ia l P iezometer Sys tem Output H e a d Difference (cm) (cm) (last 30 seconds averaged) 1 1 1 0.9 -0.1 1 1 2 0.7 0.5 2 1 1 0.9 0.8 2 1 2 1 - 1 . 0 ' 0.9 1 2 1 3.2 3.1 1 2 2 5.3 5.3 2 2 1 N o t avai lable 20.2 2 2 2 3 6.3 1.6 'Sensor 2 tubing was observed to be constricted, may have been clogged. 2 May have observed incorrect tube. 3Sensor 2 tubing was observed to be slightly constricted, may have been clogged. 35 T h i s data indicates that there is good agreement between measured and outputted differential head data. T h i s is problematic , as discussed be low, because large errors are associated w i t h the cal ibrat ions o f the differential pressure sensors. Despi te these errors, the observed differential heads i n T a b l e 3-2 show that i f the ca l ibra t ion was performed more careful ly, perhaps the output o f the D P S w o u l d be usable for analysis . P r i o r to f ie ld work , and again after returning f rom F l o r i d a , ca l ib ra t ion experiments were conducted w i t h the differential pressure transducers us ing a standing manometer capable o f p roduc ing head differentials o f up to l m . De-a i r ed water was pumped into a man i fo ld and through the pressure transducers. B l e e d va lves were used to ensure that no air bubbles w o u l d be present, w h i c h might act as cushions to any pressure fluctuations, reduc ing the data accuracy. Once the manometer tubes were f i l l ed , a syringe was used to add or remove water f rom the co lumns . The differential heads were var ied to obtain the widest possible range o f output i n m i l l i v o l t s f rom the pressure transducers. The relationships determined f rom these experiments were used to produce ca l ibra t ion equations to process the raw D P S data. 3.3 Resul ts and D i s c u s s i o n 3.3.1 Nearshore Aquifer Characterization Onshore Horizontal Hydraulic Gradient In order to estimate groundwater f l ow to the nearshore mar ine environment , an onshore hor izonta l hydrau l ic gradient is calculated. A hydrau l ic gradient is calculated is based on the Leve logge r data col lected at we l l s P I and N I as descr ibed above. T h e hor izonta l distance between P I and N I is 56 .1m. The fluctuations caused b y the t idal 36 signals i n each w e l l d i d not s ignif icant ly affect the hydrau l ic gradient ca lcu la t ion because the elevat ion difference remained near ly constant dur ing these t imes, ranging on average between 4 and 5cm. In that the s torm events o n the 2 8 t h and 3 0 t h o f A u g u s t created two periods o f steady water table rise and decl ine, each per iod was examined separately. The first per iod w i l l l i k e l y produce a more accurate portrayal o f the average hydrau l i c gradient because it is over a longer t ime interval al though both the first and second periods were used to calculate a gradient. F o r the first per iod, the ver t ica l water table difference between we l l s P I and N I ranged between 1.36 and 1.40m, averaging 1.38m. A n average gradient o f 0.025 was determined. The second per iod produced a hydrau l ic gradient ve ry close to that o f the first, determined to be 0.026. W h i l e the Leve loggers are actual ly recording the pressure head i n the onshore we l l s , the e levat ion recorded is used as a p r o x y for the water table elevation. A n interesting observat ion ment ioned p rev ious ly is that the A u g u s t 2 8 t h event caused a 6 c m rise i n the water table at w e l l P I w h i l e o n l y a l c m rise is noted i n w e l l N I . T h i s observation is further discussed later i n the chapter. It should be noted that independent measurements o f the water l eve l i n we l l s N I and P I placed the water table 0.5 to 3 .7cm higher i n w e l l N I and 4.9 to 5 .1cm higher i n w e l l P I than measured w i t h the Leveloggers . One possible explanat ion is that the co i l ed wi re used to suspend the Leveloggers i n the we l l s had not been straightened enough pr io r to insert ion and that the slight he l i ca l shape o f the w i r e produced f r ic t ion against the inside o f the we l l s . The Leveloggers w o u l d thus not extend to their expected fu l l depth. The Leve loggers were r emoved and reinserted two to three t imes over the course o f the f ie ld study (for s lug testing, water sampling) and each reinsert ion w o u l d produce a poss ib i l i ty for error. A n independent measure was made at each r e m o v a l , g i v i n g the 37 opportuni ty to correct for such errors. I f a m a x i m u m difference between the w e l l elevations for each per iod is used to calculate the hydraul ic gradient, the same gradients reported above are produced. Thus , the difference between the independent measurements and the sensor measured data does not affect the calculated hydrau l ic gradient o f 0.025. Ave rage water table depth be low the ground surface is -2 .086m for w e l l P I and -0 .934m for w e l l N I . Slug Test Results S l u g testing o f we l l s N I , P I and P 2 produced hydrau l ic conduc t iv i ty estimates o f 1 .02xl0" 4 m/s, 9 x l 0 " 6 m/s and 2 . 7 9 x l 0 " 4 m/s, respectively, suggesting m e d i u m grained sand. These conduct ivi t ies seem reasonable and are s imi la r to those found i n other onshore wel l s f rom previous studies (Findlatter, 2001). A compar i son is d i sp layed o n Table 3-3 (adapted f rom Findlatter, 2001). 38 T a b l e 3-3. H y d r a u l i c C o n d u c t i v i t y o f W e l l s at F S U M L M o n i t o r i n g W e l l M i d - s c r e e n depth E l e v a t i o n o f m i d - H y d r a u l i c b e l o w sediment screen (m, C o n d u c t i v i t y (m/s) interface (m) N A V D 8 8 ) P I 4.62 -0.75 9 . 0 x l 0 " b P 2 2.79 1.16 2 . 8 x l 0 " 4 N I 3.86 -0.245 l . O x l O " 4 A l 3.13 a -2.61 3 . 1 x l 0 " b A 2 2.87 - 9 . 1 x l 0 " b b A 3 1.73 - V e r y permeable c A B 1 3.63 - 2 . 2 x l 0 " 4 A B 3 1.28 - 1.4x l0" 4 B l 4.29 - 5 . 8 x l 0 " b B 4 1.07 - 2.0x10" ' B 5 3.67 - 3 . 5 x l 0 " b B 6 2.78 - 5 . 1 x l 0 " b B C 1 3.87 - 1.1x10"' B C 4 1.23 - 1.4X10"5 C I 3.75 - 2.3x10"" C 3 1.38 - 5 . 6 x l 0 " b D I 4.38 - 4.9x10" 5 b D 3 1.32 - 1.9x l0" b a The mid-screen depth has changed since reported in Rasmussen, 1998 (initially recorded as 3.93m). This is likely due in part by infilling of sediment after the standpipe was opened at ground level during a fire. b For wells where more than 1 test was performed, the arithmetic mean of the K values was reported (see Findlater, 2001). c The rate of recovery during the testing was too rapid to be accurately measured. A significant decrease i n hydraul ic conduct iv i ty w i t h depth is noted between we l l s P 2 and P I , the value o f K decreasing b y a factor o f 31 . T h i s imp l i e s a change i n ver t ical structure between these two wel l s . T h i s l oca l ver t ica l difference is not consistent throughout the f ie ld area. Findlater (2001) recognized that the rest o f the w e l l s at the f ie ld site do not have an obv ious trend o f hydraul ic conduc t iv i ty values w i t h depth. Despi te this, S m i t h and Z a w a d z k i (2003) adopted a ver t ica l layer m o d e l to see i f that w o u l d exp la in the rate o f S G D observed at the site. H y d r a u l i c conduct iv i t ies determined f rom the 2002 s lug tests established that values were w i t h i n one order o f magni tude o f those found b y Rasmussen (1998) and Findla ter (2001). 39 Approximate Discharge Rate at the Shoreline F r o m the onshore hydrau l ic gradient estimate and s lug test data o f onshore we l l s , an approximate discharge rate at the shoreline can be determined. A n ari thmetic mean o f the hydrau l ic conduct ivi t ies calculated at we l l s P I , P 2 and N I produced an average conduct iv i ty o f 1.3xl0~ 4 m/s. A geometric mean gives an average hydrau l i c conduc t iv i ty value o f 6 .4x l0~ 5 m/s. B o t h are used to calculate an approximate range o f submarine groundwater discharge values expected at the site. The unit length o f shorel ine (below) was chosen to match the w i d t h o f the f ie ld area used b y the 2000 in tercompar ison study researchers. T h i s area extends 2 0 0 m offshore o f the mean tide l ine . The result o f this ca lcula t ion w i l l then be comparable to their summary results, pub l i shed i n Burnet t et a l . (2002) and the results o f the models produced b y S m i t h and Z a w a d z k i (2003) whose numer ica l m o d e l was based o n this study area. The f o l l o w i n g values were used i n the calcula t ion: H y d r a u l i c C o n d u c t i v i t y (K m e a n): 1 .3xl0" 4 m/s = 7 . 8 x l 0 " 3 m / m i n H y d r a u l i c C o n d u c t i v i t y (Kgeomean): 6.4x10" m/s = 3.8x10" m / m m H y d r a u l i c Gradient [dhjdx): 0.025 U n i t Leng th o f Shorel ine: 100m Thickness o f Sur f i c ia l Aqu i f e r : 7 m U s i n g these variables, total discharge rate (Q) at the site is estimated to be: Q = Kn,eanA— = (7-%xl0"3m/min)(0.025Xl00m\7m) = 0 . 1 4 m 3 /min dl Q = K S e o m e a n A — = ( 3 - 8 * 1 0 " 3 m 1 min)(0.025XlQ0m\lm) = 0.07OT 3 /min 40 These results indicate that another source o f freshwater is required to main ta in the S G D rates measured b y Burnet t et a l . (2002). T h e y report discharge rates r ang ing f rom 1.6 to 2.5 m 3 / m i n over the same region indicated above based o n measurements f rom chemica l tracers and manua l and automated seepage meters. Al te rna t ive ly , S m i t h and Z a w a d z k i (2003) predicted seepage rates between 0.005 and 0.15 m / m i n f rom their first m o d e l , w h i c h incorporated input o n l y from the Sur f i c ia l Aqu i f e r . T h e i r h igher value o f 0.15 m 3 / m i n , is based o n a mode led ver t ica l structure where hydrau l ic conduc t iv i ty decreases w i t h depth. The second ca lcula t ion above (wi th K g e o m e a n ) produces an estimated discharge rate s imi la r to their m o d e l . It seems l i k e l y then, that leakage across the Intracoastal formation, as proposed b y M o o r e (2003) and modeled b y S m i t h and Z a w a d z k i (2003), is one potential source for addi t ional freshwater input to the system. The hydrau l ic conduct iv i ty values determined for the onshore w e l l s are not necessari ly representative o f the rest o f the onshore region. The hydrau l i c conduc t iv i ty varies b y two orders o f magnitude between we l l s P I (9x10~ 6 m/s) and P 2 ( 2 . 