Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Calhoun: The NPS Institutional Archive Theses and Dissertations Thesis and Dissertation Collection 1984 Design of a freestanding noise measurement and recording system to predict the intensity and location of electromagnetic radiation from earthquakes. Ross, Mickey V. http://hdl.handle.net/10945/19245 DUDLEY KNOX LIBRARY NAVAL POSTGRADUATE SCHOOL MONTERFY, CALIFORNIA 93943 NAVAL POSTGRADUATE SCHOOL Monterey, California THESIS DESIGN OF A FREESTANDING NOISE MEASUREMENT AND RECORDING SYSTEM TO PREDICT THE intensity! AND LOCATION OF ELECTROMAGNETIC RADIATION FROM EARTHQUAKES | by Mickey V. Ross December 1984 Thesis Advisor: R. W. Adler Approved for public release; distribution unlimited T225131 SECURITY CLASSIFICATION OF THIS PAGE (Whtn Data Sntarmd) READ INSTRUCTIONS BEFORE COMPLETING FORM REPORT DOCUMENTATION PAGE 1. 4. REPORT NUMBER 2. GOVT ACCESSION NO. r\Ti,E (and Subtitle) 3. RECIPIENT'S CATALOG NUMBER 5. TYPE OF REPORT 4 PERIOD COVERED Master's Thesis; Design of a Freestanding Noise Measurement and Recording System to Predict the December Intensity and Location of Electromagnetic PERFORMING ORG. REPORT Radiation from Earthquakes 6. 7. AUTHORCa; 8. NUMBER CONTRACT OR GRANT NUMBERf*; Mickey V. Ross 9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT, PROJECT, TASK AREA 4 WORK UNIT NUMBERS 12. REPORT DATE Naval Postgraduate School Monterey, California 93943 11. CONTROLLING OFFICE NAME AND ADDRESS December 1984 Naval Postgraduate School Monterey, California 93943 13. NUMBER OF PAGES 81 14. MONITORING AGENCY NAME ADDRESSC// 4 d///oranr Irom Controlling Olllca) IS. SECURITY CLASS, (of thla report) UNCLASSIFIED t5«. 16. DISTRIBUTION STATEMENT Co/ (h/» DEC LASSIFI CATION/ DOWN GRADING SCHEDULE ReporO Approved for public release; distribution unlimited 17. DISTRIBUTION STATEMENT 18. SUPPLEMENTARY NOTES 19. KEY WORDS (ol the abttract entered In (Continue on reverse aide Block 30, It different Irom Report) neceeaary and Identity by block nutnber) It Noise Measurement; Prediction; Electromagnetic Radiation from Earthquakes 20, ABSTRACT (Continue on reverae aide If neceaaary and Identify by block number) The design, construction, and testing of a freestanding radio frequency (RF) noise measurement and recording system Placement of in the 30 MHz and 150 MHz range is presented. the system along the San Andreas fault to permit the DO , :iZ, 1473 EDITION OF S/N 0102- 1 NOV LF- 014- «5 IS 6601 OBSOLETE SECURITY CLASSIFICATION OF THIS PACE fl»7>»n Data Bntarad) SECURITY CLASSIFICATION OF THIS PAGE (Whmt DMm En(«r«cO establishment of a correlation between increased background RF noise to earthquake fault activity for potential as an earthquake prediction tool is described. S'N 0102- LF- 014- 6601 SECURITY CLASSIFICATION OF THIS PACE(TI7i«n Dmtm Enfrmd) Approved for public release; distribution is unlimited. Design of a Freestanding Noise Measurement and Recording System to Predict the Intensity and location of Electromagnetic Radiation from Earthquakes by Mickey V. Ross Lieutenant, United States Navy B.S., United States Naval Academy, 1977 Submitted in partial fulfillment of the reguirements for the degree of MASTER OF SCIENCE IN ELICTRICAL ENGINEERING from the NAVAL POSTGRADUATE SCHOOL December 1984 ABSTEACT The design, construction, radio frequency (RF) in the 30 MHz and 150 and testing of a freestanding noise measurement and recording system MHz range is presented. the system along the San Andreas lishment of a correlation to earthquake Placement of fault to permit the estab- between increased background RF activity for potential earthquake prediction tool is described. noise fault as an TABLE OF CCNTENTS I. II. INTEODUCTION 11 A. EVOLUTION OF EARTHQUAKE PBEDICTION 11 B. FOCUS OF STUDY 13 BACKGROUND A. B. C. 14 EARTHQUAKE GROUND MOTION SOURCES OF ELECTROMAGNETIC RADIATION IN 14 EARTHQUAKES Insulators and Conductors 1. 