UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The construction of a feedback seismograph station and an analysis of the Long Shot data from the Canadian… Jensen, Oliver George 1966

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

Item Metadata

Download

Media
831-UBC_1966_A6_7 J4.pdf [ 4.74MB ]
Metadata
JSON: 831-1.0053376.json
JSON-LD: 831-1.0053376-ld.json
RDF/XML (Pretty): 831-1.0053376-rdf.xml
RDF/JSON: 831-1.0053376-rdf.json
Turtle: 831-1.0053376-turtle.txt
N-Triples: 831-1.0053376-rdf-ntriples.txt
Original Record: 831-1.0053376-source.json
Full Text
831-1.0053376-fulltext.txt
Citation
831-1.0053376.ris

Full Text

THE CONSTRUCTION OP A FEEDBACK SEISMOGRAPH STATION AND AN ANALYSIS OP THE LONG SHOT DATA PROM THE CANADIAN SEISMOGRAPH STATIONS by OLIVER GEORGE JENSEN B.Sc. University of B r i t i s h Columbia, 1964 A THESIS SUBMITTED IN PARTIAL FULFILMENT OP THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE i n the Department of Geophysics We accept t h i s thesis as conforming to the required standard University of B r i t i s h Columbia June, 1966 In presentmg th i s thesis i n p a r t i a l f u l f i l m e n t of the requirements for an advanced degree at the University of B r i t i s h Columbia, I agree that the Library s h a l l make i t f r e e l y available f o r reference and study,, I further agree that permission f o r extensive copying of t h i s thesis f o r scholarly purposes may be granted by the Head of my Department or by his representatives. I t i s understood that copying or publication of t h i s thesis for f i n a n c i a l gain s h a l l not be allowed without my written permission. The Uni v e r s i t y of B r i t i s h Columbia Vancouver 8, Canada i i ABSTRACT A p r a c t i c a l and v e r s a t i l e feedback seismograph station has been constructed. Using feedback techniques developed by R.D. Meldrum, the r e l a t i v e low frequency response and the damping r a t i o of a Willmore Mk. I seismo-meter have been s i g n i f i c a n t l y increased f o r use i n a broad-band, low frequency bandpass seismograph station. Within l i m i t s imposed by a very high ambient ground noise l e v e l at the University of B r i t i s h Columbia s i t e , c i r c u i t noise and instrument amplifier c h a r a c t e r i s t i c s , i t i s possible to vary the damping r a t i o and resonant period through modification of the feedback loop transfer function. The seismograph has been continuously operating since November 1965 and has recorded over 40 l o c a l tremors and distant earthquakes from as f a r away as the mid Indian Ocean. It has shown that i t i s a useful demonstration and research instrument. A p i l o t analysis of the Long Shot nuclear explosion data received by the Canadian seismic stations indicates a consistent compressional f i r s t a r r i v a l as expected from an impulsive explosion source. S i g n i f i c a n t t r a v e l time discrepancies are observed i n the commencement of the P a r r i v a l which arrived up to 6 seconds early at a l l stations with the largest residuals at the most distant s i t e s . A comparison earthquake i n the Rat Islands area indicates a simi l a r bias trend. i l l The P a r r i v a l amplitudes appear to be anomalously low i n the central B.C. area and high i n eastern Canada. The e f f e c t i s also evident i n the u n i f i e d magnitude deter-minations which are based on these amplitudes. The causes of the variations of magnitudes and the anomalously low amplitudes have not been explained. The average magnitude and standard deviation f o r a l l Canadian stations Is shown to be 6.01 ±0.40 which agrees well with the world-wide average determination of 5-99 t 0.52. Spectral investigations demonstrate that there are both common and i n d i v i d u a l c h a r a c t e r i s t i c s among the ground amplitude spectra of the d i f f e r e n t stations. These c h a r a c t e r i s t i c s have not been correlated to the explosion source mechanism or to geological structure although some causative suggestions have been made. i PREFACE The present thesis i s concerned with two separate but related problems: the construction of a feedback seismograph and an analysis of the Long Shot data from the Canadian seismograph stations. The construction of a p r a c t i c a l seismograph using feedback p r i n c i p l e s developed by R.D. Meldrum and Dr. R.D. Russell of the Department of Geophysics, University of B r i t i s h Columbia i s discussed i n Part I. Upon completion of the instrument, an attempt was made to record seismic wave a r r i v a l s from the United States A i r Force Long Shot nuclear t e s t . The magnitude of the a r r i v a l s was found to be an order of magnitude lower than predicted and observed i n many other parts of Canada. This important f i n d i n g was confirmed by a systemmatic analysis of recordings of t h i s nuclear event from a l l Canadian seismograph stations. A report of t h i s analysis forms Part II of t h i s t h e s i s . An Analysis of the Long Shot Data From the  Canadian Seismograph Stations has also been published as a separate report of the A r c t i c I n s t i t u t e of North America. i v TABLE OP CONTENTS PART I THE CONSTRUCTION OP A FEEDBACK SEISMOGRAPH STATION CHAPTER I INTRODUCTION 1-1 The Seismograph Station 1 1-2 Preliminary Noise Studies 2 1-3 Summary of Meldrums Results 3 1- 4 The P r i n c i p l e of the Present Feedback Seismometer 4 CHAPTER II CONSTRUCTION OF THE FEEDBACK SEISMOGRAPH 2- 1 C a l i b r a t i o n and Bridge Design 9 2-2 The Ele c t r o n i c Construction of the Feedback Function 10 2-3 Amplifier Requirements 20 2-4 Design D e t a i l s 21 2-5 F i l t e r i n g and Recording Apparatus 25 2-6 Seismograph C a l i b r a t i o n 28 2-7 Limitations of the Feedback Seismograph System 30 2-8 Operational Testing of the Feedback Seismograph 33 CHAPTER I I I CONCLUSIONS 38 REFERENCES 40 PART II AN ANALYSIS OF THE LONG SHOT DATA FROM THE CANADIAN SEISMOGRAPH STATIONS CHAPTER I INTRODUCTION 1-1 Purpose of Long Shot 42 1-2 Present Understanding of Explosion Seismology 43 V 1- 3 Contributing Stations 46 CHAPTER II ANALYSIS OP LONG SHOT DATA-GENERAL 2- 1 Character of Records 50 2-2 R e l i a b i l i t y of Data 50 2-3 A r r i v a l Times and Residuals 51 2-4 F i r s t Motions 56 2- 5 Ground V e l o c i t i e s and Unified Magnitudes 60 CHAPTER III CHARACTER OF SIGNAL 3- 1 Special Records 70 3-2 A F i l t e r i n g Technique f o r Improved Character of the Wawa Records 73 3-3 Spectral Analysis 78 3-4 Results of Spectral Analysis 86 REFERENCES 87 APPENDIX I 90 LIST OP TABLES PART I TABLE I S t a t i s t i c a l Results of Test on Time Mark Generator 28 TABLE II Earthquakes Recorded by the University of B r i t i s h Columbia Seismograph Station 35 PART II TABLE I L i s t of Contributing Stations 48 TABLE II Long Shot Time Residuals 53 TABLE III Earthquake Time Residuals 54 TABLE IV Magnitude Determinations 63 v i i LIST OP FIGURES PART I FIGURE 1 A schematic diagram of the Maxwell impedance bridge method of applying feedback to an electromagnetic seismometer FIGURE 2 Block diagram of feedback seismometer FIGURE 3 Block diagram of an operational amplifier FIGURE 4 Basic amplifier c i r c u i t f o r the generation of the feedback function FIGURE 5 The gain response f o r the basic feedback c i r c u i t FIGURE 6 Modified amplifier c i r c u i t f o r the generation of the feedback function FIGURE 7 The gain response of the modified amplifier c i r c u i t FIGURE 8 The amplification stages of the feedback loop FIGURE 9 The gain response of the f i r s t amplifier stage FIGURE 10 The gain response of the second amplifier stage FIGURE 11a Schematic of the feedback seismometer FIGURE l i b Schematic of the f i l t e r i n g and recording network FIGURE 12 The acceleration s e n s i t i v i t i e s of the o r i g i n a l and feedback seismometers v i l i FIGURE 13 The gain response of the f i l t e r network 25 FIGURE 14 The v e l o c i t y s e n s i t i v i t y of the seismograph station 29 PART II FIGURE 1 The Canadian seismograph stations 47 FIGURE 2a Long Shot time residuals 55 FIGURE 2b March 30, 1965 Amchitka Island earthquake time residuals 55 FIGURE 3a F i r s t Motions f o r Canadian stations (I) 57 FIGURE 3b F i r s t Motions f o r Canadian stations (II) 58 FIGURE 3c F i r s t Motions f o r Canadian stations (III) 59 FIGURE 4 The Maximum observed peak to peak ground v e l o c i t i e s 6 l FIGURE 5 Un i f i e d magnitudes f o r Canadian stations 64 FIGURE 6 Azimuthal Plot of Unified Magnitudes 66 FIGURE 7 Differences between horizontal and v e r t i c a l magnitude determinations 68 FIGURE 8 Special Station C a l i b r a t i o n s (1) 71 Special station C a l i b r a t i o n s (II) 72 FIGURE 9 The frequency response of the Wawa deconvolution f i l t e r 76 FIGURE 10 Comparison of f i l t e r e d and u n f i l t e r e d Wawa records 77 i x FIGURE 11a Fourier amplitude spectra of P a r r i v a l s 83 FIGURE l i b Fourier amplitude spectra of P a r r i v a l s 84 FIGURE 12 Power spectra of the v e r t i c a l component of the ground v e l o c i t y 85 X ACKNOWLEDGEMENTS The author would l i k e to thank Dr. R.D. Russell fo r h i s invaluable aid i n the construction of the seismo-graph station. He i s also grateful to R.D. Meldrum f o r many valuable discussions about the project and to P. Michalow and T.A. Thome f o r assistance with some of the technical problems. The construction of the seismograph station i s a j o i n t endeavor of the Departments of Geophysics and Geology. The Socony Mobil O i l Company and the Observatories Branch, Department of Mines and Technical Surveys supplied the funds to obtain the equipment required f o r the construc-t i o n of the station. The author would l i k e to acknowledge the valuable assistance of Dr. K. Whitham of the Seismological Branch of the Dominion Observatory who suggested many of the investigations of the Long Shot report. Thanks are due to Drs. R.M. E l l i s and R.D. Russell who often aided i n solving technical d i f f i c u l t i e s and i n preparing the manuscript, and to Dr. J.A. Jacobs, Head of the Department of Geophysics at the University of B r i t i s h Columbia, f o r the use of departmental f a c i l i t i e s i n the Long Shot project. The author would also l i k e to thank P.W. Basham bf the A r c t i c I n s t i t u t e of North America and the s t a f f of the Seismological D i v i s i o n of the Department of Mines and Technical Surveys f o r the data used i n the project. x i The researches of the writer have been supported by the National Research Council of Canada with grants to Drs. R.D. Russell and J.A. Jacobs and by the A i r Force Office of S c i e n t i f i c Research, United States A i r Force, under AFOSR grant AF-AFOSR-1022-66 to the A r c t i c I n s t i t u t e of North America. PART I THE CONSTRUCTION OP A FEEDBACK SEISMOGRAPH STATION 1 CHAPTER I INTRODUCTION 1-1 The Seismograph Station A v e r s a t i l e , v e r t i c a l component seismograph station has been constructed f o r the University of B r i t i s h Columbia Department of Geophysics using negative feedback techniques which were developed by R.D. Meldrum 9 . The instrument was designed to f u l f i l l two main purposes: i t was intended to be a demonstration seismograph f o r the purpose of p u b l i c i z i n g seismology among students and the general public and to provide data f o r research i n Geo-physics. As a demonstration apparatus, the instrument was designed to record the major P (body compressional), S (body shear) and LQ and LR (low frequency Love and Rayleigh surface) waves on one record jio] . The required instrument c h a r a c t e r i s t i c s d i f f e r greatly from those of the standard Canadian seismograph stations operated by the Observatories Branch of the Department of Mines and Technical Surveys which record separately short period (0.2 to 1.0 second) and long period (15 to 80 seconds) disturbances. Periods between 1 and 15 seconds are highly attenuated by these stations i n order to attenuate microseism signals which have a peak amplitude at about 5 seconds period j^2J. Generally, the P a r r i v a l s are received by the short period instruments and the long period Love and Rayleigh surface waves are 2 received by the long period instruments. T. Matumoto [8J has shown that much of the energy of the S wavr a r r i v a l s occurs i n the periods attenuated by these separated instruments. Thus, i n order to obtain a single record showing the major a r r i v a l s , i t was required that the University of B r i t i s h Columbia station (hereafter c a l l e d U.B.C. station) should operate with longer periods than the short period Dominion Observatory instruments; a bandpass of 1 to 4 seconds period was chosen. As a research seismograph, Drs. W. Milne and H. White, of the Dominion Astrophysical Observatory, V i c t o r i a , had suggested that the station could be useful i n the general studies of seismicity on the west coast. The instrument ^as designed to provide general data f o r such studies and to give immediate a r r i v a l time and ground motion amplitude information f o r preliminary determinations of the energy and loc a t i o n of l o c a l events. Therefore, i t was desirable that the seismograph be equipped with a vi s u a l recording system rather than the normal photographic recorder. The feedback technique lent i t s e l f p a r t i c u l a r l y well to f u l f i l l m e n t of the requirements. 