8 x l 0 " 4 m/s) over a ver t ica l distance o f 1.9 m . A d d i t i o n a l l y , the wel l -screen o f w e l l A l has an elevat ion ~2 m lower than w e l l P I ( F i g u r e 3-1) and a s imi la r hydrau l ic conduc t iv i ty Q . l x l O " 6 m/s). W e l l N I is re la t ive ly deep, screened at a point 0.5 m higher than P I and w i t h a hydraul ic conduct iv i ty estimated at 1 .02xl0" 4 m/s, w h i c h is c loser to that o f w e l l P 2 . I f w e assume that we l l s A l and P I are part o f the same sediment unit , a s lop ing "contact" might run between the two we l l s . Unfortunately, it is u n k n o w n whether the heterogeneity o f the onshore Sur f i c i a l A q u i f e r is la teral ly extensive or i f there are discont inuous pockets o f h igh and l o w conduct ivi ty . The t ida l s ignal analysis section 41 explains one hypothesis for w h y the site m a y have zones o f h i g h and l o w hydrau l ic conduct iv i ty . W h i l e the offshore Sur f i c i a l A q u i f e r we l l s also range i n hydrau l i c conduct iv i ty , no apparent ver t ical structure w i t h depth was found b y Findla ter (2001). S m i t h and Z a w a d z k i (2003) do suggest a structure i n their numer ica l m o d e l , as ment ioned previous ly . A g a i n , the K g e omean ca lcula t ion above matches w e l l w i t h their s imula ted discharge across the seabed f rom the Sur f i c ia l Aqu i f e r . T h i s lends credence to the hypothesis o f ver t ica l structure i n the onshore region. Tidal signal analysis A s described prev ious ly , t ida l signals are present i n both w e l l P I and w e l l N I . The s ignal at P I is not iceably stronger (signal ampli tude o f 1 to 1.5 cm) than at N I (signal ampli tude o f 0.75 to 1 cm) despite P I be ing farther onshore. W e l l data f rom P I and N I are presented w i t h t idal e levat ion data i n F i g u r e s 3-18 a n d 3-19. T h e storm events inhib i t a continuous assessment o f the relat ionship but there are enough clear signals i n the w e l l data before and after the ra in events to explore h o w the water table onshore is inf luenced b y the tides. The ra in events occur dur ing a per iod o f transi t ion between neap and spr ing tides. Semi-d iurna l tides are di f f icul t to ident ify i n the w e l l data but 12.5 hour signals are present i n the data, as w i l l be discussed i n the P o w e r Spect rum Dens i ty section o f this chapter. A n analyt ica l so lu t ion is attempted to predict the t ida l ampl i tude damping expected at we l l s P I and N I . The equation,H(x) = h0 exp(-x^nS s j t p T^j , calculates the 42 ampli tude o f the t ida l s ignal (H(x)) at a w e l l x meters f rom the mean tide l ine , w i t h parameters o f the t ida l ampl i tude (h0), specif ic storage (Ss), t ida l pe r iod (tp) and aquifer t ransmiss ivi ty ( I ) , w h i c h is Kb or hydraul ic conduct iv i ty m u l t i p l i e d b y the aquifer thickness ( Y i m and M o h s e n 1992). T h i s t idal analysis is a transient boundary value p rob lem equation that assumes a semi-infini te aquifer, w h i c h is perfect ly confined. The Sur f i c ia l A q u i f e r is not conf ined and thus, this assumption is v io la ted . A correct ion is made, however , i n that the specific y i e l d (Sy) is used instead o f the specif ic storage (Ss) because Sy is defined for calculat ions i n v o l v i n g an unconfined aquifer. T h e f o l l o w i n g values are used i n the equation: Dis tance f rom Shorel ine, P I (x): 92 m Dis tance f rom Shorel ine, N I (x): 36 m T i d a l ampl i tude (h0): 0.5 m Speci f ic Y i e l d (Sy): 0.20 T i d a l pe r iod (tp): 12.67 hours = 45612 seconds Transmiss iv i ty , P I (T=Kb): ( 9 x l 0 " 6 m/s)(7m) = 6 . 3 x l 0 " 5 m 2 / s Transmiss iv i ty , N I (T=Kb): ( l x l O " 4 m/s)(7m) = 9 . l x l O " 4 m 2 / s W i t h these parameters, t ida l s ignal amplitudes o f l x l O " 1 7 c m and 0 .6cm are estimated for we l l s P I and N I , respectively. The value calculated for w e l l P I is essential ly equal to zero, an est imation that is not reflected i n the observed w e l l data where ampli tudes o f up to 1.5 c m are observed. The value calculated at w e l l N I is approximate ly correct, be ing very close to the observed 0.75 c m amplitudes. A s an experiment, the hydrau l i c conduc t iv i ty i n the equat ion was changed to see what value w o u l d produce a t ida l s ignal ampli tude equivalent to that seen i n the field. A t ida l s ignal o f 1.5 c m observed at a w e l l 43 92 m in land f rom the shoreline w o u l d require an aquifer hydrau l ic conduc t iv i ty o f 1.3xl0~ 3 m/s, a value 144 times higher than the value estimated at w e l l P I . A conduc t iv i ty o f 1.4x10" 4 m/s produces the ampli tude that is seen i n the f ie ld at w e l l N I , ampl i tude o f approximate ly 0 .75cm. T h i s value for N I is approximate ly the same as the estimated value. It is not entirely clear w h y the t idal signals at N I are damped relat ive to P I . T i d a l signals are not as evident dur ing the water table decl ine at N I pr ior to the s torm events yet the signals are ve ry evident i n the data after the s torm events, perhaps i n response to the large spr ing tide observed September 4 t h to 7 t h . A s ment ioned above, the hydrau l ic conduct iv i ty o f the sediment immedia te ly surrounding w e l l N I is ~31 t imes higher than that o f the sediment surrounding w e l l P I ( T a b l e 3-3), i m p l y i n g that greater attenuation, and thus smaller signals, should be seen at P I , as reflected i n the calculat ions . T h i s , however , does not appear to be the case at the f ie ld site. One poss ible explanat ion is that w e l l N I is immedia te ly shoreward o f a l o w conduc t iv i ty unit, w h i c h w o u l d dampen pressure signals f rom the tides at the shoreline. It is apparent f rom T a b l e 3-3 that we l l s A l and A 2 , the deepest we l l s i n the A-nest , have conduct ivi t ies i n the range o f 10" 6 m/s. Th i s represents a l o w hydrau l ic conduct iv i ty zone between the shorel ine and w e l l N I that m a y act to inhib i t t ida l influences. A d d i n g weight to this argument is the fact that w h e n the first large storm ( A u g 28 t h ) infil trated into the Sur f i c i a l A q u i f e r , it caused a notable increase i n the water table i n w e l l P I (6 cm) w h i l e o n l y a m i n o r increase was noted i n w e l l N I (2 cm) , despite w e l l N I be ing 31 t imes more conduct ive than P I ( F i g u r e s 3-4 a n d 3-5). T h i s impl i e s that w e l l N I is near some k i n d o f l o w hydrau l ic conduc t iv i ty unit. It is also poss ible that some k i n d o f h igh conduct iv i ty unit exists at depth, t ransmitt ing the 44 t idal s ignal deep onshore w h i l e l eav ing hydraul ica l ly- insula ted , nearshore w e l l s l i ke w e l l N I mos t ly unaffected. T h i s hypothesis is u n l i k e l y and w o u l d require further f i e ldwork before be ing rejected. The best assumption that can be made is that w e l l N I is adjacent to a l o w conduct iv i ty unit that cou ld be part o f the A-nes t l o w K zone. 3.3.2 Submarine Groundwater Discharge A s ment ioned previous ly , two sites i n the offshore reg ion were chosen for discharge experiments; site 1 located 2 6 m offshore f rom the l o w tide l ine and site 2 located 7 8 m offshore f rom the l o w tide l ine . The results o f the automated seepage meter and the differential piezometer system are presented separately and then examined for dominant frequencies v i a power spectrum density analyses, discussed later i n the chapter. Direct Seepage Measurement Results The rate o f submarine groundwater discharge at the F S U M L f ie ld site var ied extensively over the course o f the f ie ld examinat ion. T i d a l signals are apparent i n the seepage data both v i s u a l l y and i n the power spectrum analyses. Sta t i s t ica l ly significant relationships also exist between some o f the seepage data and the sea surface e levat ion al though most relat ionships are quite weak. The f o l l o w i n g d i scuss ion is b roken d o w n b y ind iv idua l runs at each site. The data set col lected dur ing run 1 o f site 1 (S1R1) is cont inuous between Augus t 16 t h and A u g u s t 2 3 r d and displays discharge rates i n the range o f 10 to 80 cm/day, steadily increasing w i t h t ime (Figure 3-13). It is not understood w h y the discharge w o u l d be 45 increasing over this interval , namely because no significant ra infa l l events occurred that might increase overa l l seepage rates. E a r l y seepage results do not exhibi t strong t ida l influences un t i l approximate ly A u g u s t 19 t h . Da ta col lected between A u g u s t 1 7 t h and Augus t 2 3 r d is plotted i n Figure 3-20. F r o m v i sua l examinat ion o f the data, peak discharge occurs after the h i g h tide has passed but before the l o w tide is reached. T h i s observation is explored and reproduced w i t h the numer ica l m o d e l i n Chapter 4. Paulsen et a l . (2003) found a strong inverse relationship at Wes t N e c k B a y , L o n g Island, N e w Y o r k , b y di rec t ly p lo t t ing t ida l stage against discharge. T h i s behavior seems to be site specific, however , as an inverse, l inear relat ionship is not observed i n the data f rom F S U M L . Groundwater discharge dur ing the second run at site 1 ( S 1 R 2 ) , col lec ted between September 3 r d and September 5 t h , is diff icul t to characterize (Figure 3-15). In i t ia l rates dur ing the first 24 hours are 10 to 20 cm/day (wi th a few higher spikes) but taper o f f to less than 10 cm/day. P r i o r to 18:20 on the 4 t h , the data has one recognizable peak early on, corresponding to a drop i n the tide (Figure 3-21). A t 18:20 o n September 4 t h , the meter started recording seepage rates 3 times higher than before. D u r i n g the second h a l f (after 18:20, Sep 4 t h ) , there are two peaks that correspond to d ropp ing or l o w tides but v i s ib le correlat ion is diff icul t . It is not clear w h y there is a sudden increase i n the discharge rate part w a y through the run. Tide/seepage interactions dur ing run 1 at site 2 ( S 2 R 1 ) are easier to explore than other data sets because the interval o f co l lec t ion was over a re la t ive ly l o n g per iod (8 days). The seepage data (Figure 3-14) has obvious t ida l influences f rom the beg inn ing o f the run, A u g u s t 2 4 t h un t i l about Augus t 3 1 s t , w i t h the seepage rate ranging i n magnitude 46 f rom 6 to 49 cm/day. F r o m this point un t i l the end o f the run o n September 2na, the seepage rate dropped and remained steady at about 15 cm/day. It shou ld be noted that this drop i n seepage coinc ides w i t h the two major s torm events o n A u g u s t 2 8 t h and 3 0 t h ( F i g u r e 3-22). It is not clear w h y seepage w o