15 20 2. Contact Electrification 22 3. Fracture of Quartz-bearinj Rocks 25 4. Porous Rock SOURCES OF ELECTROMAGNETIC PRECURSOR 27 EVENTS 32 1. Electric Fields and Atmospheric Electric Potential 2- III. Magnetic Fields and Animal Behavoir ... 33 ELECTROMAGNETIC PROPOGATION CRITERIA FOR EARTHQUAKES IV. 32 38 A. THE ELECTRIC FIELD EQUATIONS 39 B. SKIN DEPTH AND RADIATED FREQUENCIES 43 C. THE PROPOGATION CRITERION 44 DESCRIPTION OF RESEARCH 46 A. DESIGN GOALS 46 B- SYSTEM COMPONENTS 49 1. Receivers 49 2. Recorders 52 3. Interface Amplifiers 53 Timer 54 V. PERFORMANCE OF SYSTEM 62 VI. CONCLUSIONS AND EECOMMENIATIONS 4. . 68 A. CONCLUSIONS 68 B. RECOMMENDATIONS 69 APPENDIX A: RECEIVER SENSITIVITY DATA AT VHF FREQUENCIES APPENDIX B: 70 ELECTROMAGNETIC EMISSIONS FROM EARTHQUAKES 73 A. COALINGA 73 B. RUSSIA 73 C. JAPAN 74 LIST OF REFERENCES 79 INITIAL DISTRIBUTION LIST 81 LIST OF lABLES I. Time Variations in the Geomagnetic Field II. Order of Magnitude Estimates of the Various 34 Possible Physical Mechanisms for Earthquake-Related Field Anomalies 36 III. Chart Speed and Trace Density 54 IV. Frequency 38.45 MHz RF Irput in uV and Signal Output V- inmV 71 Frequency 150.75 MHz RF Input in uV and Signal Output in mV 72 LIST OF FIGOBES Elastic-Rebound Model of Zarthciuakes Oblique View Across the San Andreas Fault Zone Typical Right- Lateral Offset of Stream Offset 16 is About 450 Feet 17 2.3 Idealized Model of Earthquake Source 18 2.4 Double Couple Source Model of Fault Slippage 2.1 2.2 Similar to Spheres Pulled by Wire Through 2.5 Frictionless Tube Shock-Induced Polarization Signals of Plexiglas Below Transition Pressure 2.7 Shock-Induced Polarization Signals of Plexiglas Above Transition Pressure Shock-Induced Polarization Signals of 2.8 Polystyrene Above and Belcw Transition Pressure Shock-Induced Polarization Signals of 2.6 Polystyrene and Plexiglas as a Function of Shock Pressure 2.9 22 . . 23 Spectral Content of Fracture-Related EM 26 Examples of EM Transients Accompanying Fracture of Piezoelectric Specimens 2.11 21 24 Emission 2.10 19 Electric Double Layer and Velocity Profile in 28 a Capillary 29 Spectrum of the Geomagnetic Field 35 3.1 Cross-Section of Earthquake Fault Line 39 3.2 Electromagnet Vectors at Fault Hypocenter Position of Antennas Along the San Andreas Fault 4 2. 4. 12 8 1 47 U.2 Block Diagram of Data Reception System 48 4.3 VHF Converter Schematic fcr 38.45 MHz 50 4.4 VHF Converter Schematic fcr 150.75 MHz 51 4.5 IF Audio Module Schematic 53 4.6 Eustrak Model 2146 Schematic 55 4.7 Rustrak Model 2146 with Signal Processor Schematic 56 4.8 Single Channel Modificaticn (S-MOD) 57 4.9 Bias Stepper Demodif icaticn 58 4.10 Interface Amplifier with Voltage Generator 4.11 Time Tick Generator 5. Signal Level Data from Noise Measurement System . . 63 5.2 Signal Level Data from Noise Measurement System . . 64 5.3 Signal Level Data from Noise Measurement System . . 65 5.4 Signal Level Data from Noise Measurement System . . 66 5.5 Signal Level Data from Noise Measurement System . . 67 B.I Earthquake Epicenter and B.2 Eseudo 3-D Plot of Doppler Disturbance Path B.3 Electromagnetic Noise Levels at Caucasus Electromagnetic Noise Levels at Sugadaira Electromagnetic Noise Levels at Suginami B. B. .... 59 61 HF Path Locations .... .... 74 75 76 77 78 ACKNOWLEDGEMENTS The author is indebted to a multitude of persons for their assistance and patience during the preparation of this thesis. Foremost, I wish to thank my Lord and Savior, Jesus Christ, whose grace I found sufficient. I and my wish to son, express sincere gratitude to my wife, Mary, confidence and assurance Timothy Joel, for throughout this endeavor. W. Adler, and Dr. also wish to thank Dr. Richard Stephen Jauregui, Jr., whose guidance and support were incessant. Naval Postgraduate School under, I thank all the professors of whom I had classes Finally, CAPT Warren Averill, and their superb support. 