1-2 Preliminary Noise Studies Preliminary investigations of the seismic noise at the U.B.C. s i t e showed that, although the noise l e v e l at frequencies above 1 cycle per second i s much greater than on the V i c t o r i a seismograph station pier, the noise 3 l e v e l does not preclude the construction of a useful station. Measurements on the V i c t o r i a p i e r indicated a maximum peak to peak ground noise amplitude of 18 m i l l i -microns at 0.7 cycles per second and, i n the basement of the University of B.C. Chemical Engineering building, a maximum peak to peak ground amplitude of 130 millimicrons at 1.2 cycles per second. These measurements were made on successive microseismically quiet days. The lower frequency microseism noise (0.2 to 0.5 cycles per second) did not appear to be s i g n i f i c a n t l y d i f f e r e n t f o r the two si t e s , although at frequencies above 2 cycles per second, the noise l e v e l at the U.B.C. s i t e was several times higher. Since i t was intended to operate the instrument with a r e l a t i v e l y low frequency bandpass, the higher frequency noise did not pose immediate d i f f i c u l t i e s as i t could be attenuated by low pass electronic f i l t e r i n g . In an attempt to f i n d a quieter s i t e on the University of B.C. campus, noise studies were conducted i n the base-ment of the Forestry and Geology building. I t was found that a p r o h i b i t i v e l y high noise l e v e l at about 1 cycle per second was generated by student t r a f f i c . The background noise i n the Chemical Engineering b u i l d i n g was found to be of s i g n i f i c a n t l y lower amplitude and higher frequency which allowed easier separation of the noise from seismic signals. 1-3 Summary of Meldrum's r e s u l t s In the l i t e r a t u r e , many feedback seismograph 4 systems have been suggested f o r v a r i e t i e s of purposes jll] . However, most of these systems require complicated methods fo r adding the feedback signal to the input ground signals. Meldrum showed that a v e r s a t i l e seismometer could be con-structed using e l e c t r o n i c negative feedback applied d i r e c t l y into the transducer c o i l of an electromagnetic seismometer through a Maxwell impedance bridge. He showed that the e f f e c t i v e resonant period of the seismometer could be increased by feeding a simulated ground acceleration into the seismometer and that the damping could be increased by feeding i n a simulated ground v e l o c i t y . With a Willmore MK. I, one second resonant period seismometer, he obtained an e f f e c t i v e resonant period of 17 seconds using these techniques. He suggested that increased period and damping could be accomplished simultaneously by l i n e a r l y combining the acceleration and v e l o c i t y feedback signals. Meldrum's feedback techniques offered the required v e r s a t i l i t y f o r the construction of the U.B.C. station, and were used to obtain the desired 4 second resonant period from an available one second period Willmore Mark I seismometer. 1-4 The P r i n c i p l e of the Present Feedback Seismometer In Meldrum's feedback technique, the undamped seismometer i s placed i n the "unknown" p o s i t i o n of the bridge which has been previously balanced to remove the purely e l e c t r i c a l e f f e c t s of the transducer c o i l impedance. Willmore 13 has shown that, i f the bridge resistance Pig. 1. A schematic diagram of the Maxwell impedance bridge method fo r applying feedback to an electromagnetic seismometer. 6 and capacitance values are c o r r e c t l y chosen, any voltage signal applied to the inpmt terminals of the bridge produces the same e f f e c t on the seismometer as a ground acceleration of magnitude: . where v(j^) i s the applied voltage s i g n a l . In Figure 1, R R J R D J R B J C B are bridge impedances, 3 represents the seismometer and Vi i s the input feedback s i g n a l . In Meldrum's method f o r applying feedback, the seismometer i s continuously operated within the bridge In a v i r t u a l l y open c i r c u i t e d configuration. The output signal i s obtained across the bridge output terminals, amplified, modified and fed back into the bridge input terminals, producing a simulated ground acceleration on the seismometer proportional to the feedback s i g n a l . The equivalent block diagram f o r t h i s l i n e a r negative feedback system i s shown i n Figure 2. F i g . 2. A block diagram of the feedback seismometer. In the diagram, the parameter S represents the complex frequency 5 = <f+ju) . The transfer function of the seismometer i s represented by: _ _ OLS G ( s ) ~~ S a-+^fu)hS4oO n a-where a determines the maximum acceleration s e n s i t i v i t y , £ i s the damping r a t i o and U)n i s the resonant angular frequency of the seismometer. The feedback path transfer function Is represented by F ( s ) a n d the approximate transfer function of the Maxwell impedance bridge by j / f ^ g 9 • represents the current equivalent of the input ground acceleration signal and V ( s ) , the output voltage signal from the feedback seismometer due to the input ground motion signal [6J. The equivalent transfer function of the entire l i n e a r feedback system i s e a s i l y obtained: _ . I j / \ - Gi (s)  H C S j | + F (S )6 (S ) /R B The form of the feedback transfer function determines the e f f e c t on the seismometer c h a r a c t e r i s t i c s . Meldrum investigated feedback of the forms: 1/s, 1, and s, where s Is the complex frequency. These forms correspond to integration, amplification and d i f f e r e n t i a t i o n of the output signal from the seismometer respectively. He showed that f o r negative feedback, with these basic functions, the equivalent resonant period of the seismo-meter could be increased or decreased by more than an order of magnitude and the damping r a t i o could be increased to almost any desired value. Combinations of these e f f e c t s 8 could also be achieved by using l i n e a r combinations of the basic feedback functions. For the proposed station seismograph, an increase of the resonant period of an available Willmore Mk. I seismometer from 1 to 4 seconds and an increase of the damping r a t i o to c r i t i c a l was required. Meldrum's investigations indicated that the feedback transfer function: F ( s ^ = b + c s produced t h i s desired e f f e c t . Under negative feedback of t h i s form, the equivalent transfer function of the feedback seismometer can be e a s i l y determined using l i n e a r feedback techniques where: i a's T T CLC CL is a. b I _ PL I t i s notable that the maximum s e n s i t i v i t y of the instrument to ground accelerations i s decreased as the resonant period i s increased. 9 CHAPTER II CONSTRUCTION OP THE FEEDBACK SEISMOGRAPH 2-1 C a l i b r a t i o n and Bridge Design A Willmore Mk. I seismometer, obtained on loan from the Dominion Observatory, was used i n the development of the seismograph. C a l i b r a t i o n of t h i s p a r t i c u l a r seismometer was accomplished using Willmore's impedance bridge technique 13 . The following instrumental constants were obtained: a = 3030 v o l t sec kgm"^" 1 9 - 113.5 v o l t sec m - 1 2TT x 1.0 sec" 1 i - 0.037 c r i t i c a l damping Rc - 519.4 ohms L c - 170 m i l l i h e n r i e s The seismometer was operated within a Maxwell impedance bridge which had previously been balanced to compensate the e f f e c t s of the e l e c t r i c a l impedance of the transducer c o i l with respect to signals applied to the bridge input terminals. In t h i s configuration the following bridge constants were used: R B = 105.3K ohms Ro = 10.0K ohms RR = 49.2 ohms = 0.0 microfarads See Figure 1. 10 I t was found necessary to remove the capacitance Cg because the amplifiers used i n the feedback network could not d i r e c t l y drive a capacitive load. At the very low frequencies at which the instrument was designed to operate, the impedance due to c o i l inductance does not exceed 1$ of the resistance of the transducer c o i l and only very s l i g h t error i s introduced by the omission of the capacitance required to balance the inductance of the c o i l . 2-2 Ele c t r o n i c Construction of the Feedback Function For the required seismometer c h a r a c t e r i s t i c s of 4 second resonant period and c r i t i c a l damping, the feedback transfer function constants b and c were required to be about 1500 and 500 seconds respectively. To obtain t h i s function, operational amplifiers of the type used i n analogue computers were used. Operational amplifiers are very high gain amplifiers with a f l a t frequency response from 0 to 100 cycles per second possessing a constant phase s h i f t of l8o° and a single corner frequency above which the gain decreases assymptotically at 6 decibels per octave. By applying external Input and feedback impedances to such amplifiers, many desireable mathematical transfer functions can be obtained e l e c t r o n i c a l l y -O F i g . 3. Block diagram of a closed loop operational amplifier. 11 The block diagram of Figure 3 represents the amplifier i n i t s closed loop form; that i s with input and feedback response of the operational amplifier. I t i s e a s i l y shown, that f o r very high gain amplifiers, the r a t i o between the input and output voltages approaches 1 : Thus, the transfer function of the closed loop amplifier depends only on the chosen external impedances. Various operations and transfer functions can be produced by proper choices of the external impedances; the basic available operations are: M u l t i p l i c a t i o n (amplification) H(s) = -a Integration (low pass f i l t e r ) = -1/s DIfferentation (high pass f i l t e r ) = -s Addition (summing) = - Sum of input These operations are possible using only r e s i s t o r s and capacitors f o r the external impedances; combinations of these operations are also possible. impedances applied. and H;t a r e input and feedback impedances respectively and-(Si(s) i s the complex frequency signals The basic transfer function for the feedback seismograph was obtained using the operational amplifier configuration shown i n Figure 4. 12 o c, A A / W V Rsi w v w - o o - - o P i g . 4. The basic c i r c u i t f o r the generation of the feedback function. The transfer function of t h i s configuration i s : where: H(s) = - K , ( l + T s ) The frequency response f o r t h i s transfer function i s plotted i n Figure 5. P r a c t i c a l l y , t h i s i d e a l transfer function was s l i g h t l y modified to reduce the amplification of high frequency ground and c i r c u i t noise. This was accomplished by placing a small capacitance i n p a r a l l e l with the resistance as shown i n Figure 6. 