u l d drop o f f immed ia t e ly after these ra in events. It is l og i ca l to assume that the opposite w o u l d occur, w i t h discharge increasing after ra in events o f this magnitude. F r o m the f ie ld observations, it can be seen that seepage fo l lows a c y c l i c pattern that fo l lows the t idal cyc le . A t the t idal extremes, discharge rates are at a m i n i m u m and near zero. D u r i n g the t ransi t ion o f h i g h to l o w tide, seepage rate reaches a m a x i m u m . Converse ly , w h e n the tide is t ransi t ioning f rom l o w to h igh tide, seepage is at a m i n i m u m . Hence , w i t h o n l y discharge occur r ing (no recharge), a plot o f seepage rate against t ida l height should give a non-l inear relat ionship that is convex d o w n , highest seepage at mid-t ides, and l o w seepage at the extremes. T h i s process is i l lustrated extremely w e l l i n the I D m o d e l results o f Chapter 4. The seepage data col lec ted dur ing second run at site 2 ( S 2 R 2 ) was not used i n an analysis because the output produced is b e l o w the detection l i m i t o f the meter (1 cm/day) . T h i s si tuation highl ights one o f the design problems w i t h the Tan iguch i - type automated seepage meters. The seepage measurements can be examined o n l y after the system has been removed f rom the seabed, w h i c h i n this case meant the loss o f the entire data set. I f there was a w a y to remotely check the output wi thout d isassembl ing the apparatus, these problems c o u l d be avoided . Notab le i n F i g u r e s 3-14 a n d 3-15 are extended periods o f near constant discharge rates o f 10 to 20cm/day that extend f rom 10 to 72 hours at a t ime. D i scha rge is recorded 47 b y the automated seepage meter but these periods do not d i sp lay the effects o f t ida l fluctuations. T h i s is unusual and it is not k n o w n w h y this behavior is observed. In general, the seepage data sets col lected for this project are s imi l a r to the seepage data col lected for the 2000 intercomparison project. The m a i n s imi l a r i t y between the seepage results is that the range o f discharge rates (averaging about 20-40 cm/day) is s imi la r and the temporal variat ions i n the data w i t h t idal fluctuations is s imi la r . In this regard, the discharge is observed to be greatest dur ing the t ransi t ion between h i g h and l o w tides. Seepage and Rainfall A n invest igat ion o f seepage response to storm events was conducted o n the seepage data. F i v e significant ra infa l l events occurred over the course o f the experiment. A significant ra infa l l event was considered to be any storm that p roduced 8 m m / h r o f ra infa l l w i t h i n a 15 minute per iod or more over the f ie ld area. A peak discharge i n the seepage meter data after the two ra in fa l l events o f Augus t 2 8 t h and Augus t 3 0 t h was expected but a significant increase i n discharge was not observed. O n the contrary, seepage dropped to less than 20cm/day and stayed re la t ive ly th constant after the A u g . 30 event. In the Leve logge r data, these events were captured as a rap id and significant rise (0 .28m over 3.5 days) i n the water table at w e l l P I and a s l ight ly more gradual and moderate rise (0 .16m over 4 days) i n w e l l N I . T h e heightened water table at w e l l P I reached its peak on A u g u s t 3 l s t /Sep tember 1 s t . It is not clear w h y offshore seepage rates w o u l d decl ine after such significant inf i l t ra t ion has occurred. 48 O n A u g u s t 1 9 t h and 2 2 n d , dur ing run 1 at site 1, smaller events occurred (8 .6mm/hr and 13mm/hr) than those l is ted above yet an increased seepage rate is notable i n the seepage meter output for both events ( F i g u r e 3-20). The peaks i n seepage are caused b y the t idal influences but m a y be enhanced b y the storms. Differential Piezometer System The Dif ferent ia l P iezometer System d i d not successful ly resolve pressure differences i n the seabed dur ing the per iod o f observat ion at F S U M L . T h e reason for the d i f f icul ty arises from two k e y issues: the ver t ical distance (Az) between the two ports on the piezometers and the cal ibrat ions used to process the data. F o u r ca l ibra t ion tests were conducted on the differential pressure transducers under laboratory condi t ions , two before the f ie ldwork and two after the f ie ldwork . A standing manometer was used to vary the heads, the output plot ted and a regression l ine fit to each l ine . The cal ibrat ions were not designed to m i m i c the environment o f the F l o r i d a f ie ld area but were intended to measure the output o f the pressure sensors i n a control led setting. It became apparent after running the pos t - f ie ldwork cal ibrat ions o n the D P S sensors that the output had "dr i f ted," shift ing a l l values for sensor 1 approximate ly 2 c m to the left and a l l values for sensor 2 approximate ly 1cm to the left ( F i g u r e s 3-23 a n d 3-24). It is not k n o w n w h e n the sensors drifted, because no other cal ibrat ions had been performed i n the f ie ld . A s a result o f the lack o f in format ion o n the drift (i.e. w h e n or h o w it occurred), investigations were made into h o w to use the ca l ibra t ion error and m i n i m i z e the error associated w i t h the drift. In order to use the differential pressure data, a regression analysis was performed on the a l l ca l ibra t ion data f rom both before and after 49 the f ie ldwork. A s part o f the regression analysis, an R M S error is p roduced w h i c h describes h o w far away a g iven point falls f rom the regression l ine . T h e equations produced f rom the a l l ca l ibra t ion data are: F o r Sensor 1: Ah = (AP- 1616.1) / 34.62 where the rms error is Ah ± 1.05cm F o r Sensor 2: Ah = (AP - 1418.2) / 35.39 where the rms error is Ah ± 0 .68cm In these equations, AP is the sensor output (differential pressure) i n m V and Ah is the differential head. A t a 6 8 % confidence interval , the R M S error is ± 1.05cm for sensor 1 and ± 0 .68cm for sensor 2. A t a 9 5 % confidence interval , these errors become ± 2 .1cm and ± 1.36cm, respect ively. Because the errors i n v o l v e d have the same magni tude as most o f the data, the D P S output cannot be trusted to resolve processes affecting the seabed. Despi te this, the regression equations produced for each sensor were used to process the r a w data f rom the differential p iezometer system so that they c o u l d be plotted. Figure 3-24 displays sensor data processed w i t h both a l l ca l ibra t ion data (thin lines) and o n l y post-f ie ld w o r k cal ibra t ion data (thicker l ines). The cal ibrat ions performed before and after the f ie ld study were conducted so that a l l procedures were the same, as m u c h as possible. Y e t the cal ibrat ions are o b v i o u s l y o f f b y a considerable amount. Potent ia l errors i n the cal ibra t ion tests c o u l d have been caused b y mul t ip le factors. The process used to de-air the water was changed after conduct ing pre-f ield cal ibrat ions. In i t ia l ly , a v a c u u m o f 8 5 k P a was appl ied to a v o l u m e o f water for 24 hours. F o r the p re l imina ry f ie ld studies i n Vancouve r , the actual f i e ldwork i n F l o r i d a and the post-calibrations, bo i l ed water was used to de-air water for the sensors. Ano the r potential source o f error is the density o f water used i n D P S cal ibrat ions m a y have been different f rom the density o f water used i n D P S experiments i n F l o r i d a . V a n c o u v e r tap 50 water (used for a l l ca l ibra t ion experiments, bo th pre and post f i e ldwork) has a l ower T D S concentration than that o f tap water at F l o r i d a State U n i v e r s i t y M a r i n e L a b . T h e density differential o f the water used is l i k e l y to be extremely sma l l and w i l l not have m u c h o f an effect o n the calibrat ions. A greater chance o f error that m a y have caused the drift i n the sensors is the severe temperature fluctuations experienced inside the plast ic b o x where the sensors were housed dur ing experimentat ion i n F l o r i d a . The box underwent a m a x i m u m da i ly s w i n g o f 2 2 ° C dur ing run 2 at site 1, w i t h the temperature ranging f rom 2 5 ° C just pr ior to sunrise to 4 7 ° C i n the late afternoon (See F i g u r e 3-26). Pressure transducers can be sensitive to temperature fluctuations and m a y cause offset i n the data. The temperature fluctuations inside the D P S box dur ing the experiments are w e l l w i t h i n the operating temperature o f the sensor, w h i c h is -18 to + 8 0 ° C . There is a potential as w e l l that the resistors used i n the D P S box for t ransforming the e lect r ical output f rom the sensors were not proper ly insulated against the h i g h temperatures ins ide the box . A s the resistors were not tested for temperature effects, it is u n k n o w n whether this affected the D P S output. In l ight o f these errors, it is s t i l l useful to examine the differential p iezometer system output. A s discussed previous ly , the differential pressure transducers used i n the f ie ld are capable o f r e so lv ing pressure differences o f up to +/-0.34m o f hydrau l ic head (+/-0.5 ps id or +/- 3.45 Pa) , w h i c h is w i t h i n the range o f head differences measured at the f ie ld area (Findlater 2001) . P r e l im ina ry experiments conducted at a beach a long the Spanish B a n k s o f V a n c o u v e r pr ior to the F l o r i d a f ie ld w o r k indicated that the port separation o f 0 .31m was sufficient to produce head differences o n the order o f 5 to 11cm for sensor 1 and 1 to 6 c m for sensor 2. It is not clear w h y the sensors p roduced different 51 readings. It was not anticipated that the head differences measured at F S U M L w o u l d require a larger port separation. Recharge was observed b y both sensors (more c o m m o n i n sensor 2), a h i g h l y u n l i k e l y occurrence w h e n constant discharge was recorded b y the seepage meter 0 .4m away. Dif ferent ia l head measurements were sampled at a rate o f one measurement every 2 seconds in i t i a l l y so as to ensure that no signals affecting the D P S w o u l d be lost. It was determined, however , that one measurement every 6 seconds w o u l d be suitable and easier for data storage. Based on the cal ibra t ion equations (discussed be low) , sensor 1 exhibi ted head differences for run 1 that ranged f rom -1.5 c m to +4cm (average o f 1cm) at site 1 (Figure 3-27) and a range o f - 2 c m to +4 .5cm (average