10 I wish to the Bullard Hall gang for ^' A. INTRODCCTION EVOLUTION OF EARTHQUAKE PREEICTION The problem involving searching for a reliable precursor luch attention throughout the to earthquakes is being given reality that within minutes With the world. act of nature leaving can trigger an incredibly disasterous thousands and property damage in death tolls in the tens of the anywhere billions an earthquake world. the in The desire for prediction capabilities has evolved into an intense investi- scientific gative activity data was undertaking. collected recently, Until to establish only earthquake has the greatest potential to occur. potentially destructive because they occur seismic in remote, shocks seismic where an Hundreds of escape sparsely populated unnoticed areas of the earth. The present state of the art in earthquake prediction may be compared to that of weather forecasting some 50 years Seismological data acquired continuous and detailed from every area of the world that is subject to seismic disturbances. Various experimental methods for prediction of seismic activity are widely used such as electric and magnetic field, tilt, creep, and strain motion measurements to establish a practical data base [Ref. 1]. ago. Evidence of radio (RF) must be emission associated with precur- sors to earthquake activity is very sparse. Electromagnetic radiation associated with earthquakes appears impulsively [Ref. 2], which suggests that these electromagnetic waves are radiated in a wide range of radio frequencies. The characteristics of the electromagnetic radiation depend greatly, however, on such factors as the magnitude of the earthquake. 11 the depth of the earthquake region, the geology of focus, the distance between epicenter and point of observation, the condition of ionospheric propagation, and others. This is possibly the reason why the intensity of the electromagnetic radiation is different for different earthquakes, with values varying from tens or hundreds of microvolts per meter up to tens of hundreds of volts per meter. Measurements possibility being made are presently existence of the to explore of electro-magnetic the fields. However, there presently appears to be no directed research, or a plausible explanation for, reported emission interference coinciding with radio frequency an increase in earth- quake activity. Derr noted an observation of noise on the FM broadcast band (88-108 MHz) earthquake lights [Eef. 3]. MHz radio noise by J-W. with The reception of earth-based 18 Warwick of the High Altitude at the time of the famous Chilean Observatory, Boulder, Co, has been earthquake still cccuring simultaneously unanswered [Eef. 4]. Research personnel involved in RF interference study, in interviewing citizen band and amateur radio operators in the Hollister, Caarea, noted a consistent report of "increase in backnoise level ground preceding earthquake frequency range of CB bands (50-54 MHz and noticeable (10 (27 MHz) activity" in the and six and two meter amateur 144-148 MHz). Reports typically cited increase in dB) background noise a which returning to normal just before the tremors were felt (one hour to 1 (field minutes) Observations of the electrostatic field mills) have reported that diurnal variations at selected an earthquake preceded by 12 to 24 hours, . sites are accompanied by "high frequency, apparently have no explanation. approached boundary, induced on what whether is actually spiky noise" that Various theories have been occuring at the impact of dissimilar materials polarization have developed 12 the necessary fissure or shock charge density to produce the electric field and specific area surface behaves as a subsequently the constant current generator. B. FOCOS OF STUDY and testing of a freestanding The design, construction, radio frequency (RF) in the 30 MHz and 150 noise measurement and recording system MHz range is presented. the system along the San Andreas lishment of noise a correlation to earthquake fault fault to permit the estab- between increased activity background RF for potential earthquake prediction tool is described. 