13 Pig. 6. Modified c i r c u i t f o r the generation of the feedback function. 14 The modified transfer function f o r t h i s c i r c u i t i s : H,(s ) - t \ i ( | + l i s ) where and ~~[~| remain the same and ~fa_= Ra,C^. The impedances used i n the p r a c t i c a l construction of t h i s stage were: RI = 6.7 Meg ohm C | = 0.05 microfarad r\£ = 10.0 Meg ohm Ca.= 0.0015 microfarad which give: K,- 1.5 T i = °-34 seconds = 0.015 seconds It i s necessary that "T^,«Ti s o that the feedback transfer function maintains the form b + C S t o s u f f i c i e n t l y high frequencies. The cutoff corner frequency, $x~~: /^-^Tx of t h i s stage i s 10 cycles per second. The frequency response of t h i s modified c i r c u i t i s shown i n Figure 7. Using t h i s single amplifier stage, i t was not possible to achieve the required gain of about 1,000 to give the required feedback function constants and two stages of further a m p l i f i c a t i o n of the feedback signals were required. The schematic of these further a m p l i f i c a t i o n stages Is shown i n Figure 8. R3 O A / W W R 3 O V W W Pig. 8. The amplification stages of the feedback loop. 16 50 20 H 10 •a! 5.0 2.0 / 1 — 1.0 0.1 0.2 0.5 1.0 2.0 5.0 10 20 FREQUENCY (CYCLES/SECOND) F i g . 7« The gain response f o r the modified feedback function c i r c u i t . 500 200 100 53 H g '50 .20 10 50 100 0.1 0.2 0.5 1.0 2.0 5.0 10 20 50 100 FREQUENCY (CYCLES/SECOND) F i g . 9. The gain response of the f i r s t amplifier stage. 17 In the f i r s t a d ditional amplifier stage, a d i f f e r e n t i a l input was required to i s o l a t e the bridge output terminals and to allow grounding of one bridge input terminal to accomodate the use of single ended output amplifiers. The transfer function of the f i r s t amplifier stage i s : where |^\^ = and = R-fOf- • T n e impedances used i n p r a c t i c a l construction of t h i s stage were: R 3 = 100 K ohms = 10 M ohms Qq. = 0.0015 microfarads to give: = 0.015 seconds The frequency response of the stage Is shown i n Figure 9. The capacitance C4 i n p a r a l l e l with resistance Rij was again required f o r the attenuation of high frequency noise; the f i l t e r cutoff corner frequency was chosen well outside the designed bandpass of the seismograph station. The f i n a l amplifier stage was capacity coupled with capacitance £5" to eliminate d i r e c t current and reduce low frequencies i n the network. The transfer function of t h i s stage i s : H ^ ( s ) = (| + T^(|+~&s) 18 where: Ti = RcCs This stage attenuates low frequencies below is and high frequencies above ^ = J/^J]~ "J^ • The attenuation was chosen to be well separated from the required bandpass of the feedback seismometer and between the two corner frequencies, the amplifier gain i s r e l a t i v e l y constant: The impedances used i n the construction of the f i n a l a mplifier stage were: = 1 M ohm R4 = 10 M ohm G 5 * = 2.0 microfarads = 0.006 microfarads which give: "TV = 2 seconds "T& = 0.06 seconds Ti -<3 20 seconds The o v e r a l l frequency response of t h i s amplifier i s shown i n Figure 10. 19 0.05 0.1 0.2 0.5 1.0 2.0 5.0 10 20 FREQUENCY (CYCLES/SECOND) F i g . 10. The gain response of the second amplifier stage The t o t a l feedback network transfer function can be represented by the simple form: F T ( S ) = Fes) H(s) where F(s) has the required form of the feedback function: F(s) =-K,KaK30+Ts) and H(s) represents the noise and d r i f t f i l t e r i n g properties inherent In the designed feedback network: H(s) T7S (l+Tasyi+T^)(l+-&i)(|+T<s) 20 The required constants b and C of the feedback transfer function, P(s) — b+CS > a r e related to the c i r c u i t constants: b = K| KSLK3 = 1500 = 500 seconds 2-3 Amplifier Requirements Tests conducted on seismically quiet days indicated that very low l e v e l signals are obtained from the seismo-meter i n the frequency band of 0.1 to 1.0 cycles per second due to background ground noise. During periods of very low noise and microseism, the signal from the seismometer can be l e s s than 10 microvolts peak to peak maximum i n t h i s bandwidth. Thus, i n order that seismometer s e n s i t i v i t y i s not l i m i t e d by the feedback loop electronic noise, i t was necessary to use a f i r s t stage amplifier with an input noise l e v e l below that generated by the background noise. A P h i l b r i c k Researches P25A s o l i d state operational amplifier which has a t y p i c a l input noise l e v e l of 5 microvolts peak to peak i n the bandwidth 0.016 to 1.6 cycles per second was adequately used f o r t h i s f i r s t stage and operated with a gain of 100. Because the signal l e v e l i s greatly amplified by t h i s stage, the input noise l e v e l requirements of successive stages i s not so c r i t i c a l and higher noise, l e s s expensive P h i l b r i c k P65AU amplifiers were used i n these stages. The operational amplifiers were operated within 21 a P h i l b r i c k Model MP Operational Manifold which accepted standard p i n jacks and double pin plugs on which the r e s i s t o r s and capacitors required f o r the c i r c u i t construc-t i o n were mounted. A well regulated (0.05$ l i n e and load variation) balanced +15 v o l t , -15 v o l t power supply was provided with the manifold to power the operational amplifiers. The P h i l b r i c k Operational Manifold with the associated amplifiers and power supply was chosen because i t provided the convenient patchboard f o r c i r c u i t construction and was r e l a t i v e l y compact. However, amplifiers of other manufacture should be usable with equal seccess providing the minimal noise requirements of the f i r s t a mplification stage are met. Low noise carbon deposit p r e c i s i o n r e s i s t o r s with ± 1% tolerances and stable, low leakage, mylar d i e l e c t r i c capacitors with - 10$ tolerances were used i n order to minimize the thermal noise generated by the external c i r c u i t impedances. 2=!k Design D e t a i l s The feedback technique was intended to decrease the e f f e c t i v e resonant frequency of the Willmore Mk.I seismometer from 1.0 to about 0.25 cycles per second and to Increase the damping r a t i o from 0.037 c r i t i c a l to unity. For t h i s purpose, the feedback function was designed with constants: b = 1500 C = 500 seconds SCHEMATIC OP THE FEEDBACK SEISMOMETER 1.5uuF 1.5uuF 2 UF 1M | ( — V W -To F i l t e r Network 1.5uuF 10M e a l i b r a t i o n C > A / W l Input 5HK O V\AA o -FILTER AND RECORDING NETWORK 68K 68K 5uP Helicorder Amplifier -r-2uP Recorder Galvonometer Pig. U f a} Schematic diagram of the feedback" seismometer. 11(b) Schematic diagram of the f i l t e r i n g and recording network. 23 Prom the t h e o r e t i c a l formulae, the precise natural frequency and damping r a t i o expected f o r the feedback seismometer under t h i s feedback function are e a s i l y calculated: bridge technique was used f o r the d i r e c t determination of the instrument response to ground accelerations by applying sinusoidal signals to the input terminals of the impedance bridge. Low frequency signals of known voltage (less than 5 v o l t s peak to peak) were applied to the c a l i b r a t i o n input of amplifier #3 (Figure 11a) of the operational manifold. The input signals were incremented i n frequency by approximately one octave steps from 0.05 to 1.0 cycles per second. An oscilloscope was connected to the output of amplifier #2 to measure the response of the Instrument to the input signals. For each frequency, the r a t i o of the output to the input signal was calculated; these r a t i o s were plotted against frequency on log-log scales to graph the r e l a t i v e acceleration s e n s i t i v i t y of the seismometer. To obtain the absolute acceleration s e n s i t i v i t y , each r a t i o The c a l i b r a t i o n acceleration s e n s i t i v i t y curves are compared fo r the o r i g i n a l and feedback seismometers i n Figure 12. = 0.26 cycles second -= O.87 c r i t i c a l damping A c a l i b r a t i o n procedure based on Willmore's -1 24 100. Eh H Eh H CO § 1 0 . 0 CO H E-t K W o o < 1.0 < 1 \v 0.1 FREQUENCY 1.0 (CYCLES/SECOND) 10.0 F i g . 12. The acceleration s e n s i t i v i t i e s of the o r i g i n a l and feedback seismometers. The capacitor (Figure 11a) e s s e n t i a l l y determines the value of the feedback constant c and the re s u l t i n g resonant frequency. The 15$ discrepancy between the observed and calculated values of the resonant frequency of the feedback instrument i s probably a r e s u l t of the low tolerances i n t h i s capacitor value. The constant b and hence the damping r a t i o i s la r g e l y determined by the 25 c i r c u i t resistances; the very s l i g h t error of le s s than 1% i n the experimentally determined damping r a t i o i s well within that expected due to the possible errors i n the pre c i s i o n r e s i s t o r s . 2-5 F i l t e r i n g and Recording Apparatus Because an amplifier-recorder system rather than a galvonometer-recorder system was used, the low pass f i l t e r i n g properties of the galvonometer were l o s t and electronic f i l t e r i n g was required to attenuate unwanted high frequency signals. A l i n e a r f i l t e r - a m p l i f i e r of three low pass and one high pass stages with corner frequencies of 0.03 and 1.0 cycles per second and an o v e r a l l gain of 26 1.5 was constructed to eliminate low frequency noise and d r i f t and to reduce the high frequency noise. (Figure l i b . ) The transfer function of the f i l t e r i n g network i s : — G H ( S ) ~ ( l ^ s ) ( l + T 5 S f ( l + T 3 s ) where the gain Q = 1.5, T| = 5 seconds, ~~[x_ = 0.014 seconds, and ~]~3 = 0.015 seconds. The f i l t e r amplitude frequency response i s shown i n Figure 13. The very severe f i l t e r i n g above 1 cycle per second was made necessary by a high ground noise l e v e l generated by t r a f f i c , students and machinery i n the v i c i n i t y of the seismograph s i t e . Most of the noise was found to occur above 2 cycles per second and during noisy daytime periods, the signal l e v e l s generated by the noise approaches 1 m i l l i v o l t peak to peak from the seismometer or about 100 times greater than the noise background i n the operating bandpass of the feedback instrument. In the seismograph, the seismic signal i s obtained from the second amplifier stage of the feedback loop, separated from the noise by the f i l t e r network, further amplified by a Geotech Helicorder model 4983 D.C. amplifier and recorded by a Helicorder model 2484 recorder. The recorder produces a continuous h e l i c a l record of 8 to 72 hours variable duration with chart speeds of 10 to 90 millimeters per minute on sensitized paper with a heated stylus. The amplifier provides a maximum gain of 500 and a maximum zero to peak output of 27 150 v o l t s . An incremental attenuator with 6 decibel (factor of 2) steps i s provided i n the Helicorder amplifier f o r stepwise l e v e l adjustments. The heated stylus of the recorder i s deflected by a galvanometer with a d e f l e c t i o n s e n s i t i v i t y of about 50 v o l t s per inch and a maximum zero to peak d e f l e c t i o n of 3 inches. This d e f l e c t i o n , however, was l i m i t e d to about 3/4 inches by placing a p a i r of 30 vo l t zener diodes across the input terminals of the galvanometer to prevent damage to the stylus on large d e f l e c t i o n s . The timing of the seismograph station records i s presently accomplished by a synchronous time-mark generator constructed from a 60 cycle per second synchronous motor which operates from the power l i n e . The timing impulses are obtained from a photocell which i s separated from a collimated l i g h t source by an opaque disc turned by the synchronous motor. A sl o t cut i n the disc allows the collimated l i g h t beam to st r i k e the photocell f o r a period of two seconds each minute. Extraneous l i g h t i s prevented from a f f e c t i n g the c e l l by placing the c e l l , motor and source l i g h t within a l i g h t t i g h t box. The timing impulses are added d i r e c t l y to the seismic signal at the second tube stage of the Helicorder amplifier. Power l i n e frequency i n Vancouver i s maintained to phase errors of les s than 0.3 seconds by the B r i t i s h Columbia Power and Hydro Authority. This accuracy i s s u f f i c i e n t f o r the present station timing requirements. Tests of accuracy were conducted on the time 28 mark generator to determine the accuracy with which i t follows the l i n e frequency. An accurate c r y s t a l controlled Aptimeter frequency counter was used to count the number of cycles of the power l i n e voltage during one minute as measured by the time mark generator. The generator was used d i r e c t l y to gate the counter; that i s , to turn i t on and then o f f one revolution of the disc l a t e r . The s t a t i s -t i c a l r e s u l t s of 100 such observations i s shown i n Table I. TABLE I Number of Cycles Number of Times Counted 3598 and fewer 0 3599 5 3600 91 3601 3 3602 1 3603 and more 0 Average 36OO cycles (1 minute); Standard deviation: 0.3^ cycles (0.0057 seconds) These tests on the time mark generator using the period meter and subsequent operation have shown that the timing i s accurate within a few tenths of a second maximum error from Universal Time barring power f a i l u r e s . However, i t i s intended to acquire a more accurate c r y s t a l chronometer fo r c r i t i c a l timing. 