o f 1.5cm) at site 2 (Figure 3-28). Sensor 2 had output ranging f rom - 1 c m to +4cm (average 0.5cm) at site 1 (Figure 3-29) and f rom -3.5cm to +2cm (average 0cm) at site 2 (Figure 3-30). It should be noted that significant noise is present i n the data, l i k e l y caused b y w i n d and/or wave disturbances. Future experiments w i l l require more v ig i l an t efforts to reduce noise i n the system. The port separation was increased to 0 .62m for run 2, p roduc ing better results but also in t roducing new concerns about the data. D u r i n g run 2 at site 1, the head differences for both sensors were somewhat higher than those observed dur ing run 1 and no negative readings were recorded (Figure 3-31). Notab le , however , is that the sensor outputs are not just offset i n magnitude as i n the first run, they are also offset i n di rect ion. Whereas p rev ious ly the sensors produced differential head measurements that osc i l la ted i n the same direct ion, S 1 R 2 shows the sensors f luctuating i n different direct ions. That the head difference recorded at each piezometer (measured at the same depths and separated b y 52 o n l y 1.4m, hor izonta l ly) w o u l d record opposite osci l la t ions w i t h t ide is a surpr is ing result and ve ry dif f icul t to interpret. Included i n Figure 3-31 is the seepage discharge data. T h i s data has been inc luded because the o r ig ina l intent o f the D P S experiment was to correlate a l l three data sets: t idal fluctuations, seepage rate and differential head i n the seabed. A s the D P S cal ibra t ion errors d i d not a l l ow the data to be resolved i n a usable format, this goa l c o u l d not be achieved. Despi te this, S 1 R 2 contains the highest qua l i ty D P S data co l lec ted at the f ie ld site. R u n 2 at site 2 recorded differential head for less than one t ida l c y c l e and thus had s imi la r issues as w i t h the seepage data. Trunca t ion o f the experiment was mandatory due to the large spr ing tide occur r ing at that t ime, forc ing the r e m o v a l o f the D P S datalogger and sensors. A s w i l l be described i n the next section, no realist ic analysis can be performed due to ca l ibra t ion issues. In summary, the differential pressure experiment was not successful because ( A ) run 1 had to be discounted f rom analysis due to l o w head differences and observed recharge, (B ) the R M S error f rom the regression o f the ca l ibra t ion data is at the same magnitude as the data i t se l f and (C) the cal ibra t ion o f the differential pressure sensors d i d not take into account environmental affects l i k e l y to be experienced i n the offshore reg ion o f the F S U M L . Future studies should inc lude careful ca l ibra t ion o f equipment under condi t ions l i k e l y to be experienced i n the f ie ld . 53 Power Spectrum Density Analyses P o w e r Spect rum Dens i t y ( P S D ) transformations determine the dominant frequencies present i n a g iven set o f numbers. A s s u m i n g that t ida l influences are present i n both the seepage meter data and the onshore Leve logge r data f rom we l l s P I and N I , these data sets were examined. The numer ica l analysis too l , M a t l a b ™ , was used to process the data for these experiments. The w i n d o w size for each analysis was defined as the number o f samples i n the data set, a l l o w i n g for the highest resolu t ion possible . A s a result the figures p rov ided tend to be no i sy at higher frequencies. P S D analyses were performed o n the t idal data p r io r to the process ing o f other data sets. Th i s w o u l d determine w h i c h frequencies should be returned b y the analysis o f the seepage and differential pressure system data. The t ida l e levat ion was co l lec ted at two sampl ing frequencies; 10 minute intervals i n i t i a l l y (1912 samples) and then reduced to 5 minute intervals (3195 samples). The dominant periods returned f rom these analyses are 24.5 hours (diurnal) and 12.3 hours (semi-diurnal) for the 10 minute data and 24.2 hours (diurnal) and 12.7 hours (semi-diurnal) for the 5 minute data. T h e per iodograms for these estimates are shown i n Figures 3-32 and 3-33, respectively. These periods are s imi la r to those reported b y T a n i g u c h i (2002) w h o measured S G D and t ida l f luctuations i n Osaka B a y , Japan, over a pe r iod o f 4.3 months. Tan iguch i transformed 3 months o f data, determining periods o f 24.1 hours (diurnal) and 12.3 hours (semi-diurnal) for both the S G D and t ida l measurements. Tan iguch i also found a 341.4 hour s ignal , ind ica t ive o f the 14 day lunar cyc l e (spring-neap tide). The t idal data col lected at T u r k e y Po in t , w h i c h the most tempora l ly extensive data set o f a l l the experiments, d i d not extend beyond 24 days 54 and thus w o u l d not have captured more than one lunar cyc l e and w o u l d not have seen the lower frequency s ignal o f the b i -mon th ly lunar tide. P S D analyses were performed on three o f the four seepage data sets; S 1 R 1 , S 1 R 2 and S 2 R 1 . Seepage data col lected dur ing the second run at site 2 ( S 2 R 2 ) was b e l o w the detection l i m i t o f the meter and was not i n place l ong enough to col lec t over a fu l l , d iurnal t idal cyc le . F o r these reasons, S 2 R 2 data was not analyzed. The S 2 R 1 seepage data was investigated i n two parts: first, as a part ia l data set w i t h o n l y the first 4 days o f seepage data and then the entire data set. The reason for this is because data co l lec ted between A u g u s t 2 8 t h and September 2 n d does not contain a v i s u a l l y s ignif icant t ida l influence. D a t a col lec ted between Augus t 2 4 t h and Augus t 2 8 t h , however , exhib i ted strong t idal effects. The results for these analyses are summar ized i n Table 3-4. Per iodograms are shown i n Figures 3-34, 3-35 and 3-36. Table 3-4. See page Data PSD results Site and R u n N u m b e r o f Samples 1 s t Dominan t Pe r iod (hours) 2 n d D o m i n a n t P e r i o d (hours) 3 r D o m i n a n t P e r i o d (hours) S 1 R 1 1669 46.5 23.2 11.6 S 2 R 1 992 (partial) 27.6 11.8 8.3 S 2 R 1 2555 42.6 12.5 N A The power spectrum density results show that some o f the seepage data sets exhibi t d iurna l and semi-diurnal t idal frequencies w h i l e others have not captured eas i ly identif iable frequencies. Per iodograms for S 1 R 1 and S 2 R 1 (Figures 3-34, 3-35 and 3-36) exhibi t frequencies that are nearly diurnal and/or nearly semi-d iurna l . The 42.6 and 46.5 hour signals observed i n S 1 R 1 and S 2 R 1 sequences are a two-day t ida l cyc l e that is not v i s u a l l y observed i n the non-transformed data (Figures 3-13 and 3-14). 55 S 1 R 2 P S D results (actual data plotted i n Figure 3-15) do not have a v i s u a l l y obvious t ida l frequency because o f anomalous ly l o w discharge rates for over h a l f o f the run t ime. A l s o , later i n the run, discharge rates increase sharply but do not appear to be dr iven b y t ida l signals. T h i s data d i d not produce t ida l periods and is not presented. F i n a l l y , the onshore w e l l data was investigated. A l t h o u g h not v i s i b l e i n the plot ted water table elevations at we l l s P I and N I , the analysis o f these we l l s showed that they captured semi-d iurna l as w e l l as diurnal tide signals ve ry w e l l . B o t h we l l s , sampled every 10 minutes, had near ly continuous data over the entire experiment pe r iod (Augus t 14 t h to September 6 t h ) . T i d a l signals analyzed at w e l l P I , w i t h 2891 samples, had a dominant per iod o f 24.09 hours and a secondary per iod o f 12.04 hours (Figure 3-37). W e l l N I , w i t h 3178 samples, experienced t idal influences w i t h a dominant per iod o f 24.08 hours and a secondary per iod o f 12.04 hours (Figure 3-38). W h i l e l a c k i n g sharp peaks indica t ing strong signals, the t idal influences expected w i t h i n the coastal aquifer, captured b y the onshore we l l s , are present. 56 Water Table at F S U M L field site at 12:10pm, Aug. 25th, 2002 Note: Ground surface elevation is approximate between wells - View is to the East 105 100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0 Dis t ance onshore f r o m L o w T i d e L i n e , m Figure 3-1. Onshore Cross-section showing location of P, N and A nests (and the water table position at 12:10pm on August 25th, 2002) 57 Elevation above average sea level, m ( N A V D 1988) to "1 n 6\ VJ V 65 3 a cr •1 S5 -cr JT «T a tt •a a I I I 1 i k to 1111 i 1 k '"0 111 1 3 CD O ^ n a" P- 5T B •a g © °-to g. n 1 1 1 1 to i H i | 1 ki •z w « & p " ro a" •P- rt *" 3 n> P •a £ to 3 § °" to g. I o 3 3 o" 3 o 3 sf o 3 o IT3 Vi 2 < O £?• ft C/l O 33 CD m Figure 3-7. Waterloo Profiler sampling at offshore site (site 1). Note: Wells behind the scaffolding are those of BC-nest 63 Elevation of Well Mid-Screen, N A V D 8 8 (m) ON crq s n u» 00 —-B" o s cr O « T3 C era c » • t/3 a •B <-*• <B 5 c re O © ro co cn O 5>" fi> 3 O CD o 5? 3 " O (D •n 3 3 r-o S Q. CD (D ro o o o CO o o o ro o ro o o CD o co o o o ro o o o • o •a C L • * CD | l » •a CD C L J ro H J r o r o ro CD O <J> OlO) Ul • • • • • • ro , . t • ro • I • ro co < Elevation of Well Mid-Screen, NAVD88 (m) fflSOUi^UM-'O-'MW^ O S 0\ (W e re « 5 —. 3' 3" *a •t o re* | . S-" cr © re 13 cr I > c era c C/3 re re 3 cr re i K> O o © D t o ' sr 3 O a> O 3-O o 3 r -o a. 3 CD O ro O O 05 O 00 O O O ro o -p». o ro r~ co co -«• * • ro ro • ro • co ro • ro • co ro O 3 CD C/> ft o Figure 3-11. Tide Meter 67 7 2 F i g u r e 3-17. Ins ide the D P S B o x 73 P1 water table elevation (m) CTQ C a H c.' s c S" 3 V 5 (ui '88QAVN) uoiJBAaig |epu (A c 22. o o 3 •a s> 5>" o 3 0 ) 0) CD m CD < 01 o' 3 fi) <D fl) 3 a a. w m CD < 0 ) 5' 3 > < D oo 0 9 3 "2 m d 5 < a) < £U — oo 5' 00 3 —1 ore c re w p re re •a &> CfQ tt re a E. W f tt S' s tt 3 tt 3' 5 tt « • re a c 3' oro in Seepage Rate (cm/day) Rainfall (mm/hr x 5) 8/16/02 0:00 8/16/02 12:00 8/17/02 0:00 8/17/02 12:00 8/18/02 0:00 8/18/02 12:00 8/19/02 0:00 O 8/19/02 12:00 Si. => 8/20/02 0:00 Q. <» 8/20/02 12:00 8/21/02 0:00 8/21/02 12:00 8/22/02 0:00 8/22/02 12:00 8/23/02 0:00 8/23/02 12:00 8/24/02 0:00 73 CD CD T 3 tu CQ CD 73 0) Q. SL m CD < 5" 3 0) 3 a 73 w 3' 5T 7J 0) » era' Head Difference (cm) co co o —^  ro 9/3/02 6:00 9/3/02 12:00 9/3/02 18:00 9/4/02 0:00 9/4/02 6:00 O 0) CD 5 9/4/02 12:00 3 CD 9/4/02 18:00 9/5/02 0:00 9/5/02 6:00 9/5/02 12:00 9/5/02 18:00 3>' CD 73 c 3 ro • o i W ? i E. O "< 3 1 -w fi) •* 3 =! CL <L ro h 2 ° 5- 8 ~ CD 0 w 0) CO ~ CD i i a = 1 S3 cr i o' 3 Q. § tt tt 3 a o ->• o ro O w o 4^  O o o o oo o to ->• o o o (Aep/iuo) 3}ey sBedaas (ui) 89QAVN 'UOUEA3|3 |epu. (w) 88QAVN 'uo ! ieAa |3 |epu. oo C i n K» v© «2 I C/3 re in O 1 65 B re" < SS S" s Differential Head (cm) o b b oo N5 go oo ST <o 00 S3 f3 m — o m -1 ^ 3 ro 7) D "0 W CO CD 3 (A O fi) 3 Q. Q. m CD < B) 5" 3 _» o o (ui) uo|}eAe|3 |ep|i 0 0 era c n W2 I O C/3 C/3 P O s a » 3 E .ST S" p Head Difference (cm) 9/3/2002 6:00 9/3/2002 12:00 9/3/2002 18:00 9/4/2002 0:00 9/4/2002 6:00 O S-| 9/4/2002 12:00 3 (D 9/4/2002 18:00 9/5/2002 0:00 9/5/2002 6:00 9/5/2002 12:00 H 9/5/2002 18:00 4 (Aep/wo) aiey aBedaas (ui) 88QAVN 'U0!1BA3|3 |epu Periodogram PSD Estimate for 10 minute Tidal Data 10 F ! ( •;• •• • ' ; i , • ; '--'.I .'