2-6 Seismograph C a l i b r a t i o n The o v e r a l l response of the f i l t e r s , Helicorder amplifier and recorder was determined by applying low l e v e l 29 5000 i | 2000 1000 o o m CO EH s > 500 200 EH H h 100 EH H CO S3 W CO EH 50 o o 20 10 r^*" 1 \ / i 1} 1/ // // IJ IT \ \ V t \ • \ \ \ • \ / / / • \ 0.02 0.05 0.1 0.2 0.5 1.0 2.0 5.0 FREQUENCY (CYCLES/SECOND) F i g . 14. The maximum v e l o c i t y s e n s i t i v i t y of the seismograph (0 attenuation). sinusoidal signals to the input of the f i r s t stage of the f i l t e r network. The response i s shown i n Figure 13. This response was, then, m u l t i p l i e d by the v e l o c i t y s e n s i t i v i t y of the feedback seismometer, (determined by multiplying the acceleration s e n s i t i v i t y by the angular frequency f o r each point on the graph) to obtain the o v e r a l l seismograph v e l o c i t y s e n s i t i v i t y (0 attenuation on Helicorder amplifier) i s shown i n Figure 14. The half power points of the 30 instrument bandpass are approximately 0.1 and 1.0 cycles per second. Below 0.2 cycles per second, the feedback seismometer attenuates assymptotically at 12 decibels per octave and above 1 cycle per second, the f i l t e r network attenuates assymptotically at 18 decibels per octave. 2-7 Limitations of the Feedback Seismograph System It i s possible f o r closed loop negative feedback systems to become unstable and o s c i l l a t e without bound 1 £7] • This i n s t a b i l i t y occurs when the phase s h i f t of the feedback loop reaches ± 180 degrees, which r e s u l t s i n the feedback signal adding to rather than subtracting from the input. Mathematically, t h i s condition can be seen by considering the closed loop transfer function: H , , - F ( s ) At some complex frequency, i t i s conceivable that the denominator may vanish and produce indeterminate responses to any input s i g n a l . For the feedback seismograph, a maximum phase lead of l 8 o ° cannot quite occur under feedback of the form considered and the system remains unconditionally stable. In fa c t , the e f f e c t of applying the feedback s t a b i l i z e s the system, providing the feedback i s so applied that s u f f i c i e n t damping i s achieved. P r a c t i c a l l y , however, other e f f e c t s such as low l e v e l signals, amplifier saturation, and high c i r c u i t and ground noise do severly l i m i t the amount of feedback which can be used. 31 The low l e v e l seismic signals are separated from the ground noise by the frequency f i l t e r i n g c h a r a c t e r i s t i c s of the seismograph system. It has already been mentioned that on days of low background noise, the signals from the seismometer are les s than 10 microvolts peak to peak i n the instrument bandpass. Since i t i s desirable that the instrument electronic noise does not l i m i t the ultimate s e n s i t i v i t y of the instrument, the low l e v e l signals necessitate the use of a high q u a l i t y low noise amplifier i n the f i r s t stage of the feedback loop. At low frequencies, below 1 cycle per second, the v e l o c i t y s e n s i t i v i t y of the Willmore Mk. I seismometer decreases asymptotically at the rate of 12 decibels per octave and very low frequency signals are highly attenuated. Consequently, the e f f e c t i v e instrument resonant frequency cannot be i n d e f i n i t e l y moved toward lower frequencies unless a very sophisticated low noise i n i t i a l stage amplifier i s used. Otherwise, the instrument noise may well exceed the low frequency back-ground noise signal l e v e l s and even moderate seismic event signals. In the present feedback seismograph the input noise l e v e l of the f i r s t amplifier stage i s reasonable l e s s than the l i m i t i n g ground noise l e v e l s i n the instrument bandpass and does not l i m i t the ultimate s e n s i t i v i t y . A further d i f f i c u l t y a r ises from the saturation of the present low output voltage t r a n s i s t o r operational amplifiers which can d e l i v e r maximum signals of only ± 11 v o l t s . When an operational amplifier i s driven to signal 32 outputs which exceed t h i s maximum voltage, saturation occurs and the external impedances no longer determine the transfer function of the committed amplifier. A few seconds i s usually required f o r the saturated amplifier to recover and restore normal operation a f t e r the removal of the input signal causing the saturation. The e f f e c t causes a complete breakdown of the feedback loop function; about 10 seconds i s required f o r the recovery of the feedback operation. The form of the feedback required to obtain the desired instrument response causes greater amplification of high frequencies. This i s unfortunate because of the very high noise l e v e l above 1 cycle per second at the U.B.C. s i t e . Careful f i l t e r i n g within the feedback loop i s necessary to attenuate the noise without disturbing the operation of the feedback loop. But, even with t h i s i n t e r n a l feedback loop f i l t e r i n g , transient ground noise signals may drive the f i n a l stage amplifier to saturation. A reduction i n the ef f e c t i v e instrument resonant frequency requires an increase i n the gain of the feedback loop. Consequently, the possi-b i l i t y of amplifier saturation due to more highly amplified noise transients i s enhanced. This saturation d i f f i c u l t y l i m i t s the amount of feedback which may be used under the U.B.C. s i t e conditions with the present low voltage amplifiers. The d i f f i c u l t y could be circumvented by using a higher voltage f i n a l amplifier stage or by reducing the value of the balancing bridge r e s i s t o r Rg (Figure 1). The l a t t e r alternative causes a p a r t i a l breakdown of the ground motion 33 simulation c h a r a c t e r i s t i c s of the bridge and a r e s u l t i n g i n s t a b i l i t y [13J . Rather, i t i s suggested that a f i n a l amplifier stage, capable of d e l i v e r i n g ± 1 0 0 v o l t s to the Maxwell bridge, would eliminate or reduce the saturation problem. The t r a n s i s t o r amplifiers used i n the present feedback seismograph are very sensitive to temperature variations which cause low frequency noise and d r i f t . The ef f e c t Is a re s u l t of temperature induced variati o n s of the base currents and base voltages of the input t r a n s i s t o r s i n the amplifiers. There i s some d i f f i c u l t y i n separating low frequency signals from t h i s temperature induced noise and d r i f t which prevents unlimited reduction of the resonant frequency. For good s t a b i l i t y and low noise at low frequen-cies, chopper s t a b i l i z e d amplifiers are required. 2-8 Operational Testing of the Feedback Seismograph Upon completion of the seismograph, a procedure of t e s t i n g was i n s t i t u t e d . Since the U.B.C. Department of Geophysics was to undertake analysis of the Long Shot, Long Range Seismic Measurements Experiment, seismic records f o r a l l Canadian seismograph stations, an e f f o r t was made to record the seismic a r r i v a l s from the event with the newly constructed instrument. A s i t e was chosen at the Williams Lake, B r i t i s h Columbia airport to monitor the expected a r r i v a l s . No record was obtained. Following t h i s apparent f a i l u r e of the instrument, a procedure of further t e s t i n g 34 and r e c a l i b r a t l o n of the seismograph showed i t to be i n perfect operating condition. Subsequent analysis of the Long Shot data f o r the Canadian stations indicated that anomalously low signal amplitudes were received i n the general area of the Williams Lake s i t e . The apparent signal l e v e l was at least one order of magnitude lower than anticipated and below the s e n s i t i v i t y l e v e l of the seismograph. Since the Long Shot experiment, the feedback seismograph has been continuously operating at the U.B.C. si t e and more than 40 l o c a l and distant earthquakes have been recorded. The instrument has been sensitive to l o c a l events of magnitude 3 on the Gutenburg-Richter u n i f i e d scale [lo] and to magnitude 5 events at teleseismic distances. Records of large events, approaching u n i f i e d magnitude 7, have shown several d i s t i n c t seismic wave a r r i v a l s ; the major compressional, shear and surface waves to about 20 seconds period are usually discernable on the records of the largest earthquakes. Table II l i s t s the major events recorded i n the f i r s t quarter of 1966. The event i d e n t i f i c a t i o n and magnitudes were obtained from the U.S. Coast and Geodetic Survey Seismological B u l l e t i n s 12 . 