.I 1 ' • • ' I; !•;• -|0"7 I L L r -i i' r I I I JL J r •• , ( J I I 1 1 1 ! I !/ I I I i J • I; i f j j . I, „q ~? ~1 0 1 10 10 10 10 10 Frequency (Hz) Figure 3-32. Periodogram for Tidal Data with 10 minute sampling frequency 88 Periodogram PSD Estimate for. 5 minute Tidal Data 10 | r 1 > .,: i. , | , . . . . . | r • • I ! i !.. 1 o" I : i I I i i i . . i; -i i i. i r l t . i r i i I r i 1 - .—i r :. i i i I io"3 io"2 10"' 10° io' Frequency (Hz) Figure 3-33. Periodogram for Tidal Data with 5 minute sampling frequency 89 Periodogram PSD Estimate - S1R1 Seepage Data 10 r. , , .. ; . ;j i .. . • ;..M..,| r ; I • • Frequency (Hz) Figure 3-34. Periodogram for Seepage Data of S1R1 90 91 6 Periodogram PSD Estimate - S2R1 Seepage Data - Whole Data Set - Window Size 2555 10 £ , r , , , , , , | , ... .. .. .. . - . ...,| r •—; • .• i | — • — r - •—r-...-•a •• -1 • Q - - V 10 TO 10 10 10 Frequency (Hz) Figure 3-36. Periodogram for Seepage Data of Run 2 at Site 1 92 Peripdqgran* PSD Estimate - Onshore Well PI - 2891 samples 10 ' | r r , ; , i , , , ' ' ' ' ' ' I- I I Frequency (Hz) Figure 3-37. Periodogram for W e l l P I W a t e r Table Elevat ion Data 9 3 Periodogram PSD;Estimate;-;Onshore Well NI - 3178 samples 10 ' | r i—... • . K> vj i 7—.. • ... i-i :LI r 7—; •  i > ill Frequency (Hz) Figure 3-38. Periodogram for Well NI Water Table Elevation Data 94 4. Numerical Model of the FSUML Site 4.1 O n e - D i m e n s i o n a l Compute r S imu la t i on 4.1.1 Rationale and FRA C3D VS Description Researchers examin ing wave , tide and groundwater interactions i n coastal aquifers and nearshore environments have used numer ica l models to shed l ight on the phenomena o f submarine groundwater discharge, saltwater in t rus ion and freshwater/seawater m i x i n g (Ata i e -Ash t i an i et a l . 1999; U c h i y a m a et a l . 2000; L a n g e v i n 2003; S m i t h and Z a w a d z k i 2003). The results o f a few o f these invest igat ions are w o r t h descr ib ing here. A t a i e - A s h t i a n i et a l . (1999) mode led the effects o f t ida l fluctuations on saltwater in t rus ion i n unconfmed aquifers, where one f ind ing was that s m a l l discharge ve loc i ty vectors are produced dur ing a rising tide and that large discharge ve loc i ty vectors are produced dur ing a fa l l ing tide. In addit ion, A t a i e - A s h t i a n i et a l . (1999) discovered that the seaward freshwater f lux i n the aquifer has considerable influence o n both the shape and loca t ion o f the seawater/freshwater contact. F i n a l l y , the authors found that b y neglect ing t idal f luctuation effects resulted i n an inaccurate evaluat ion o f the water table elevat ion i n the onshore reg ion o f the aquifer. U c h i y a m a et a l . (2000) used a numer ica l m o d e l to simulate nutrient discharge f rom groundwater discharge into the K a s h i m a Sea o f f the east coast o f Japan based o n f ie ld measurements and compared w i t h nutrient concentrations measured i n the Tone R i v e r . T h e y d iscovered that w h i l e nutrient fluxes from S G D were m i n o r compared to those from the Tone R i v e r , it is recognized that S G D m a y have been underestimated because macropore and wave-setup effects were not inc luded i n their mode l . L a n g e v i n (2003) used a density-dependent, three-d imens iona l f l ow m o d e l , S E A W A T , to predict S G D into B i s c a y n e B a y , F l o r i d a . The 95 author determined that for 6 o f the 9 years modeled , the S G D magnitude was approximate ly 10% o f the surface water discharge to B i s c a y n e B a y w h i l e the 3 driest years produced S G D exceeding surface water discharge. L a n g e v i n (2003) notes, however , that results contain a h igh l eve l o f uncertainty as the f ie ld measurements o f S G D proved problemat ic and were not used to val idate the m o d e l . N u m e r i c a l m o d e l i n g i n this context is useful to test assumptions made about coastal aquifers and the processes d r i v i n g S G D . 4.1.2 Construction Details A 1-dimensional c o l u m n m o d e l was developed to simulate seepage and differential pressures i n the seabed, g iven t ida l condi t ions at the site. The c o l u m n m o d e l dimensions are l m b y l m b y 10m ( X Y Z ) w i t h the c o l u m n subd iv ided at the centimeter scale (0 .01m ver t ica l spatial resolution). The ver t ica l subdivis ions enable the ca lcu la t ion o f small-scale changes i n hydrau l ic head. O n l y non-density-dependent f l o w was used i n the s imula t ion . T h i s is because unrealist ic head distr ibutions were p roduced us ing densi ty-dependency as a result o f i r resolvable errors. The doma in for the m o d e l includes the Sur f ic ia l A q u i f e r ( S A ) and the Intracoastal Fo rma t ion (IF), bo th o f w h i c h are assumed to be homogenous. T h e upper 6 m o f the c o l u m n are designated as S A , this depth is based o n the refusal depth o f the Wate r loo Prof i le r at site 1. The bot tom four, meters o f the c o l u m n are designated as IF . T h e hydrau l ic conduc t iv i ty o f the S A was chosen to be 1 x 10" 5 m/s , w h i c h is the geometric mean o f we l l s B C 1 and B C 4 , the we l l s closest to site 1 where K values had been determined. The hydrau l ic conduct iv i ty o f the IF was chosen to be 1 x 10" 9 m/s, estimated 96 f rom descriptions o f the IF b y Schmid t (1984). Other mater ial properties i nc luded i n the m o d e l are l is ted i n Table 4-1. A conceptual d iagram o f the 1-dimensional m o d e l is presented i n Figure 4-1. Table 4-1. Material properties of Surficial Aquifer and Intracoastal Formation Parameter Surficial Aquifer Intracoastal Formation H y d r a u l i c C o n d u c t i v i t y (m/s) 1 x 10" 5 1 x 10" y Storage Coeff ic ient 0.001 0.0001 Poros i ty 0.3 0.05 A constant head o f 1.5m was appl ied at the top o f the U p p e r F l o r i d a n A q u i f e r , w h i c h is the base o f the m o d e l at the I F / F l o r i d a n A q u i f e r contact. The m o d e l was run to steady-state w i t h constant heads at both the top and the bo t tom o f the m o d e l . O n c e steady-state was reached, heads ranging f rom 0.331 to 1.403m were appl ied at the top o f the S A , s imula t ing t idal fluctuations. The fluctuations are based on real t ida l e levat ion data col lected at the site dur ing each per iod o f t ime under examinat ion . T h e t ime periods examined were between 2 and 4 days, enough t ime to a l l ow for at least two t ida l cycles to be analyzed. T h e t ida l elevations, referenced to N A V D 8 8 , were corrected to be representative o f the height o f the water c o l u m n above the seabed at site 1. T h e head fluctuations i n the m o d e l are programmed to occur at either 5 or 10 minute intervals over the course o f the run, depending on the co l lec t ion rate o f tide data. T h e m o d e l is s imula t ing a zone 2 0 - 4 0 m beyond the l o w tide l ine where no input f rom the surf ic ia l aquifer is considered. Observa t ion points are inc luded i n the m o d e l at intervals o f 0 .2m (ver t ical depth) to observe the hydrau l ic head fluctuations w i t h depth. Differences between observed 97 hydraul ic head at depth simulate the differential piezometer system output. T h e depths used f rom the s imula t ion are 0 .4m and l m b e l o w the seabed, p r o v i d i n g a A z o f 0 .6m. Despi te the errors associated w i t h the calibrations o f the D P S , it is useful to examine h o w hydrau l ic heads are fluctuating i n the seabed. Seepage rates (described as f lux rates i n F R A C 3 D V S ) across the seabed are also predicted us ing the I D m o d e l . A f lux output node was specif ied at the top o f the c o l u m n (at the seabed), w h i c h shou ld approximate the observed submarine groundwater discharge rate measured w i t h the automated seepage meter. T h e m o d e l was calibrated b y va ry ing the hydrau l ic conduc t iv i ty o f the S A and the IF to see h o w these changes w o u l d affect the S G D and D P S results. 4.2 Resul ts and D i s c u s s i o n The m o d e l results indicate that g iven the t idal data recorded i n the f ie ld and w i t h the estimated aquifer properties indicated above, discharge and recharge shou ld be observed at the f ie ld site ( F i g u r e s 4-2). In this figure, a t ida l sequence o f approximate ly four semi-diurnal cycles is superimposed w i t h the s imulated discharge rate and the s imulated differential head. The s imulated processes o f discharge and differential head are d r iven i n the same di rec t ion b y the fluctuating tide. A s s u m i n g the automated seepage meter was per forming accurately, no recharge was recorded i n the experiments dur ing Augus t and September 2002. W h i l e the seepage meter cannot measure recharge, a reading o f zero or the m i n i m u m reading is expected i f recharge is occur r ing . It is possible , a l though u n l i k e l y , that the uncalibrated seepage meter was p roduc ing discharge readings w h e n i n real i ty recharge was occurr ing , yet there is no w a y to ver i fy i f this was the case. Recharge m a y exp la in the periods o f absent t ida l signals i n the seepage data but 98 again, ver i f ica t ion o f this hypothesis is not an option. The discrepancy between the m o d e l and the f ie ld results is not currently understood. N o recharge at the F S U M L site was reported b y Tan iguch i et a l . (2003) for the intercomparison study nor has previous w o r k at the mar ine lab discussed observations o f recharge. Y e t recharge can occur at h i g h tides as documented b y Paulsen et a l . (2003), where two ultrasonic automated seepage meters spaced 2 meters apart i n Wes t N e c k B a y , N e w Y o r k , recorded discharge i n one meter and recharge at the other. The authors conc luded that spatial heterogeneities m a y exist at the meter scale and this m a y account for the f ie ld seepage measurements. In regards to the t im ing o f the f lux rates, it is observed f rom the m o d e l that the highest discharge rates occur dur ing the transit ion from highest to lowest t ide (Figure 4-2). Burnet t et a l . (2002) also note that the highest f lux rates observed at F S U M L i n A u g u s t 2000 occurred dur ing the transi t ion f rom highest to lowest t ide (Figure 2-3), an observation ident ica l to what was found dur ing the current research (Figures 3-17, 3-19, 3-21). Converse ly , recharge is at a m a x i m u m i n the m o d e l results dur ing the transit ion from l o w to h i g h tide. T h i s corresponds to discharge be ing at a m i n i m u m at the f ie ld site both i n 2000 and 2002. The fact that the s imula t ion is able to predict the occurrence o f seepage osci l la t ions w i t h tide as seen i n the f ie ld impl i e s that the condi t ions appl ied w i t h i n the m o d e l are accurate to some degree. It is interesting to note, however , that the f ie ld results f rom Pau lsen et a l . (2003) do not match those found b y this study or Burnett et a l . (2002). Paulsen et a l . (2003) record that seepage rates are at a m a x i m u m immedia te ly after the l o w tide i n a l l data d isp layed yet they do not speculate on this. The discrepancy m a y be caused b y the fact that the aquifer at Wes t N e c k B a y is composed o f g lac io - f luv ia l m e d i u m to coarse-grained sand, w h i c h is h i g h l y conduc t iv i ty mater ial 99 w h i l e the Sur f i c ia l A q u i f e r at F S U M L is composed o f s i l ty sand (geometric mean hydraul ic conduc t iv i ty o f S A is 8 x 10"6 m/s). N o hydraul ic conduc t iv i ty data is p rov ided b y Paulsen et a l . (2003). H e a d vs. depth profi les created f rom the I D m o d e l are presented i n F i g u r e 4-3. Th i s figure demonstrates head fluctuations w i t h depth over a pe r iod o f 17 hours. N o t i c e that there is continuous discharge f rom the Intracoastal Fo rma t ion , albeit w i d e l y v a r y i n g w i t h the tides, caused b y the appl ied boundary condi t ion at the base o f the m o d e l . The sequence o f discharge and recharge across the seabed is also cont ro l led b y the t ida l influences appl ied to the seabed surface at the top o f the Sur f i c i a l A q u i f e r . C o m p a r i n g F i g u r e s 4-2 a n d 4-3, the s imulated discharge can be expla ined b y c lose ly examin ing h o w the t idal osci l la t ions affect the sediments at depth. A t h igh tide, hydrau l ic head i n the seabed is static and discharge is at zero but begins to c l i m b immedia te ly after h i g h tide is passed, ind ica t ing that pressure is decreasing and a l l o w i n g b rack i sh water to discharge. A s the tide drops, the rate o f discharge increases w i t h the decreasing pressure. A t just after m i d w a y between h i g h and l o w tide, w h e n the slope o f the fa l l ing tide is at its m a x i m u m (highest rate o f pressure change or highest differential head), the seepage rate reaches its m a x i m u m , w h i c h impl ies that the rate o f pressure change at the seabed controls the rate o f change i n the discharge. A s the rate o f t ida l e levat ion decl ine begins to s low (decreasing slope o f the tide), the discharge rate begins to drop as w e l l . W h e n the tide reaches its m i n i m u m , the seepage rate is again at zero because the head has equal ized across the S A . W i t h increas ing pressure at the seabed f rom the increasing tide, recharge rates increase, reaching a m a x i m u m at the point o f m a x i m u m t idal increase ( m a x i m u m slope). F i n a l l y , as the t ida l 100 elevat ion flattens again towards h i g h tide, the recharge rate decl ines, reaching zero at h i g h tide. Thus , the m o d e l indicates that the rate o f discharge from (or recharge into) the Sur f i c ia l A q u i f e r is rea l ly caused b y the rate o f change i n pressure caused b y the fluctuating tides and the hydraul ic head assigned at the top o f the U p p e r F l o r i d a n Aqu i f e r . One-d imens iona l s imulat ions were run f rom t idal data col lec ted dur ing run 1 at both sites ( S 1 R 1 and S 2 R 1 ) and run 2 at site 1 (S1R2) . The hydrau l ic conduc t iv i ty at site two, 3.59 x 10" 6 m/s, based o n the geometric mean o f the hydrau l ic conduct iv i t ies o f we l l s C I and C 3 , the we l l s c losed to site 2. A l l other variables were kept the same for this s imula t ion . Seepage data was plotted w i t h m o d e l results a long w i t h the t ida l fluctuations inputted to the m o d e l because it w o u l d be useful to see h o w the f i e ld results compare to the s imula t ion . The S 1 R 2 , S 1 R 1 and S 2 R 1 plots are presented i n Figures 4-4, 4-5 and 4-6, respectively. No tab le i n Figure 4-5 is the observat ion that the m o d e l c lea r ly predicts temporal variat ions i n discharge rates measured i n the f ie ld . Discharges peak at about l c m and w i t h hydrau l ic conduct ivi t ies o f 1 x 10" 4 m/s for the S A and 1 x 10" 5 m/s for the IF , discharges peaked at about 2 c m . The m o d e l cal ibrat ions, however , as descr ibed above were unable to predict accurate discharge magnitudes, the s imulated discharge rates a lways fa l l ing m u c h lower than the f ie ld rates even at extreme parameter values. The relat ionship between t idal height above the seabed and seepage rate is presented i n Figures 4-7, 4-8 and 4-9. A s discussed previous ly , a non- l inear relat ionship appears to exist between the seepage rate and the t ida l fluctuations. A s can be seen c lear ly i n these figures, a dist inct c i rcular cyc le is noted, f o l l o w i n g the t ida l cyc le . W h e n 101 the tide is at an extreme, high or low, discharge is at or near zero. When tides are at mid-levels, discharge is at a maximum or minimum, depending on the direction of the tide. As expected, simulated differential heads behave in a way very similar to that of simulated discharge as described above. The greatest differential heads are produced when the highest rate of pressure change is achieved, which is at the point of greatest slope of the falling or rising tides. These two processes act in tandem and quality field data should allow for prediction of the magnitude and direction of one process from the other. While the current research did not produce data where prediction of this nature could be achieved, future work should aim to minimize chances of calibration and instrument error in order to collect high quality differential head data. 1 0 2 Speci f ied Head at Top F a c e : 0.25m to 1.40m (Tidal Osci l la t ions) C h a n g e at 5-minute intervals 6m S A Vert ical Discret izat ion of 0.01m (not to sca le) K S A = 1 x 10"5 m/s K F = 1 x 10"9 m/s 4m IF 1m Speci f ied Head at Bottom F a c e : 1.5m Figure 4-1. Conceptual Model for 1-Dimensional Simulation 103 o S1 9/4/02 12:00 9/4/02 15:00 9/4/02 18:00 9/4/02 21:00 9/5/02 0:00 9/5/02 3:00 9/5/02 6:00 9/5/02 9:00 9/5/02 12:00 9/5/02 15:00 Simulated Seepage Rate (m/day) Simulated Differential Head (m) o b o b 9/3/02 6:00 9/3/02 9:00 9/3/02 12:00 9/3/02 15:00 9/3/02 18:00 9/3/02 21:00 9/4/02 0:00 9/4/02 3:00 9/4/02 6:00 o 9/4/02 9:00 o b o ro o o b o (w) paqeas eAoqe ep j i jo )i|B;eH era o ON o 5 g , era c B 3 3 1 re i 4 -O a A 2! O 3" re 69 «> e a CO re re - K -a 3" » re era «. » £. J? sr 2, Q . 8 Sr s T3 te era » -a 65 c « 5 ° = 3 65 M I ^ CO - re g B re n » » re 69 S 8  3 a  » e 3 5 & & •j re » 2 re re B 65 re 3 1 S re 63 9/3/02 6:00 9/3/02 9:00 9/3/02 12:00 9/3/02 15:00 9/3/02 18:00 9/3/02 21:00 9/4/02 0:00 9/4/02 3:00 9/4/02 6:00 g • 9/4/02 9:00 3 —i 9/4/02 12:00 I ro 9/4/02 15:00 9/4/02 18:00 9/4/02 21:00 9/5/02 0:00 9/5/02 3:00 9/5/02 6:00 9/5/02 9:00 9/5/02 12:00 9/5/02 15:00 Simulated Seepage Rate (m/day) Simulated Differential Head (m) o o o o o o o (Aep/uio) ajey eBedees pajnsea|/\| (01., ui) paqeas aAoqe e p u JO }u6|aH o o 5 2, era* 3 c 5 re a 4^ M JL re • » ? c © "1 J?" 9 Z a cr cc as re 5- re *g cr » re (ro 2 rt 3"' ,53 CL ° 2 3 6 72 ffi -o T. — M O <-• re sf-MM* o» c. r-H CP a r » SE 9 i 3 E a S. 5/3 re * -a K » re n 9 » re &i 3 " re & ti g a a 3 3 a B S- re re a —• » s te Simulated Seepage Rate (m/day) Simulated Differential Head (m) 8/19/02 0:00 8/19/02 6:00 8/19/02 12:00 8/19/02 18:00 8/20/02 0:00 8/20/02 6:00 8/20/02 12:00 8/20/02 18:00 8/21/02 0:00 D s * 8/21/02 6:00 BI 3 H 8/21/02 12:00 I 8/21/02 18:00 8/22/02 0:00 8/22/02 6:00 8/22/02 12:00 8/22/02 18:00 8/23/02 0:00 8/23/02 6:00 8/23/02 12:00 8/23/02 18:00 o o o o o o o o o o o o o o o o o o o o (Aep/uio) aiey aBedaas pajnseaw (0 V » w) paqeas a A o q e apu ;o |l|B|8H o oo ft 5* 2. era 3 re P sr ^ 2. » * ss 2. 3 * a CP <73 LS n S- re «. -a CP SB re CTQ BO rt £ sf « cf ? 3 re ss 73 era h -« T3 3 — ss o re _j-5' ^ Ml* S CP D ^ =,• c« w g s 3 3 5 ss 6 i «-«S « . * re < re S-tl 2. ss re n re ss 3 * re ss ss 3 g Q. 3 2 a 3 •j re •*. re re cj J S" »§ B fD ss ss a> '< 3 8/24/02 6:00 8/24/02 12:00 8/24/02 18:00 8/25/02 0:00 8/25/02 6:00 8/25/02 12:00 8/25/02 18:00 8/26/02 0:00 « 8/26/02 6:00 CD | 8/26/02 12:00 3' 8/26/02 18:00 CD 8/27/02 0:00 8/27/02 6:00 8/27/02 12:00 8/27/02 18:00 8/28/02 0:00 8/28/02 6:00 8/28/02 12:00 8/28/02 18:00 Simulated Seepage Rate (m/day) Simulated Differential Head (m) • i i i i i i i o o o o o o o o o o o o o o o o o o o o o o o o o b b b b b b b b b b b b b b b (Aep/wo) ajey eBedeas pejnsee|Aj (Ol. » ui) paqees aAoqe ap j i jo luBjei-i 3 CD 0) (A c co 3 i i E CD a> CD CD a, a. to 0 CD f i CD • £ w B> ro - 73 CD 0) X! a- o 3 CD a a. a 5 CD 3" 1 W 3^ 1 tu cr o < CD CO CD »•§ ft) CD CT CD a. tu CD ft) 3 a 3 1 % co ft ffi a era e n Go 3 c_ 5" m a. GO re 13 & era re rt c SO ce re a o s a . o cc I—' re a Ti-er re a era 65 3' En re 33 re era* Si o < re CO re » cr re a-Simulated Seepage Rate (m/day) o b o b o b o ro o o o cn H CL. CO o I b 2. CD' 3 P ju —J cr O S- o CO bo <T> B> cr <D O °- b o o Ol o o o o o O l o b o b o b ro o o ro O l ro • • • • • •«* m+ <•» • • • • • • • • • • • •»!