35 TABLE II EARTHQUAKES RECORDED BY THE UNIVERSITY OP BRITISH COLUMBIA SEISMOGRAPH STATION Date Time (U.S.C.G.S.) Magnitude (U.S.C.G.S.) Location (U.S.C.G.S.) November 22, 1965 20:25:30.4 5.9 Andreanof Islands 23 2:17:49.2 5.6 Andreanof Islands December 3, 1965 (15:16:22.6) _ _ _ Unidentified (about 2300 km. distant) 6 11:34:53.7 5.9 J a l i s c o , Mexico 9 6:07:48.6 6.0 Guerrero, Mexico 15 23:05:20.7 6.0 South of Panama 16 (17:19:25) Unidentified 22 (22:02) ___ Unidentified 22 19:41:23.0 6.5 Kenai penninsula, Alaska 23 20:47:37.5 5.4 South eastern Alaska January 12, 1966 (7:19:45) Unidentified (possibly l o c a l ) 20 19:51:26 4.2 Queen Charlotte Is. 22 14:27:07.9 5 .8 South of Alaska February 7, 1966 8:48:34.5 4.5 Queen Charlotte I s. 7 14:03:04 3.7 Vancouver I s. 10 14:21:10.9 6.2 Mariana Is. 13 4:57:57.7 6.3 Eastern Kazakh, USSR (Russian underground nuclear test) 36 TABLE II Continued: Date Time Magnitude Location February 16 3:18:27.2 6.5 New Hebrides 17 11 :48:00.8 6.4 Mid-Indian Rise 22 5:02:37.2 6.2 New B r i t a i n region 26 0:41:12 5.3 Near Is., Aleutians March 7, 1966 18:09:43.6 4.7 Montana 12 16:31:21.8 6.7 Taiwan 22 8:19:33.8 6.0 North Eastern China 29 2:17:38.5 5.9 Volcano Is. region 30 12 :40:01.0 5.3 Vancouver Is. region A p r i l 7, 1966 9:42:32.1 5.7 Ryukyu Islands 8 22:10:59.3 5.1 Kodiak Island 11 23:00 :24.0 5.4 Kodiak Island 16 1:27:15.3 5.7 Kodiak Island 17 16 :46:50.9 4.5 Queen Charlotte Is. 22 23:27:20.5 5-9 Kodiak Island May 5 (15 :04) Unidentified 11 (14:27) Unidentified 11 (21:49) — Unidentified 18 (7:37) — Unidentified 37 TABLE II Continued: Date Time May 19, 1966 (7:11) 19 (13:59) 20 (9:27) 20 (24:00) Magnitude Location Unidentified Unidentified Unidentified Unidentified 38 CHAPTER III CONCLUSIONS It has been shown that a p r a c t i c a l and v e r s a t i l e seismograph station can be constructed using the electronic feedback techniques developed by R.D. Meldrum. With these feedback methods, I t has been possible to create a c r i t i c a l l y damped 5 second period seismometer from an underdamped Willmore Mk. I instrument by applying negative feedback of a modified form of the output signal from the seismometer to i t s transducer through a Maxwell impedance bridge. Within l i m i t s imposed by the ground and c i r c u i t noise and the c h a r a c t e r i s t i c s of the amplifiers used i n the feedback, i t i s possible to vary the damping r a t i o and resonant period of the feedback seismometer to any desired values through minor modifications of the transfer function of the feedback loop. The feedback seismometer has been u t i l i z e d i n the construction of a U.B.C. seismograph station. Although there i s a very high ambient noise l e v e l at the seismograph station s i t e , i t has been possible to separate seismic signals from the noise by means of electronic f i l t e r i n g f o r recording purposes. However, f i l t e r i n g i s not possible within the feedback loop of the instrument i t s e l f and the high amplitude noise can disturb the function of the feedback and cause i t s temporary breakdown. C i r c u i t noise within the feedback loop has been held below the ground noise signal l e v e l s to ensure that the associated e l e c t r o n i c s does 39 not l i m i t the available s e n s i t i v i t y of the seismograph. The U.B.C. stat i o n has been continuously operating since November 1965 and has shown that i t can be useful f o r demonstration and research purposes. More than 40 s i g n i f i -cant large teleseisms from as f a r away as the Indian Ocean and l o c a l tremors have been recorded by the station during i t s operation. 40 REFERENCES 1. Bonn, E.O., The Transform Analysis of Linear Systems, Addison-Wesley Publishing Co., Inc., Reading, Mass., 1963. 2. Brune, J.N., and J . Oliver, The Seismic Noise of the Earth's Surface, B u l l . Seism. Soc. Am., 49, 349-353, 1959. 3. Burr-Brown Research Corp., Handbook of Operational Amplifier Applications, Burr-Brown Research Corp., Tuscon, A r i z . , 1963. 4. Department of Mines and Technical Surveys, Canadian Seismgraph Calibrations, Department of Mines and Technical Surveys, Ottawa, Canada, 1965. 5. Gutenburg, B., and C.F. Richter, Magnitude and Energy of Earthquakes, Annali d i Geofisica, 9, 1-15, 1956. 6. Kollar, F., and R.D. Russell, Seismometer Analysis Using an E l e c t r i c Current Analogue, B u l l . Seism. Soc. Am. (Submitted f o r Publication). 7. Kuo, B.C., Automatic Control Systems, Prentice H a l l , Inc., Englewood C l i f f s , N.J., 1962. 8 . Matumoto, T., On the Spectral Structure of Earthquake Waves, B u l l . Earthquake Res. Inst., Tokyo Univ. 38, 13-27, I 9 6 0 . 9. Meldrum, R.D., An Application of Feedback to E l e c t r o -magnetic Seismometers, M.Sc. Thesis, Department of Geophysics, University of B r i t i s h Columbia. 10. Richter, C.F., Elementary Seismology, W.H. Freeman and Co., San Fransisco, 1958. 11. Sutton, G.H., and G.V. Latham, Analysis of a Feedback Controlled Seismometer, J. Geophys. Res. 69, 3865-3882, 1964. 41 12. United States Coast and Geodetic Surveys, Summary of Preliminary Determinations of Epicenters, U.S. Department of Commerce, Environmental Science Services Administration, November 1965--May 1966. 13. Willmore, P.L., The Application of the Maxwell Impedance Bridge to the C a l i b r a t i o n of Electromagnetic Seis-mographs, B u l l . Seism. Soc. Am., 49, 99-114, 1959-PART II AN ANALYSIS OP THE LONG SHOT DATA PROM THE CANADIAN SEISMOGRAPH STATIONS CHAPTER 1 INTRODUCTION 1 Purpose of Long Shot Before agreement i s reached on a policed nuclear test ban treaty, p a r t i c i p a n t s require the a b i l i t y to detect v i o l a t i o n s . Seismic methods have been shown to be the best available f o r detection, l o c a t i o n and i d e n t i f i c a t i o n of underground nuclear b l a s t s . The d i f f e r e n t i a t i o n of man-made from natural seismic events i s a subject of active current research. Advancement i n t h i s f i e l d should reduce the number of ambiguous events, and thus the number of on-site inspections of suspected v i o l a t i o n s requested by ce r t a i n countries. The purpose of Long Shot, the long range seismic measurement experiment, was to provide data to further the a b i l i t y to detect, locate and d i s t i n g u i s h nuclear explosions from natural seismicity at long range. LONG SHOT - LONG RANGE SEISMIC MEASUREMENT [6] DATE October 29, 1965 TIME 21:00:00.08 Universal Time LOCATION Amchitka Island, Aleutians Latitude 51°26"17 n Longitude 179°10'57 n DEPTH 2303 Peet below surface ELEVATION -2164 Peet FORM OP SHOT 80 Kiloton TNT equivalent nuclear device This report comprises a p i l o t analysis of Long Shot a r r i v a l s recorded by the 24 stations of the Canadian seismic network and a more comprehensive analysis of data from the special stations operated by the A r c t i c Institute of North America, Areas of possible further i n t e r e s t i n g study were sought together with co r r e l a t i o n s of some of the more obvious r e s u l t s with the accepted s t r u c t u r a l p i c ture. 1-2 Present Understanding of Explosion Seismology Marked differences between the source mechan-isms of earthquakes and explosions are known. The explosion creates an intense impulsive pressure wave which near the source generates stresses exceeding the e l a s t i c l i m i t of the rock mass, causing i t to be crushed and cracked. The pressure quickly attenuates with d i s -tance u n t i l the material begins to act e l a s t i c a l l y . Beyond t h i s distance, the wavefront behaves as though the source were a point with a compressional f i r s t pulse moving r a d i a l l y from i t . On the other hand, earthquakes are thought to be the r e s u l t of stresses b u i l t up over some large volume and released through shearing along a f a u l t plane. The source appears to be l i n e a r and the observed f i r s t pulse can be either d i l a t i o n a l or compressional depending on the 44 r e l a t i v e o r i e n t a t i o n of the source and observer. Because the source i s not s p h e r i c a l l y symmetric, azimuthal v a r i -ations i n signal amplitude, shape and d i r e c t i o n of f i r s t motions are often evident. Moreover, the stress release creates body shear waves and surface Love and Rayleigh waves as well as lo n g i t u d i n a l body waves. In explosions, source-generated shear and Love waves are not expected although a conversion of pressure to shear waves as well as a secondary creation of shear waves through the release of e x i s t i n g rock stresses can occur. These subsidiary phases quickly at-tenuate with distance and are not s i g n i f i c a n t at t e l e -seismic distances. Although Love waves (transverse surface waves) cannot be generated by the above i d e a l i z e d explosion mechanism, they sometimes do occur, probably as a r e s u l t of geological inhomogeneity near the shot point. Carpenter [5] has noted that the Nevada Test Site data indicate shallow earthquakes generate much greater Love and Rayleigh wave amplitudes than explosions of comparable energy. These source e f f e c t s lead to c r i t e r i a f o r d i f -f e r e n t i a t i o n of explosions and natural phenomena. The most obvious c r i t e r i o n i s that of l o c a t i o n and depth; the next i s that of the f i r s t motion which must be compres-sional for a l l explosions and can be either compressional 45 or d i l a t i o n a l , depending on source-observer orientation, for earthquakes. Other c r i t e r i a are the v i r t u a l lack of S (shear) waves at teleseismic distances, lack of azimuthal variations of amplitude and sign a l shape, a comparitively simple P phase f o r explosions, and r e l a t i v e l y reduced apparent surface wave amplitudes [ 1 0 ] . Certain p r a c t i c a l d i f f i c u l t i e s arise i n the application of these detection c r i t e r i a . Carpenter has suggested that the f i r s t motion c r i t e r i o n produces d i f -f i c u l t y i n two ways. The f i r s t pulse i s generally small compared to the successive motions i n the P wave t r a i n and may be obscured. He suggests that a minimum si g n a l -to-noise r a t i o as high as 10 may be required to allow "confident recognition" of the character of the f i r s t motion. He has also noted that f o r stations at d i s -tances greater than 3,000 kilometers, the received signal leaves the focus within a 45° l i m i t i n g cone oriented downward about a v e r t i c a l axis. I t i s possible that t h i s l i m i t e d angle could show eithe r purely compres-sional or d i l a t i o n a l character f o r the i d e a l i z e d earth-quake. The f i r s t motions can best be applied to es t a b l i s h events as earthquakes i f unambiguous d i l a t i o n s are found. It has been suggested that t h i s c r i t e r i o n be u t i l i z e d only f o r magnitude 4.5 and greater events so that a s u f f i c i e n t sample of stations may determine the f i r s t H6 motion character r e l i a b l y [ 1 0 ] For the remaining c r i t e r i a , Carpenter has pro-posed a s t a t i s t i c a l approach to determine a p r o b a b i l i t y that a given event i s an explosion. This, however, requires a large control sample and adequate data are available only f o r the Nevada Test S i t e area at the present time. These suggested detection c r i t e r i a have been considered throughout the present analysis, however, i t has not been possible to test them rigorously with the l i m i t e d data available from the Canadian stations. This study has been pr i m a r i l y concerned with cataloguing the basic data and obtaining r e s u l t s of d i g i t a l analysis of some special records. 1-3 Contributing Stations The data used i n t h i s analysis were obtained from the Canadian Dominion Observatory seismic network and f i v e special A r c t i c I n s t i t u t e of North America seismic stations. A map of the d i s t r i b u t i o n of the stations i s shown i n F i g . 1. CANADIAN SEISMIC STATIONS YELLOWKNIFEi FORT ST JAMES .PORT HARDY WILLIAMS LAKE ACOPPERMINE 'BAKER LAKE 17 LEDUC- AFLINFLONx AQREAT •EDMONTON <^  WHALE VICTORIA A ROCKY MTN HOUSE PENTICTON 300 SCALE 600 ST JOHN'S* ©SEPT-ILES SEVEN FALLS WAWA. / /^HALIFAX c ^ ^ O T T A W A A ^MONTREAL f / \ ["^SCARBOROUGH MILES f N D O N A y /  PIG. The Canadian seismograph stations, 48 TABLE I L i s t of Contributing Stations Dominion Observatories Branch Aler t (ALE) 82°29 'N 62°24 •W Baker Lake (BLC) 64°19 •N 96°01 'W Coppermine (CMC) 67°50 N 115°05 'W Edmonton (EDM) 53°13 •N 113°21 'W P l i n Plon (PPC) 54°44 'N 101°56 •w Port St. James (PSJ) 54°26 »N 124°15 'W Probisher (PBC) 63°44 'N 68° 28 'W Great Whale River (GWC) 55°l8 'N 77°45 'W Halifax (HAL) 44038 N 63°36 'W London (LND) 43°02 N 8l°ll 'W Montreal (MNT) 45° 30 N 73°37 'W Mould Bay (MBC) 7 6 ° l 4 'N 119°20 'W Ottawa (OTT) 45 0 24 N 75°43 w Penticton (PNT) 49° 19 N 119°37 w Port Hardy (PHC) 50°42 N 127°26 w Resolute (RES) 74° 41 N 94054 w St. John's (STJ) 47° 34 N 52°44 w Scarborough (SCB) 43043 N 79° 14 w S c h e f f e r v i l l e (SCH) 54049 N 66°47 w Sept l i e s (SIC) 50°10 N 66°40 w Seven F a l l s (SFA) 47°07 N 70°50 w Shawinlgan F a l l s (SHP) 46°33 N 72°46 w V i c t o r i a (VIC) 48° 31 N 123°25 w Yellowknife (YKC) 62°29 N 114°29 w 49 A r c t i c I n s t i t u t e of North America Leduc (LD6) 53°13*N 113°21'W Nova Scotia 44°31'N 63°53'W Rocky Mtn. House (RM6) 52°24'N 114057'W Wawa (WAW) 48017'N 85°02»W Williams Lake (WLC) 52°12'N 122°02'W The Leduc (LD6) and Edmonton instruments were operated on the same seismic p i e r near Edmonton. 