«»•«•< ••f • 4« • • • • • • (A CO i ' c_ Bl o a. co CD CD T3 Bl (Q CD 73 Bl r+ (ft •a o ft n a Bl ca Bl 5' CO Q. CB CQ 3" Bl CT O < CO CD Bl cr CD a Simulated Seepage Rate (m/day) o o O o o o o o o o o b O b b b b b b b b b O o o o o o o o o —^ o 00 a> -*>. to o to ro oo o 5. Conclusion T h i s research attempted to investigate the processes d r i v i n g submarine groundwater discharge from three basic perspectives: characterizat ion o f the coastal aquifer, offshore analysis o f S G D and a numer ica l m o d e l i n g approach. Character izat ion o f the coastal aquifer i n v o l v e d the analysis o f onshore water table levels i n order to calculate an appropriate hor izonta l hydrau l ic gradient, found to be 0.025, based o n the average drop i n hydraul ic head between we l l s P I and N I . S l u g testing o f these we l l s and w e l l P 2 us ing the H v o r s l e v method produced hydrau l i c conduct iv i ty values that m a y point to a ver t ical structure o f decreasing hydrau l i c conduct iv i ty w i t h depth w i t h i n the Sur f ic ia l A q u i f e r as postulated b y S m i t h and Z a w a d z k i (2003). The geometric mean and arithmetic mean o f these three we l l s is 6.4 x 10" 5 m/s and 1.3 x 10" 4 m/s, respectively, w h i c h is w i t h i n the range o f other we l l s at the site. T w o major s torm events p rov ided insight into the nature o f f l o w i n the onshore region, ind ica t ing that w e l l P I responded more q u i c k l y to inf i l t ra t ion than w e l l N I despite be ing i n sediment 31 times less conduct ive. F r o m observations o f t ida l osci l la t ions present i n these we l l s it is l i k e l y that a l o w conduc t iv i ty uni t exists just shoreward o f w e l l N I , damping the effects o f t ida l influences. A discharge rate over an offshore area comparable to the 2000 intercomparison study, based o n the onshore w e l l hydrau l ic conduc t iv i ty values and hor izonta l hydraul ic gradient, was calculated to be lower than that measured b y the intercomparison researchers b y 1-2 orders o f magnitude. T h i s result is consistent w i t h a m o d e l b y Smi th and Z a w a d z k i (2003) whose ve r t i ca l ly layered Sur f i c i a l A q u i f e r mode l obtained a value o f 0 .15m / m i n w h i l e the D a r c y ' s l aw ca lcula t ion from this study obtained values o f 0.14 m / m i n and 0.07 m / m i n for the 112 onshore w e l l K value ari thmetic and geometric means, respect ively. One poss ib le explanat ion for this is leakage across the Intracoastal Fo rma t ion into the Sur f i c i a l Aqu i f e r . Submar ine groundwater discharge experiments were focused o n two offshore sites where an automated seepage meter was coupled w i t h a new apparatus, a differential piezometer system, w h i c h was designed to measure differential head fluctuations i n the seabed. Seepage rates across the seabed had measured rates v a r y i n g between 20-40 cm/day o n average al though peaks o f up to 80cm/day were observed. W h i l e recharge cannot be recorded w i t h the seepage meter, continuous discharge as observed is equivalent to no observed recharge. Occas iona l periods o f " l o w " discharge rates (10-20 cm/day) w i t h m i n i m a l t idal influence were observed over the course o f several t idal cycles . La rge ca l ibra t ion errors associated w i t h the D P S differential pressure transducers, w h i c h are o f the same magnitude as the measured data, prevent the use o f the D P S data i n an analysis. Despi te this, a p re l iminary assessment o f the data is comple ted . Di f f i cu l t i e s w i t h the differential piezometer system can be avoided i n the future b y careful lab calibrations o f any system planned for use i n the f ie ld and running cal ibrat ions w h i l e i n the f ie ld . A c q u i r i n g a c o m m e r c i a l l y avai lable 2-port, differential pressure probe should be considered for future experiments. A 1-dimensional numer ica l s imula t ion was used to examine h o w s imulated seepage rates compare w i t h observed rates. The m o d e l results support the f ie ld evidence i n regards to temporal per iod ic i ty o f seepage w i t h t idal fluctuations. It is observed that the relat ionship between the seepage rate and the t ida l stage appears to be a non-l inear one w i t h m a x i m u m discharge or recharge occur r ing at t imes o f m a x i m u m slope o f t ida l 113 oscillations. It is determined that the rate of pressure change at the surface and the hydraulic head assigned at the top of the Upper Floridan Aquifer drive the changes discharge rate. 114 Appendix A - Slug Test Recovery Data Slug Test - Well N1 - September 7th, 2002 approximately from 9:05 to 9:40am 230 -, 220 Time (hr:min) 115 Slug Test - Well P1 - September 7th, 2002 approximately from 13:40 to 14:15 Time (hr:min:sec) 116 250 240 230 220 210 200 190 180 170 160 Slug Test - Well P2 - September 7th, 2002 approximately from 14:19 to 14:45 E o H Q. a> > o n ro c o ro > o LU La 0) I •M- • •* •* •<r •sf •* •* •>f •<f • f •q- •<f •tf •* o O o o o o O O O O O o o o o O O o cb oo o CN •ir CD 00 o CN •if CD CO o CN •if CD CO o T— CN CN CN CN CN CO CO CO CO CO •«f •* •* •* •*f If) •if •ir •ir •ir •if •if •if' •if •if •if •ir •if •ir •if •if •if •if Time (hr:min:sec) 117 Appendix B - Slug Test Recovery Data Slug Test Results for new onshore wells at FSUML well N1 -r 0.0254 m well screen radius Le 0.4064 m screen length R 0.0254 m borehole radius t37 min K [m/min] K[m/d] K[m/s] Method N1 i1 in 0.367 0.0059966 8.64 0.00010 Hvorslev N1 01 out 0.45 0.0048905 7.04 0.00008 Hvorslev N1 i2 in 0.41 0.0053677 7.73 0.00009 Hvorslev N1 o2 out 0.46 0.0047842 6.89 0.00008 Hvorslev N1 i3 in 0.379 0.0058067 8.36 0.00010 Hvorslev N1 03 out 0.225 0.0097811 14.08 0.00016 Hvorslev N1 Average K [m/s] = 1.02E-04 (med. to coarse sand) well P1 r 0.0254 m well screen radius Le 0.4445 m screen length R 0.0254 m borehole radius t37 min K [m/min] K[m/d] K[m/s] Method P1 i1 in 3.83 0.0005423 0.78 0.0000090 Hvorslev P1 Average K [m/s] = 9.04E-06 (med. to coarse sand) Note: P1 - Only the first slug test (i1) was used Well P1 had not fully recovered when the slug was removed from the borehole well P2 r 0.0254 m well screen radius Le 0.9144 m screen length R 0.0254 m borehole radius t37 min K [m/min] Kfm/dl Krm/sl Method P2 i1 in 0.092 0.0137412 19.79 0.0002290 Hvorslev P2 01 out 0.15 0.0084279 12.14 0.0001405 Hvorslev P2 i2 in 0.077 0.016418 23.64 0.0002736 Hvorslev P2 02 out 0.0875 0.0144478 20.80 0.0002408 Hvorslev P2 i3 in 0.093 0.0135934 19.57 0.0002266 Hvorslev P2 o3 out 0.0375 0.0337116 48.54 0.0005619 Hvorslev P2 Average K [m/s] = 2.79E-04 (med. to coarse sand) 118 Appendix C - Slug Tests - Normalized drawdown versus time Wel l N1 - s l u g 1 (in) 1.00 o E a o S a. in 0.10 0.01 * 1 . . +t *^**»» ***** *~% ********* ****, **** **** * ***' 0.000 0.100 0.200 0.300 0.400 0.500 0.600 0.700 0.800 0.900 1.000 Time (minutes) We l l N1 - s l u g 1 (out) 1.00 c di E o o ro a. 0.10 0.01 *** *********4 * ** * * ***i B ***** 0.000 0.100 0.200 0.300 0.400 0.500 0.600 0.700 0.800 0.900 1.000 Time (minutes) 119 WellNI - slug 2 (in) 1.00 Q. (A =5 0.10 T3 N 0.01 * \ • • * ****«~i• i •*< ******** ***, • • • • r • . * • " ***** **** * • * **\* * 0.000 0.100 0.200 0.300 0.400 0.500 0.600 0.700 0.800 0.900 1.000 Time (minutes) Well N1 - slug 2 (out) a E o o JS Q. w •o •o o N 0.10 0.01 til —^****i **^. * * *****< •••• *********** 0.000 0.100 0.200 0.300 0.400 0.500 0.600 0.700 0.800 0.900 1.000 Time (minutes) 120 Well N1 - slug 3 (in) 1.00 (A ••5 0.10 © N 0.01 • **i ******** ****», ******* ***** *********J. **** ***** * **** * ***< ****** • ***** **** ******. ****** * * * **\ * i^**** J —• 0.000 0.100 0.200 0.300 0.400 0.500 0.600 0.700 0.800 0.900 1.000 Time (minutes) Well N1 - slug 3 (out) 1.00 ID E a o ra a. .2 '•u x> <». 0.10 0.01 • •*( '"-******< ***** ->. **•* **** * *** * * * * * * * * ***»*^ \* ~ * * i 0.000 0.100 0.200 0.300 0.400 0.500 0.600 0.700 0.800 0.900 1.000 Time (minutes) 121 lisplacement D O W e l l P1 - s l u g 1 ( in) lisplacement D O r~~ 00 lisplacement D O lisplacement D O lisplacement D O lisplacement D O lisplacement D O lisplacement D O lisplacement D O lisplacement D O normalized d ^ C normalized d ^ C normalized d ^ C normalized d ^ C normalized d ^ C normalized d ^ C normalized d ^ C normalized d ^ C normalized d ^ C U . U 1 i 0.000 0.5 00 1.C 00 1.S 00 2.C 00 2.£ Time (n 00 3.C linutes) 00 3.E 00 4.C 00 4.E 00 5.C 122 Well P 2 - s l u g 1 (in) 1-00 -r* 1 1 1 1 1 =5 0.10 0 . 0 1 _| i i i i | i i i i | i i i i | i i i i | . i | . i i i | i . . i | i i i i | i i i i | i i i i | 0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350 0.400 0.450 0.500 Time (minutes) Well P2 - slug 1 (out) LOO -r— 1 1 1 1 1 r a E o u re a. .52 T 3 •a o N 0.10 • • • Q I 0 1 _ | I I I I | I I I I | I I I I | I I I ! | I I I I | I I I I | I I | I I . I | I I I I | I I I I | 0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350 0.400 0.450 0.500 Time (minutes) 123 1.00 i c 01 E o o ra Q . (A ™ n m Well P2 - slug 2 (in) 00 • • • * • i • • • • • • • • « • T3 U.1U T3 01 N ra o c 0.01 • O.O • • • 1 • • • • • 00 0.0 50 0.1 00 0.1 50 0.2 00 0.2 Time (tr 50 0.3 linutes) 00 0.3 50 0.4 00 0.4 50 0.5 1 00 < Well F 3 2 - s l u c j 2 (out) displacement displacement displacement displacement • displacement • • displacement displacement displacement • • • • • *-i normalized i 3 | normalized i 3 | • • • normalized i 3 | • * normalized i 3 | • • • • normalized i 3 | . • • • • normalized i 3 | • normalized i 3 | • • • normalized i 3 | U.U 1 0 0. 05 0 1 0. 15 0 2 0. Time (n 25 0 linutes) 3 . 0. 35 0 4 0. 45 0.5 124 Wel l P 2 - s l u g 3 (in) 0.10 * • • • , • • • • • • • • * • * • • • • < • • • • • • • • 0.01 4 0.05 0.1 0.15 . 0.2 0.25 0.3 0.35 0.4 0.45 0.5 Time (minutes) Wel l P 2 - s l u g 3 (out) 1.00 I 4. * ' • • • • • , * * • » • • • • • • • • • • • » CL in =5 0.10 •a s 0.01 0.05 0.1 0.15 0.2 0.25 0.3 Time (minutes) 0.35 0.4 0.45 0.5 125 co ro co co ~v i cn cn -F^ co ro CO 0) 3 •o CD O D" fi> oo C O C O 00 CO CO co co CO co co CO 00 ro fo ro fo fo fo fo fo fo fo fo CJl CJl CJI CJI on O l O l CJl CJl CJl CJl ^ ro ro ro i o ro fo fo fo fo fo fo fo fo o o o o o o o o o o o O o o o o o o o o o o o o O o ro ro ro ro ro ro ro ro ro ro ro ro ro o o CD" o 5" 3 D fi) CD CO O CO ^ SL Z SL <D 5-' —• S." O < Q. _ <D _ CO m a D OTSOTOTOCIO>>OCD £ ^ • ' c n - » . 4 ^ - » - N i - » - W Q 3 ct> CT> O D "0 Tl CO CO CO CO £ £ CD 3 CD CD _ o o * 3 CD 2 o as o 3 CD 0 3 * S ° S o C L CD ->• ro CD ro CJI CO ro ro 00 o o C T ) C D i o ^ C J l C D M C O b ) 4 ^ b o co ro ro CO CO —»• CD CD —k X CD 4^ 00 i o b CO SL pi" •0 CD C ^ CD C L cr ro co ro CD 00 CO co CJI i o i o bo N O O) 4^ o CD O CO 4^ CJl CJl ro CJl CJl 00 —v — X •>J 0 CD ro CD CD ro CJl CO 0 CO CD CD ro CO CO CO co 00 4^ 4^ CD — ^ CD —»• -»• ro 00 4> CJl ^-J CJl ro CO ro ro CD CJl o o 4^ 4^ CO ro ro co co 0 CO CO 0 ro CD CO CO CJl —». o CO SL < o •a T3 CO c CD •a 3 X CD f—t-o ' D C CO CD C L cS —I Q . E" o ' CO oo co co co -vi CT> cn 4^ W C A ) « C O M M W N 3 W c o r o - » - o c o o o ~ - j C T ) C T i ro ro ro 4 -^ co ro ro ro —^ O CD CO to 3 o) cn =j CD CD CD CD C D C D C D C D C D C D C D C D C D 4^ 4^ U U U Q W W W U U ro ro ro ro f o f o f o f o f o f o f o f o f o o o o o o o o o o o o o o o o o o o o o o o o o o o ro ro ro ro r o r o r o r o r o r o r o r o r o co CO 00 go oo 00 go 00 ro ro ro ro fo fo fo fo CD CD CD oo CO go go ro ro ro ro ro ro fo fo o o o o o o o o o o o o o o o o ro ro ro ro ro ro ro ro o O go go Jo" fo fo 2. ro ro 3 o o 3 o o o ro ro fi> (D > > ro co 5 ro co 4--^ ^ CD ixi CD r- r- O O O cn CD -t> co ro c/> (D o o 9. =fi =fc CD I—t-CD cn cn o o o o =fi =ft =ft =fi -T> co ro o o ID co co co 00 O <D ZJ == -»• o o - i to o o '->• ro _».ro - - i M r o r o r o w M -CD -T- c n c D c n r o r o - ^ - > - C D C D j - c n o r o ^ ' - ^ ^ c o - r ^ - v j - 1 -oo ro ro ->• o oo cn oo co ro ro ro co CD CO N O cn bo CD o co to SL < o o — ro ro -Q "D O ro CO ro ro CO CO co CO 4 - 4 - ro 4^ CO 4 - 4 - 4 - cn cn P co ro ->j 00 4 - — cn cn CT> co o CO co o o cn CT) CO - — ro ro -vl cn ro co CO ;-— CD cn cn ro 4^ cn cn o CO L_ CD cn cn CD -r- CO CO L_ cn CT) CO CD 00 ro ro CT) CT) co -r- cn --J o —>• CT) 00 •^ 1 CD CD cn cn g co co ~" ro co s to o o 4 - 4^ -—. ro - - - > - r o r o r o r o r o r o - > - co --• ro ro co ro oo S -t- ^ S 4 - . r 0 O M C D C D C 0 O CO CT) S ^ O l M ^ CD r o c n c o u b c o r o b - co oo ^ co co co o 4 -o o — _ ; » v •—• O CO 4 - - J T3 cn cn 00 -v! cn CO cn cn cn cn cn cn cn 4^ 4^ co ro —^  o CD 4i. -P^  -t^  4^ 4^ 4^ co -vj co cn 4^ co 4^ 4^ IO -»• CO fi> 3 4^ CO W O O CD CO JfJ cn CD 3 •D CD CL ro o o co CD CD ro ro o o o o co co CD ro o o co CD ro o o CO CD CD CD CD CO CD ^ ^ ^ ~3 ro ro ro ro fo fo o o o o o o o o o o o o ro ro ro ro ro ro CD CO CD CO CD CD CO CO CO CO CO CO fo fo fo fo fo fo o o o o o o o o o o o o ro ro ro ro ro ro CD CO CO CD CD fo fo fo fo fo O o o o o O o o o o ro ro ro ro ro o O CD" o a &) CD co cn 4^ co ro ro co cn 4^ co —»• co cn -t^  co ro o O T) "0 CO CO •JJ > cn cn 5 CO i— f—K CO CD CD ro ro <p_ o a oo ro ro ro s z o i z a) > b > «-» CO CO ro ro io ro CD cn co cn cn cn cn co CO CD 00 ro ro ro ro ro coco 4 i O l Ol O l O O l O oo b bi b bi fo 4^ - b CO SL ini p» *< P - w T3 >>>>>> ro co 4^ 4^ CO ro co 4^ co co ro ro cn ro co CD ^ o CO 4^ CD ro CD co 00 o — ^ — X o ^1 ro 00 CO — * -4 oo ro CO CO CO CO b -4 4^ 4^ b b CO b b b b b b b b oo b CD cn o cn CO — » • 4^ co ro CO 4^ 4^ ro CO >>>>>> ro ro ro ro ^ oo oi a> CD co cn b b 4^ fo b b ro ro ro ro oi ^ oi 4^ co co b b co ro ro co ro co CD b b o b b ^ ° References A t a i e - A s h t i a n i , B . , R . E . V o l k e r and D . A . L o c k i n g t o n , 1999. T i d a l effects o n sea water int rusion i n unconfmed aquifers. J. Hydrology 216, 17-31. Burnett , W . C . , M . Tan iguch i and J . A . Oberdorfer, 2001 . Measurement and signif icance o f the direct discharge o f groundwater into the coastal zone. Journal of Sea Research 46 , 109-116. Burnett , W . C . , J . Chanton, J . Christoff , E . Kon ta r , S. K r u p a , M . Lamber t , W . M o o r e , D . O 'Rourke , R . Paulsen, C . Smi th , L . Smi th , and M . Tan iguch i , 2002. Asse s s ing methodologies for measur ing groundwater discharge to the ocean. EOS 83 , 117-123. B o k u n i e w i c z , H . , 1980. Groundwater Seepage into Great South B a y , N e w Y o r k . Estuarine and Coastal Marine Science 10, 437-444. Cable , J . E . , W . C . Burnett , J .P . Chan ton and G . L . Weather ly , 1996. Es t ima t ing groundwater discharge into the northeastern G u l f o f M e x i c o us ing radon-222. Earth Planet. Sci. Let. 144, 591-604. Cab le , J . E . , W . C . Burnet t and J .P . Chanton, 1997a. M a g n i t u d e and variat ions o f groundwater seepage a long a F l o r i d a marine shoreline. Biogeochemistry 38, 189-205. Cab le , J . E . , W . C . Burnett , J .P . Chanton, D . R . Corbett and P . H . Cab l e , 1997b. F i e l d Eva lua t ion o f Seepage Mete rs i n the Coas ta l M a r i n e Envi ronment . Estuarine, Coastal and Shelf Science 45 , 367-375. D o m e n i c o , P . A . and F . W . Schwar tz , 1998. P h y s i c a l and C h e m i c a l H y d r o g e o l o g y . J o h n W i l e y and Sons, N e w Y o r k . Fang , W . W . , M . G . Langseth, P . J . Schultheiss, 1993. A n a l y s i s and appl ica t ion o f i n situ pore pressure measurements i n marine sediments. Journal of Geophysical Research 98 (B5) , 7921-7938. Findlatter , L . L . , 2000. A n a l y s i s o f Groundwater F l o w i n a Coas ta l A q u i f e r , T u r k e y Poin t , F l o r i d a . B . S c . Honours Thesis , Dept . o f Ear th and Ocean Sciences, U n i v e r s i t y o f B r i t i s h C o l u m b i a . H v o r s l e v , M . J . , 1951. T i m e lag and s o i l permeabi l i ty i n ground water observations. U . S . A r m y Corps o f Engr . W a t e r w a y E x p . Stat. B u l l . 36, V i c k s b u r g , M S . K e l l y , S. E . and L . C . M u r d o c h , 2002. M e a s u r i n g the H y d r a u l i c C o n d u c t i v i t y o f S h a l l o w Submerged Sediments. Ground Water 41 (4), 431-439. K o h o u t , F . A , 1966. Submarine springs: a neglected phenomenon o f coastal hydro logy . Hydrology 26, 391-413. K o h o u t , F . A . , 1960. C y c l i c f l ow o f salt water i n the B i s c a y n e A q u i f e r o f southeastern F l o r i d a . Jour. Geophys. Res. 65 , 2133-2141. L a n g e v i n , C D . , 2003. S imu la t i on o f submarine ground water discharge to a mar ine estuary: B i s c a y n e B a y , F l o r i d a . Ground Water 41 (6), 758-771. L i , L . , D . A . B a r r y , F . Stagnitt i and J . - Y . Parlange, 1999. Submar ine groundwater discharge and associated chemica l input to a coastal sea. Water Resources Research 35 (11), 3253-3259. M o o r e , W . S . , 1996. Large groundwater inputs to coastal water revealed b y 2 2 6 R a enrichments. Nature 380, 612-614. M o o r e , W . S . and T . M . C h u r c h , 1996. R e p l y to P . L . Y o u n g e r , Submar ine groundwater discharge. Nature 382, 121-122. M o o r e , W . S . , 2003 . Sources and fluxes o f submarine groundwater discharge delineated b y r ad ium isotopes. Biogeochemistry 66, 75-93. M i c h a e l , H . A . , J .S . Lube t sky and C . F . Ha rvey , 2003 . Charac te r iz ing submarine groundwater discharge: A seepage meter study i n Waquo i t B a y , Massachuset ts . Geophysical Research Letters 30 (6), 1297-1232. M i l l e r , J . A . , 1986. Hydrogeo log i c F ramework o f the F l o r i d a n A q u i f e r Sys t em i n F l o r i d a and Parts o f Georg ia , A l a b a m a and South Caro l ina . U S G S Prof . Paper 1403-B. Paulsen , R . J . , D . O ' R o u r k e , C . F . S m i t h , and T. W o n g , ( in press). Influence o f T i d a l L o a d i n g and Saltwater Intrusion on Submarine Groundwater Discharge . Ground Water 42 (7). Rasmussen, L . L . , 1998. Groundwater F l o w , T i d a l M i x i n g and H a l i n e C o n v e c t i o n i n Coas ta l Sediments. U n p u b l i s h e d M . S c . Thesis , F l o r i d a State U n i v e r s i t y , Dep t . o f Oceanography, 118 p. Schmidt , W . , 1984. N iogene stratigraphy and geologic his tory o f the A p a l a c h i c o l a Embayment , F l o r i d a ; F l o r i d a G e o l o g i c a l Survey, B u l l e t i n no. 58, 146 p . S C O R ( W o r k i n g Group 112) website ( A p r i l 2004): h t tp : / /www.jhu .edu/~scor /wgl 12front.htm Scott, T . M . , 1992. A geologic ove rv iew o f F l o r i d a . In G . L . M a d d o x , J . M . L l o y d and T . M . Scott (Eds.) , F l o r i d a ' s G r o u n d water qua l i ty mon i to r ing program: background hydro geochemistry. F l o r i d a G e o l o g i c a l Survey Spec ia l P u b l i c a t i o n no. 32, pp. 4-20. \ 130 Smi th , L . and W . Z a w a d z k i , 2003. A hydrogeologic m o d e l o f submarine groundwater discharge: F l o r i d a in tercomparison experiment. Biogeochemistry 66, 95-110. Tan iguch i , M . and Y . F u k u o , 1993. Cont inuous Measurements o f G r o u n d - W a t e r Seepage U s i n g an A u t o m a t i c Seepage Mete r . Ground Water 31 (4), 675-679. Tan iguch i , M . , W . C . Burnett , C . F . Smi th , R . J . Paulsen, D . O ' R o u r k e , S . L . K r u p a and J . L . Christoff , 2003 . Spat ia l and temporal distributions o f submarine groundwater discharge rates obtained f rom various types o f seepage meters at a site i n the Northeastern G u l f o f M e x i c o . Biogeochemistry 66, 35-53. Therr ien, R . , E . A . S u d i c k y , and R . G . M c L a r e n , 2003 . F R A C 3 D V S : A n Eff ic ien t S imula tor for Three-dimensional , Saturated-Unsaturated Groundwater F l o w and Dens i ty-dependent, Cha inODecay Solute Transport i n Porous , Discree t ly-Frac ture Porous or D u a l -poros i ty Format ions . U n p u b l i s h e d U s e r ' s G u i d e , Groundwater S imula t ions G r o u p . U c h i y a m a , Y . , K . Nadaoka , P . R o l k e , K . A d a c h i , and H . Y a g i , 2000 . Submar ine groundwater discharge into the sea and associated nutrient transport i n a sandy beach. Water Resources Research 36 (6), 1467-1479. Y i m , C . S . , and M . F . N . M o h s e n , 1992. S imu la t i on o f T i d a l Effects o n Contaminate Transport i n Porous M e d i a . Ground Water 30(1), 78-86. 131 Biographical Sketch Joshua L e e C a u l k i n s was born i n Pa lo A l t o , Ca l i fo rn i a , i n 1976 and attended the U n i v e r s i t y o f C a l i f o r n i a at Santa C r u z . H e completed a B a c h e l o r o f Sc ience i n Ear th Science f rom U C S C i n 1998, rece iv ing the W e b e r / H o l t A w a r d for A c a d e m i c Exce l l ence w i t h i n the Ear th Sciences upon his graduation. H e spent one year do ing graduate w o r k i n Geographic Information Science and Remote Sens ing at Hunte r C o l l e g e , C i t y U n i v e r s i t y o f N e w Y o r k , and interned for a summer w i t h the U S A r m y Corps o f Engineers , Coas ta l and H y d r a u l i c s Laboratory, Waterways Exper iment Stat ion i n V i c k s b u r g , M i s s i s s i p p i , us ing G I S and aerial photographs to m o d e l sediment transport i n the G u l f o f M e x i c o . D u r i n g the 2001-2002 academic year at U B C , Joshua was presented w i t h the "Outs tanding Teach ing Assis tant A w a r d " w i t h i n the Department o f Ea r th and Ocean Sciences. S o m e o f Joshua 's non-academic t ime at U B C was dedicated to the Graduate Student Soc ie ty as the Vice-Pres iden t A c a d e m i c and Exte rna l , elected to that pos i t ion b y the graduate student body o f 7,000 students for two years i n a r o w . Joshua is currently w o r k i n g on his P h . D . i n Geography at the U n i v e r s i t y o f B r i t i s h C o l u m b i a , focusing his studies on f luv ia l geomorphology. 132 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0052755/manifest

Comment

Related Items