50 CHAPTER II ANALYSIS OF LONG SHOT. DATA - GENERAL 2-1 Character of Records Host short period instrument records show a d i s t i n c t P wave a r r i v a l of several cycles duration. Some stations also recorded a core r e f l e c t e d PcP phase 2 to 3 minutes a f t e r the i n i t i a l a r r i v a l but no sta t i o n records showed shear phases v i s i b l e to eye examination. Several long period instruments recorded 20 second period surface Rayleigh waves of low amplitude 20 to 30 minutes a f t e r the f i r s t motion. Appendix I l i s t s the major a r r i v a l s with the equivalent ground v e l o c i t i e s and periods f o r a l l stations. The low amplitude surface waves and the lack of shear phases was consistent with the i d e a l i z e d explosion source mechanism. 2-2 R e l i a b i l i t y of Data The signal-to-noise r a t i o of the a r r i v a l i s an important f a c t o r i n recognition, and i n the accuracy of time and f i r s t pulse character determination. For a l l records, t h i s was obtained by d i v i d i n g the maximum P a r r i v a l amplitude of the f i r s t few cycles by the maxi-mum p r e - a r r i v a l noise amplitude i n the two minutes 51 immediately preceding the a r r i v a l . A s i m i l a r method was used f o r the surface wave signal-to-noise determinations. Ichlkawa and Basham [8] have shown that a P phase can normally be i d e n t i f i e d i f the signal-to-noise r a t i o i s 1 or greater. However, Carpenter [5] has suggested a minimum value of up to 10 i s required f o r accurate determination of the f i r s t motion. Appendix I l i s t s the signal-to-noise r a t i o s f o r each station, phase and component. Less than one t h i r d of the Canadian stations meet Carpenter's requirement but a l l except the Williams Lake and Nova Scotia stations have s u f f i c i e n t signal-to-noise r a t i o s f o r i d e n t i f i c a t i o n of the P phase. 2-3 A r r i v a l Times and Time Residuals Accurate l o c a t i o n of seismic events requires precise knowledge of the t r a v e l times of the seismic waves. The standard Jeffreys-Bullen travel-time tables were determined on the basis of analyses of selected earthquakes f o r which no independent knowledge of po s i -t i o n , depth and time were ava i l a b l e . In these analyses i t must be assumed that the crust and mantle are spheri-c a l l y symmetric. The Long Shot l o c a t i o n i n the t e c h n -i c a l l y active Aleutian Island arc i s one of strong crust and upper mantle anomaly and i t i s probable that i f time biases do occur, they w i l l be noted f o r events i n t h i s region. ^9 To determine whether biases do occur, two sets of t r a v e l time data were used: the 1958 p r i n t i n g of the Jeffrey-Bullen tables and the 1961 modification of these by Herrin f o r surface focus P a c i f i c events [6] . The actual a r r i v a l times were obtained, excluding any errors i n proper a r r i v a l i d e n t i f i c a t i o n , to +0,2 second f o r the Pominion Observatory records and +0.02 second f o r the high speed A r c t i c I n s t i t u t e of North America re-cords. The r e s u l t s indicate that f o r a l l stations, the P a r r i v a l was early compared to both sets of t r a v e l time data. See Table II and I I I . The a r r i v a l s at the closest stations were 0 to 3 seconds e a r l y and at the farthest, 3 to 6 seconds early and a general increase i n the time bias with distance i s evident. P i g . 2(a). In order to determine whether t h i s same apparent error i n t r a v e l times has occurred f o r earthquakes i n the v i c i n i t y of Amchltka, the t r a v e l times of an earthquake of March 30, 1965 (Location: 50.6°N, 177.9°E; Depth 51 kilometers - U.S.C.G.S. determinations) were studied. A s i m i l a r trend of increasing residuals with increasing distance was observed i n d i c a t i n g that the bias i s most l i k e l y an e f f e c t of transmission path rather than the s p e c i f i c l o c a l b l a s t environment or source mechanism. However, the values of the residuals vary about 8 seconds from those of Long Shot. See Pig. 2(b). This probably shows that the depth, p o s i t i o n and time determinations TABLE I I Long Shot Time Residuals Station Ale r t (ALE) Baker Lake (BLC) Coppermine (CMC; Edmonton (EDM) P l i n Flon (FFC) Fort St. James (FSJ) Frobisher (FBC) Gm& whale River (GWC) Halifax (HAL) London (LND) Montreal (MNT) Mould Bay (MBC) Ottawa (OTT) Penticton (PNT) Port Hardy (PHC) Resolute (RES) St. John's (STJ) Scarborough (SCB) S c h e f f e r v i l l e (SCH) Sept l i e s (SIC) Seven F a l l s (SFA) Shawinigan F a l l s (SHF) V i c t o r i a (VIC) Yellowknife (YKC) Leduc (LD6) N©va Scotia Rocky Mtn. House (RM6) Wawa (WAW) William's Lake (WLC) A r r i v a l Time af t e r 21:00 :57.5 :01.5 6:52.7 :32 .7 1:15.0 6 :40.0 9:16.2 9:37.4 11:11.4 10:26.5 10:36.5 6:44.3 10:30.9 7:18 .3 6:36.5 7:36 .0 11:21.5 10:29.5 10:11.0 10:34.5 10:37.4 10:34 .3 7:04.7 7:01 .8 7:32 .6 7 :28.3 9:47.7 Jeffreys-Bullen 6:54 i : l 8 6:41 9 :24 9 :41 11:14 10:30 10:41 6:47 10:36 7:21 6:37 7:38 11:25 10:33 10:14 10^38 10:42 10:39 7:05 7:05 7:36 11:14 7:31 9:51 6:57 3 Residual -1.5 -2 .5 -1 .3 - 3 . 3 - 3 . 0 -1 .0 - 7 . 8 - 3 . 6 3.6 . 5 .5 - 2 . 7 -5.1 - 2 . 7 - 0 . 5 - 2 . 0 - 3 . 5 - 3 . 5 - 3 . 0 :tt - 4 . 7 - 0 . 3 - 3 . 2 - 3 . 4 - 2 . 7 - 2 . 3 Herrln 8 :00 . 8 :05 . 6 :55.7 7:36 .2 6:19 .3 6:42.3 9:25.0 9:42.4 11.15.1 10:30.8 10:43.0 6:47.3 10:37-1 7:22.0 6:37.7 7:39.3 11:25.9 10:34.1 10:15.5-10:40.2 10:43.3 10:41.4 7 :06 .3 7:05.1 7:36.2 7:18 .4 9 :52.4 Residual -2 .8 -3 .9 -3 .0 .5 .3 -2 .3 - 8 . 8 -5 .0 -4 .7 -4 .3 -6.5 -3 .0 -6.2 -3-7 -1 .2 3:1 -4 .6 -4.5 -5 .7 -5 .9 -2.4 -3 .3 -3 .6 54 TABLE I I I Earthquake Time Residuals J e f f r e y s -Station Travel Time Bullen Residual A l e r t 8:03.5 7:55 +8.5 Coppermine 7:00.3 6:53 +7.3 Edmonton 7:40.6 7:33 +7.6 Fort St. James 6:47.4 9:28.2 6:40 +7.4 Probisher 9:20 +Q.Z Halifax 11:15.8 11:09 +6.8 London 10:31.3 10:25 +6.3 Montreal 10:42.5 10:?7 +5.5 Mould Bay 6:51.6 6:44 +7.6 Ottawa 10:36.4 10L31 +5.4 Penticton 7:24.8 7:18 +6.8 Port Hardy 6:40.8 6:36 +4.8 Resolute 7:43.3 7:36 +7.3 St. John's 11:26.8 11:19 +7.8 S c h e f f e r v i l l e 10:17.5 10:10 +7.5 Seven P a l l s 10:42.4 10:37 +5.4 Shawinigan F a l l s 10:40.0 10:34 +6.0 V i c t o r i a 7:09.7 7:03 +6.7 Yellowknife 7:10.6 7:03 +7.6 55 TIME RESIDUALS FOR LONGSHOT F .VIC >HC» FSJ* •CMC •ALE nr-r» MBC. P N J • m 6 ^ L ° _ _ #WAW YKC« • EDM -D6 GWC« L l SJD* * •SIC e HAL S H F««SFA *0TT 30 40 50 DISTANCE IN DEGREES PIG. 2(a). Long Shot time residuals. 60 70 TIME RESIDUALS FOR EARTHQUAKE •ALE •FBC MBC. T^KC , FSJ» «CMC •EDM •RES S •SCH VIC* #P NT HAL* LND# , 0SCH #MNT F 30 40 50 DISTANCE IN DEGREES 60 70 FIG. 2(b). March 30, 1965 Amchitka Island earthquake time residuals. 56 f o r the earthquake are i n error by a small amount. 2-4 F i r s t Motions One of the s a l i e n t features of an explosion i s the compressional spherical wavefront which i s gen-erated. The character of the wavefront should be pre-served along the path from the explosion through the mantle to the recording station. However, at any p a r t i c u l a r station, the character of the f i r s t motion may be obscured by background noise or may be too low i n amplitude to be determined. FigB. 3(a), (b) and (c) show sketches of the f i r s t motions f o r the v e r t i c a l component. I t appears that of a l l the stations, only V i c t o r i a , Leduc, Rocky Mountain House and F l i n Flon recorded apparently unambiguous d i l a t i o n s . However, the high speed V i c t o r i a record which was recorded on the same seismic p i e r as the standard V i c t o r i a record, shows a compression preceding the d i l a t i o n . S i m i l a r l y , i t was possible.to pick the compressional f i r s t motion on the Leduc and Rocky Mountain House records upon comparing the Leduc and Edmonton records which were obtained on the same pi e r . No dependence on the event to s t a t i o n azimuth or distance was noted f o r the apparent d i l a t i o n s . 57 F i r s t Motions f o r Canadian S t a t i o n s ( I , ALERT BAKER LAKE COPPER MINE EDMONTON FLIN FLON FORT ST. JAMES FROBISHER • V I GREAT WHALE HALIFAX RIVER LONDON MONTREAL F i g . 3(a). Character of f i r s t motions at various Canadian s t a t i o n s (sketches, not to s c a l e ) . 56 F i r s t Notions f o r Canadian S t a t i o n s ( I I ) MOULD BAY OTTAWA PORT HARDY PENTICTON RESOLUTE ST. JOHN'S SCHEFFERVILLE SEPT ISLES SEVEN PALLS SHAWINIGAN FALLS VICTORIA ? j g . 3 ( b ) . Character of f i r s t motions at various Canadian s t a t i o n s (sketches, not to s c a l e ) . 59 P i g - 3 ( c ) . Character of f i r s t motions at various Canadian s t a t i o n s ( t r a c i n g s ) . 60 2-5 Ground V e l o c i t i e s and Unified Magnitudes A seisinogram i s e s s e n t i a l l y a record of the v e l o c i t y of the ground motion i n some bandwidth. For the standard Canadian st a t i o n short period instruments, t h i s band pass Is approximately 0.2 to 1.0 second period. At any angular frequency t h i s record ampli-tude i s a product of the ground v e l o c i t y and the s e n s i t i v i t y of the seismograph system to the ground v e l o c i t y : The v e l o c i t y s e n s i t i v i t y of the seismograph, Sv(&) , i s usually a v a i l a b l e i n the form of a " c a l i b r a t i o n curve" i n which the s e n s i t i v i t y versus period i s plotted as i n the curves of F i g . 8 . The maximum ground v e l o c i t y i n the f i r s t few cycles of the P a r r i v a l depends upon the energy release of the event and i s e a s i l y obtained from the records. i n Appendix X, the maximum peak to peak velo-c i t y i n the f i r s t few cycles Is l i s t e d f o r each station and various phases. These v e l o c i t i e s are plotted versus ep l c e n t r a l distance i n degrees f o r most Canadian stations i n F i g . 4. The superimposed s o l i d l i n e represents the expected maximum v e r t i c a l v e l o c i t i e s f o r the Long Shot P a r r i v a l s . (After Carpenter [5] ). I t i s notable 61 MAXIMUM PEAK TO PEAK GROUND VELOCITIES 20000 u CO £ t o o _ l tu > < LLI 0_ < UJ CL X 10000 5000 2 000 1000 500 200 100 •BLC GWC» • SIC •SCH L ^ t L D 6 / • w f s . ^ - t r t B C *RES \ •FBC •LND •SHF •STJ HAL •F WI6 ^ .ALE •OTT •VIC 3 H C « JWLC •FFC •FSJ 30 40 50 60 DISTANCE IN DEGREES 70 80 PIG. 4. The maximum observed peak to peak v e r t i c a l ground v e l o c i t i e s as a function of epicentral f o r the Canadian stations. 62 that the closest stations show low v e l o c i t i e s and the f a r -thest stations show high v e l o c i t i e s compared to those expected. In p a r t i c u l a r , the B.C. stations at Port Hardy, Port St. James and William's Lake (The William's Lake value i s an upper l i m i t to possible motion) show anomalously low v e l o c i t i e s . An important c h a r a c t e r i s t i c of a seismic event i s the magnitude which i s a measure of the energy re-lease. The u n i f i e d magnitude scale of Richter and Gutenburg f o r P waves i s based on a formula r e l a t i n g the a r r i v a l ground amplitude and period of the motion at the recording s i t e and the distance and depth of the event t m =log(AA) + Q(A,h) where A i s the zero to peak maximum v e r t i c a l com-ponent of the ground amplitude i n microns; T" * the period of the wave; and Q^Ajh) , a constant depending on the distance and depth of the event. I t i s evident that: A/T = Vw/4TT where i s the maximum peak to peak ground v e l o c i t y . A corre c t i o n can be applied to the magnitude formula i n order to use the resultant horizontal motion rather than 63 TABLE IV Unified Magnitude Determinations-Canadian Stations Station A l e r t Baker Lake Coppermine Edmonton P l i n Plon Fort St. James Probisher Great Whale River Halifax London Montreal Mould Bay-Ottawa Penticton s Port Hardy Resolute St. John's Scarborough S c h e f f e r v i l l e Sept l i e s Seven P a l l s Shawlnlgan F a l l s V i c t o r i a Yellowknife Leduc Rocky Mt. House Williams Lake V e r t i c a l Horizontal Average 5.5 6.3 6.1 6.0 5.5 5.2 6.2 6.7 6.2 6.5 6.0 6.1 6.1 6.0 5.8 6.4 6.5 6.7 6.0 1:1 6.0 6.1 5.7 l e s s than 5.3 5.5 6.2 6.0 6.1 5'7 5.4 6.3 6~1 6?1 6.1 6.4 5.4 5.5 1:1 6.4 6.1 6.0 6.0 5.6 Residual Corrected 5.55 5.95 5.9 5.75 5.9 5.65 5.4 5.75 6.0 5.75 6.05 5.65 (5.3) Mean Magnitude and Standard Deviation 6 . 0 1 ± 0.40 5 . 7 6 ± 0 . 1 9 Surface Magnitude Determinations COPPERMINE PROBISHER RESOLUTE YELLOWKNIFE 5.1 5.2 1:1 UNIFIED MAGNITUDES FOR CANADIAN STATIONS 7.0 6.0 5.0 GWC A c SIC A LND A W MBC LD Q •JZMC Eg l V,C* «V PHP #BLC 6 ^ AVERAGE IW ' ALL ST/ »RES f"6 .FFC • a i r-FBC • M3NITUDE mONS • "SHF A # HA #MNT SFA STJ — • -L 1 1 1 w • WALE 30 40 50 60 70 DISTANCE IN DEGREES FIG. 5. Unified magnitudes as a function of ep i c e n t r a l distance for the Canadian stations. 65 the v e r t i c a l motion. Where possible, both v e r t i c a l and horizontal methods were used and the magnitudes averaged. Table IV l i s t s the calculated magnitudes f o r most Canadian stations; these magnitudes are plotted versus distance i n F i g . 5 and indicated beside s t a t i o n positions on an azimuthal plot i n F i g . 6. The mean u n i f i e d magnitude f o r the 27 Canadian stations was found to be 6.01 with a standard deviation of 0.40. This value was consistent with the world-wide average determination of 5.99+0.52 but the scatter among the Canadian stations was s l i g h t l y l e s s [6J Ichlkawa and Basham [8J have determined magnitude residuals f o r 10 Canadian stations f o r earthquakes i n various regions. The corrections f o r Aleutian events-were applied to the calculated magnitudes f o r these stations and a revised magnitude with a reduced standard deviation f o r the 10 stations was obtained: 5.76+0.19 (Table IV). I t i s obvious that three of the B.C. stations show anomalously low magnitudes while several eastern Canadian stations show high magnitudes. The anomaly i n B.C. was v e r i f i e d by separate determinations f o r Long Shot by Vela-Uniform LRSM portable stations [10] f o r Smithers (4.95) and Prince George(4 .75). 66 NORTH AZIMUTHAL PLOT OF DISTANCE IN DEGREES PIG. 6. Azimuthal plot of the observed u n i f i e d magnitudes f o r Long Shot. 67 The ^method of magnitude determination used i n t h i s analysis averages the v e r t i c a l and horizontal values. Thus any enhancement of the v e r t i c a l amplitude at the expense of the horizontal, or vice versa, w i l l not a f f e c t the c a l c u l a t i o n s . I t i s expected that the e f f e c t s of r e f r a c t i o n of a seismic wave with respect to the normal to surface along the path and reverber-ation within c r u s t a l layers would show up as large differences i n the horizontal and v e r t i c a l magnitude values f o r the s t a t i o n receiving the modified signals. In order to determine which stations had received highly refracted or reverberated a r r i v a l s , z t h e differences between the horizontal and v e r t i c a l magnitudes f o r each sta t i o n were obtained. In Pig. 7, these magnitude differences are plotted versus epice n t r a l distance. Large differences, i n d i c a t i n g r e f r a c t i o n or reverbera-t i o n e f f e c t s , occurred f o r Penticton, V i c t o r i a and Ottawa. At V i c t o r i a and to l e s s e r extent at Ottawa, the h o r i -zontal motion of the a r r i v a l WSB enhanced, while at Penticton, the v e r t i c a l motion was enhanced. However, the Ottawa records were very noisy and i t was d i f f i c u l t to ascertain how genuine the apparent e f f e c t was. Only s l i g h t r e f r a c t i o n or reverberation e f f e c t s were indicated at Port St. James and Port Hardy which suggests that the anomalously low a r r i v a l amplitudes at these s i t e s are possibly due to absorption. 68 ID Q 3 t < N CE O X I UJ Q < o »-UJ > MAGNITUDE DIFFERENCES FOR CANADIAN STATIONS .6 .4 .2 0 -.2 -.4 -.6 — n u T CMC. J ^ RMb MBC» »YKC EDM* pD6#BLC A A • r— ,SCH # HAL p •ALL *RES •FBC •MNT •STJ F •SJ# FFC C )TT» •VIC 30 40 50 60 DISTANCE IN DEGREES 70 PIG. 7. The differences between the horizontal and v e r t i c a l magnitude determinations as a function of epi c e n t r a l distance. Magnitudes were also determined on the basis of surface waves, recorded by 4 of the Canadian stations, according to a formula developed by Bath [ l ] : M - log A z - l o g B + S(h) + mr + c(M 0-M c a l c) where A_ i s the v e r t i c a l amplitude i n microns f o r z 20 second period surface waves; B,S(h),m r, and c are constants dependent upon the geographical l o c a t i o n and 69 depth of the event; M Q i s a f i x e d constant, 7.1, and M c a l c l s t n e 8 u m o f t n e f l r s t ^ t e r m s l n t n e formula. Corrections can be applied i n order to use waves of periods 10<T<23 seconds. There i s a large range of values f o r the constants required i n the formula; f o r example, mr can vary from - 0 . 3 to +0.6 which r e s u l t s i n a magnitude v a r i a t i o n of almost 1. The constants were estimated on the basis of the methods outlined by Bath. The magnitudes determined by the surface wave method were 0.6 to 1.1 lower than the u n i f i e d magnitudes determined from the P a r r i v a l s . 1 (Table IV.) Using an extrapolation of a nomogram prepared by J.M. Nordquist f o r the determination of the u n i f i e d magnitude from the horizontal resultant motion of surface waves^ of 20 second period, surface wave magnitudes t y p i c a l l y 0.2 l e s s than the Bath values were obtained. REFERENCE: GUTENBURG, B. and G.G. RICHTER, Magnitude and Energy of Earthqquakes, Annali d i Geofisica, 9, 1956, 1-15. 7b CHAPTER I I I CHARACTER OP SIGNAL 3-1 Special Records Five s p e c i a l stations recorder* the Long Shot a r r i v a l s on high speed magnetic tape recorders or chart recorders: Leduc, Nova Scotia, Rocky Mountain House, Wawa, and William's Lake and V i c t o r i a . Of these, the William's Lake and Nova Scotia stations d i d not produce data which was useable i n the analysis. A l l instruments were short period, nominally 0.2 to 1.0 second period except those of William's Lake (1.0 to 5.0 seconds) and Nova Scotia (0.04 to 0.5 seconds). The c a l i b r a t i o n curves of the Leduc, Rocky Mountain House, Wawa and V i c t o r i a v e r t i c a l component instruments are shown In Pi g . 8. The horizontal and v e r t i c a l instruments were operated with very s i m i l a r character-i s t i c s at a l l stations. For analysis of these special records, analogue chart traces of the magnetic record-ings were obtained andiconverted to d i g i t a l form by Engineering Data Processors Ltd., Calgary. (These d i g i t a l data are available from the Uni v e r s i t y of B.C., Department of Geophysics, Vancouver 8, B.C.) 72 SPECIAL' STATION CALIBRATIONS ( I I ) 0.1 1.0 10.0 F i g . 8(c) Wawa C a l i b r a t i o n 0.1 1.0 10.0 F i g . 8(d) V i c t o r i a C a l i b r a t i o n 73 3-2 A F i l t e r i n g Technique f o r Improved Character of the  Wawa Records The most common type of seismometer i s a mechanical o s c i l l a t o r with mass M , suspended by a spring of constant U , and with damping of constant D, The o s c i l l a t o r i s mounted within a case which i s coupled to the ground. The instrument i s provided with a transducer which transforms r e l a t i v e motion between the case and the suspended mass into a voltage to drive a galvanometer or an amplifier and chart re-corder to record the ground motions. The transfer function f o r t h i s type of system can be obtained by li n e a r systems analysis techniques: H(s) _£s_ where S = (S'+jco ±s the complex frequency; g i s the transducer constant;o2.^oJ n = -pp = > f i s the damping r a t i o and U>r\ i s the resonant frequency of the seismometer 4 It i s common practice to operate seismometers with the damping r a t i o close to 1 ( c r i t i c a l damping). However, the instruments used at the Wawa station were considerably underdamped; the following dampting r a t i o s were obtained d i r e c t l y from each of the c a l i b r a t i o n curve s: 74 V e r t i c a l 0.l8 North 0.29 East 0.25. The e f f e c t of t h i s underdamping i s to allow the i n -struments to o s c i l l a t e with t h e i r natural frequency f o r any d r i v i n g s i g n a l . The o s c i l l a t i o n , c a l l e d ringing, greatly d i s t o r t s the form of the records. A f i l t e r was developed to transform the Wawa records into the forms which would have been obtained by c r i t i c a l l y damped instruments (£ - 1) with transfer function j^ 3J : The response of a c r i t i c a l l y damped instrument to a ground v e l o c i t y i s : vr'(t) = h ' f t ) * ^ where the ast e r i s k represents the convolution i n t e g r a l and Wit) Is the Impulse response of the seismometer. It can e a s i l y be shown that the r e l a t i o n between the o r i g i n a l , v ( t ) , and required record, u ' ( t ) , i s : v(t) = £-'{R(s)}*^ (t) where: R(s) = -M- -- s " ; ^ n s ^ and represents the inverse Laplace transform. The impulse response of the f i l t e r , K'(t). can be obtained from tables of Laplace transforms or by com-plex i n t e r g r a t i o n : r(t) = S(t) + 2(H 'Jn(l-w nt)e- w t where 8t f) i s the Dirac delta function. I t i s notable that the natural frequency, 6Jn , of the Wawa i n s t r u -ments i s approximately 8 radians per second and the f i l t e r impulse response damps extremely quickly. Because of the rapid damping of the time response of the f i l t e r It was possible to use f i l t e r s truncated at one second length. I t was shown that the spectral response of the truncated f i l t e r with the d e l t a func-t i o n subtracted was i n very good agreement with the th e o r e t i c a l response obtained from the analytic ex-pression f o r the f i l t e r response without the i n i t i a l impulse (Pig. 9). The spectral response of the trun-cated f i l t e r was calculated by computer using a num-e r i c a l Fourier transform method. The f i l t e r i n g of the Wawa records was accom-plished by numerically evaluating the truncated convolu-t i o n i n t e g r a l : u ' ( t ) = J o'r ( r M t - T J ) < l t 76 FREQUENCY RESPONSE OF VERTICAL FILTER 1.01 1 1 1 1 0.1 0.2 0.5 1.0 2 0 5.0 10.0 F R E Q U E N C Y ( C Y C L E S/sEC0ND > FIG. 9. The frequency response of the deconvolution f i l t e r f o r the v e r t i c a l component Wawa record. The curve i s the a n a l y t i c a l response. using a trapezoidal rule on the u n i v e r s i t y of B r i t i s h Columbia IBM 7040 computer. The o r i g i n a l and f i l t e r e d records are compared i n F i g , 10 f o r the i n i t i a l a r r i v a l s . These diagrams were obtained from a Calcomp computer p l o t t e r output of the convolution 77 COMPARISON OP FILTERED AND UNPILTERED WAV/A RECORDS Pig.10(a) V e r t i c a l : O r i g i n a l (top), f i l t e r e d (bottom) Fi g . l O ( b ) East: O r i g i n a l ( t o p ) ' F i l t e r e d (bottom) Fig.10(c) North. O r i g i n a l (top), F i l t e r e d (bottom) 78 i n t e g r a l evaluations and the small d i s c o n t i n u i t i e s are due to the d i g i t a l nature of the p l o t t i n g and sampling. I t i s evident that the sinusoidal resonant c h a r a c t e r i s t i c s of the records caused by the under-damped instruments have been removed and records appear much sharper a f t e r f i l t e r i n g . These recon-structed or deconvolved records were used i n the subsequent analysis. 3-3 Spectral Analysis The spectrum of a record of a seismic signal i s a product of e f f e c t s of source, path and seismo-graph c h a r a c t e r i s t i c s . The seismograph c h a r a c t e r i s t i c s are e a s i l y removed and the source e f f e c t s f o r a sp h e r i c a l l y symmetric explosion are i d e n t i c a l f o r a l l stations. Variations i n signal spectra between stations must then be a re s u l t of path differences from the source to the recording s i t e . Spectral analyses of the spec i a l records were c a r r i e d out to look f o r both common and unique features of the spectra. Peaks of common frequency and other s i m i l a r charac-t e r i s t i c s of spectra should be att r i b u t a b l e to source e f f e c t s or common signa l paths, while va r i a t i o n s between i n d i v i d u a l records might r e s u l t from major l o c a l s t r u c t u r a l differences. 79 In the time series analysis both Fourier i n t e g r a l transform and Tukey Power spectra were ob-tained using d i g i t a l record data and numerical methods. The e f f e c t of the d i g i t a l record data i s to a l i a s the spectrum 2 so that a l l frequencies represented i n I the record w i l l be observed In the i n t e r v a l ( O ) ^ ^ ) where A t i s the d i g i t i z a t i o n i n t e r v a l . In the present analyses the f o l d i n g or Nyquist frequency, fN * w a s chosen to be 10 cycles per second to reduce a l i a s i n g d i f f i c u l t i e s . Sharp f i l t e r i n g of the recorded signal above 5 cycles per second by the re-cording instruments also reduced the high frequency spectral component amplitudes and thus l i m i t e d the e f f e c t s of the f o l d i n g . The Fourier i n t e g r a l spectrum of any periodic signal i s defined by the i n t e g r a l : •00 where i s the time series; G("f) » the spectrum; and ud — The spectrum i s continuous and e x i s t s i f : •oo l : i # i d t OO For many p r a c t i c a l situations, t h i s condition i s v i o l a t e d but i n the present analysis, i t i s not r e s t r i c -t i v e to assume that the signal i s of f i n i t e length and 80 and amplitude and s a t i s f i e s the existence condition. The p a r t i c u l a r technique f o r evaluation of the Fourier i n t e g r a l and the basic computer program used i n t h i s analysis was developed by D. Weichert 9J . Tests on the method and computer program v e r i f i e d that the technique i s very accurate although computationally expensive. The power spectal methods f o r aperiodic data which are described by Blackman and Tukey are based on stationary time series or series with time invariant s t a t i s t i c a l properties. However, the signal data f o r the seismic a r r i v a l s from Long Shot are unique; that i s , non-stationary. For such data, Blackman and Tukey suggest that the time series must be treated e i t h e r as periodic or as a s t a t i s -t i c a l sample rather than as a unique event. The l a t t e r a l t e r n a t i v e , f o r which Blackman and Tukey's methods are v a l i d , was chosen f o r the analysis of the Long Shot s i g n a l . The power spectrum method involves compu-t a t i o n of the autocovariance function of the time series X.(~C) f o r time lags % ; C ( t ) = lim f % X ( t ) x ( t + t ) d t 81 The Fourier transform of the autocovarlance Is the power spectrum: For d i g i t a l time series of f i n i t e length, the auto-covarlance must be computed only f o r discrete lags which are i n t e g r a l multiples of the d i g i t i z a t i o n i n t e r v a l and the number of spectral estimates should be s u b s t a n t i a l l y l e s s than the t o t a l number of time series points. I t i s necessary to choose the maxi-mum l a g to be a small f r a c t i o n of the t o t a l time series length (usually l e s s than 1/10) to obtain the best estimate of the autocovarlance and power spectrum f o r a stationary random process. However, f o r a unique event which i s i d e n t i c a l l y zero outside the time series Interval such as a seismic signal, the autocovarlance function i s exact i f a l l lags are computed. In t h i s case, however, i t i s necessary to l i m i t the number of lags f o r computational economy and to obtain r e l i a b l e power spectrum estimates. The computer program used i n t h i s analysis was de-veloped by the Health Sciences Computing F a c i l i t y of TJ.C.L.A. [ l l ] . The spectra were smoothed by the "hamming" l a g window. In both types of analyses, record lengths of 30 to 60 seconds were used to include most of the 82 compressional wavetrain. However, the Wawa records were truncated at 10 seconds length because of a high wind noise l e v e l . The continuous Fourier spectra were sampled at i n t e r v a l s of 0.2 cycles per second up to a frequency of 5*0 cycles per second with the r e s u l t that the f i n e structure of the spectra bet-ween the sample points was l o s t . A more s a t i s f a c -tory procedure would require more d e t a i l e d sampling and consequent smoothing but t h i s would be compu-t a t i o n a l l y expensive. Ground noise spectra f o r the 30 seconds p r i o r to the f i r s t signal a r r i v a l were also obtained f o r the Leduc and Rocky Mountain House records. In F i g . 11(a) and 11(b) the i n s t r u -ment c h a r a c t e r i s t i c s have been removed to obtain the equivalent ground amplitude spectra. The power spectra were computed using about 1/5 the t o t a l available lags and were sampled at 0.2 cycles per second i n t e r v a l s up to a frequency of 10 cycles per second. The instrument c h a r a c t e r i s t i c s have been removed to obtain the v e r t i c a l ground velo-c i t y power spectra of F i g . 12. 83 LEDUC EAST ROCKY MTN HOUSE EAST 10 O l h 0001 P ARRIVAL NOISE 00\\- \, FREQUENCY (CYCLES/SECOND) FIG. 11(a) The Fourier amplitude spectra of P a r r i v a l s from Long Shot 84 VICTORIA WAWA 01 0 01 3 0001 00001 \ ^  VERTICAL EAST NORTH \\H\\ v-Vs. \\ '» 1 V wj*^ 0001 o o i h FREQUENCY (CYCLES/SECOND) FIG. 11(b). The Fourier amplitude spectra of P a r r i v a l s from Long Shot. 85 POWER SPECTRA VERTICAL COMPONENTS 0 1 2 3 4 FREQUENCY (CYCLES/SEC.) PIG. 12. Power spectra of the v e r t i c a l component of the P a r r i v a l ground v e l o c i t y f o r Leduc, Rocky Mountain House, V i c t o r i a and Wawa. 86 3-4 Results of Spectral Analysis Common major peaks at 1.2 to 1.4 an4 2.2 to 2.4 cycles per second separated by a trough at 1.8 to 2.0 cycles per second are evident i n a l l spectra. Several other peaks appear to be unique to the p a r t i c u l a r s t a t i o n and record. The common frequency peaks could be either source generated or a r e s u l t of si m i l a r s e l e c t i v e f i l t e r i n g along the wave paths while the i n d i v i d u a l peaks can most l i k e l y be associated with transmission properties of the crust near the stations. At the higher f r e -quencies the noise spectra and the a r r i v a l spectra at the Leduc and Rocky Mountain House stations show some sim i l a r c h a r a c t e r i s t i c peaks which suggests that these are caused by l o c a l e f f e c t s . amplitude and power with increasing frequency. In spectral studies of earthquakes, ichickawa and Basham concluded that the form of t h i s spectral amplitude decrement was not a function of ep i c e n t r a l distance but was related to upper mantle or c r u s t a l structure. To i l l u s t r a t e the character of the spectral amplitude decrement f o r the Long Shot P a r r i v a l s , the v e r t i c a l component Fourier spectra of each of the four stations were f i t t e d to a function: The spectra show a s i g n i f i c a n t decrease of 87 using a l e a s t squares technique as suggested by Ichikawa and Basham 8 . and the constant 3-was calculated with i t s 90$ confidence i n t e r v a l Station ci (seconds/cycle) V i c t o r i a 2 . 2 0 + 0 . 2 6 Leduc 1 . 2 4 + 0 . 2 2 Rocky Mtn. House 1 . 4 4 + 0 . 3 8 Wawa 1.68 + 0 .43 The f i t t e d curves are shown superimposed on the v e r t i c a l amplitude spectra i n Pigs. 11. The constant & i s only related to the r e l a t i v e frequency con-tent of the seismic signal which Is assumed to remain constant over the duration of the wavetrain of com-pressional a r r i v a l s . The wide confidence i n t e r v a l s prevented an accurate comparison of the gross f r e -quency content of the various P a r r i v a l s . However greater attenuation of high frequencies occurred at the V i c t o r i a and Wawa stations than at the Leduc and Rocky Mountain House stations. The explanation of these e f f e c t s i s not evident. 88 REFERENCES Bath, M., Earthquake Magnitude Determination From the V e r t i c a l Component of Surface Waves, Trans. Amer. Geophys. Union, 33, 81-90, 1952. Blackman, R. B., and J. W. Tukey, The Measurement of Power Spectra, Dover Publications, Inc., New York, 1959. Bogert, B.P., Correction of Seismograms f o r the Transfer Function of the Seismometer, B u l l . Seism. Soc. Am., 52, 781-792, 1962. Bohn, E. V., The Transform Analysis of Linear Systems, Addison-Wesley Publishing Co., Inc., Reading, Mass., 1963. Carpenter, E. W., Explosion Seismology, Science, 147, 1965. Clark, D. M., Long Range Seismic Measurements -Long Shot, A report prepared f o r the A i r f o r c e Applications Center by U.E.D. Earth Sciences D i v i s i o n , Teledyne, Inc., Alexandria, V i r g i n i a , 1965. Dixon, W. J., and F. J . Massey, J r . , Introduction to S t a t i s t i c a l Analysis, McGraw H i l l Publishing Co., New York, 1957-Ichlkawa, M., and P. W. Basham, Variations i n Short Period Records from Canadian Stations, Can. J . Earth S c i . , 2, 510-5^2, 1965. Weichert, D. H., D i g i t a l Analysis of Mass Spectra, Ph.D. thesis, Department of Geophysics, University of B r i t i s h Columbia, 1965. 89 10. United Kingdom Atomic Energy Authority, the Detection and Recognition of Underground Explosions, Special Report, 1965. 11. Health Sciences Computing F a c i l i t y , U.C.L.A., Biomedical Computer Programs, Los Angeles, 1964. 90 APPENDIX I Com- V e l o c i t y Signal/Noise U n i f i e d Station/Phase ponent Peak-Peak Period Ratio Magnitude A l e r t (ALE) P Z 1 . 0 8 ^ * ^ 0 . 9 15 E 0.345 0.9 10 N 0.480 0.9 8 5.5 PcP Z 0.167 0.9 Baker Lake (BLC) 6.25 p Z 5.70 0 .9 30 E 2.68 0 .9 14 N O.589 0.9 3 Coppermine (CMC) P Z 2.68 0.9 15 E 1.28 0.9 10 N 0.640 0.9 5 PcP Z 0.559 0.9 LR Z 0.25 20 1.5 E 0.20 20 1.5 N 0.20 20 2 Edmonton (EDM) P Z 4 .23 0.9 25 E 2.65 0.9 13 N 1.55 0.9 5 P l i n Plon (PPC) P Z 0.866 0.9 6 E 0.458 0.9 3 N 0.356 0.9 2.5 PcP Z O.658 0.9 E N 0.319 0.9 6.05 (5.09) 6.05 5.6 91 Com- V e l o c i t y Signal/Noise Unified Station/Phase ponent Peak-Peak Period Ratio Magnitude Port St. James (PSJ) P Z 0.4l5Ws«)0.8 2 E 0.356 0.8 2 N 0.328 0.8 2 5-3 Probisher (PBC) P Z 3.74 0.9 6.5 E 1.35 0 .9 3 N 1.77 0 .9 3 6.25 LR Z 0.140 20 0 .3 E 0.14 20 0 .5 (5.17) N 0.14 20 0 .5 Great Whale River (GWC) P Z 9.20 0.9 10 6.7 E 3.38 0 .9 8 N Halifax (HAL) Z l 2.00 1.0 3 6.15 Z2 1.54 1.0 3 London (LND) P Z 3.60 0.9 8 6.5 E Montreal (MNT) N 0.450 0.9 4 Z 1.20 1.0 4.5 E 0.438 1.0 2 N 0.550 1.0 3 6.05 92 Com- V e l o c i t y Signal/Noise Un i f i e d Station/Phase ponent Peak-Peak Period Ratio Magnitude Mould Bay (MBC) Z 2.90^/SEO) 0.9 40 E 1.71 0.9 25 N 0.846 0.9 20 6.1 PcP Z O.389 0.9 E N 0.346 0.9 LR Z E 1.0 15 2 N Ottawa (OTT) Penticton (PNT) Z 1.76 1.2 3 E 1.42 1.2 1.5 N 0.572 1.2 2 6.25 Z 4.02 0.9 17 E 0.480 0.9 8 N 0.625 0.9 9 5.7 PcP Z 3.26 0.9 E 0.252 0.9 N O.788 0.9 Port Hardy (PHC) Z O.667 0.9 1-2 E 0.340 0.9 2 N 0.320 0.9 1-2 5.45 Resolute (RES) P Z 2.67 0.9 32 E 1.56 0.9 22 N 1.29 0.9 21 5.85 LR Z 0.40 20 2 E 0.30 20 1 .5 (5.31) N 0.20 20 1 93 Com- V e l o c i t y Signal/Noise Unified Station/Phase ponent Peak-Peak Period Ratio Magnitude St. John's (STJ) 2 3.42 (y/sec) 1.0 4.5 E 1.50 1.0 2 N 1,15 1.0 1.5 6.4 Scarborough (SCB) S c h e f f e r v i l l e (SCH) P Z E N 1.86" 1.0 8 6.45 5.33 1.0 8 0.833 1.0 5 Sept l i e s (SIC) P Z 6.27 1.1 6 6.7 Seven P a l l s (SPA) P Z 1.35 0.9 3.5 6.0 Shawiningan P a l l s (SHP) P Z 2.70 1.0 5 6.3 V i c t o r i a (VIC) z 0.833 0.8 10 E 2.00 0.8 8 N 1.18 0.8 8 5.75 PcP Z 0.843 0.8 E 1.13 0.8 N 0.704 0.8 Yellowknife (YKC) r ] N 0.423 0.9 3.5 6.0 z 3.08 0.9 10 E I . 8 3 0.9 8 94' Cora- Velocity Signal/Noise Unified Station/Phase ponent Peak-Peak Period Ratio Magnitude Yellowknife (YKC) LR Z 0.75 />/s«) 18 2 E 0.50 18 2 (5.41) Leduc (LD6) N 0.37 18 1.5 Z 4.66 0.8 10 E 1>91 0.8 5 N 1.68 0,8 2 6.05 Rocky Mountain House (RM6) P Z 1.99 0.9 6 E 0.792 0.9 2 N 0,438 0.9 1 5-65 PcP Z 1.20 0.9 Wawa (WAV) P Z 0.752 (est.) 0.9 1-2 E 1.02 0.9 1-2 N 0.702 0.9 1-2 6.1 

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.831.1-0053376/manifest

Comment

Related Items