13 Placement of as an BACKGfiOgND II. A. EARTHQUAKE GROOND MOTION predicting important to Equally earthquake is determining which of is likely the an the many ways the ground the earthquake, to shake during and how long it will shaking will be, occurence of how last. strong the Knowledge of ground motion can make it possible to understand the charac- teristics of the earthquake source. Most measurements made during actual earthquakes at seismological stations are some distance from the source. The localized source of an earthits significance until not observed for quake was the San Francisco earthquake of 1906. recognized that It was earth's crust caused rocks parallel From the elastic Hopkins University had been strained to the fault basis of his geodetic rocks are elastic, fault in the study of revealed that or sheared. observations eminated of earthquakes rebound theory a The classic the earthquake. Reid of John's Harry F. slippage along which asserts and mechanical energy can blocks forming the near the opposite sides of a 5]- fault the motion elastically strains the move by a small amount, rocks that be stored in them just as it is stored in a compressed spring [fief. When two the fault. frictional bond The fails at its becomes large enough. That point of initial rupture, called the hypocenter, may be near the surface or deep below it. The seismic waves radiate from weakest point when the stress the hypocenter in all The point on directions producing the earthquakes. the surface of the earth is the epicenter of the an interval of minutes, earthquake. days 14 above the hypocenter The rocks rebound over or even years. The seismic over ground energy radiated be quite time may at any one small. This slow process is referred to as a seismic slip or creep. generated in zones where the huge which make up the outer layer of plates of the lithosphere, earth's surface, as in Figure 2.1 and Figure 2,2 the A conceptual model [Bef. 6], are shearing past each other. Most earthquakes are process, of the faulting developed from were largely along the San Andreas fault. than 20 severe of California, meter to investigations of earthquakes Over the past 137 years, earthquakes have been recorded to Cape about width from almost vertically and plunges 1.6 KM, The San of California in extends from the Gulf Mendocino varies in more in the state mainly along the San Andreas fault. Andreas fault which the South and Figure 2.4, as in Figure 2.3 a into the earth for 15 to 20 KM. These are some of the reasons why the San Andreas qualifies for conducting an intensive, RF noise increases measurement in show anomalous electromagnetic radiation investigation the intensity of longterm to precursor to earthquakes, B. SOURCES OF ELECTBOHAGNETIC RADIATION IN EARTHQUAKES The emissions of RF are theorized to be caused by the increasing of mechanical pressure at the fault causing shear stresses that convert the insulating rock to semiconductor) by reducing the energy a gap conductor (or between the valence and conduction bands. During the stress increase, microfractures cause shock waves that accelerate the electrons in the conducting bands to radiate both acoustic and electromagnetic energy. If the mechanical forces increase to the yield point, the rock fails in brittle fracture and causes a sudden decrease in the shear forces. The mechanical motion is described by a stick-slip motion that has stable 15 Figure 2. 1 Elastic-Rebound Model of Earthquakes 16 Figure 2,2 Oblique View Across the San Andreas Fault Zone Typical Right-Lateral Offset of Stream Offset is About 4 50 Feet 17 Figure 2.3 Idealized Model of Earthquake Source 18 COUf LE Figure 2.4 Doable Couple Source Model of Fault Slippage Similar to Spheres Pulled by Bire Through Frictionless Tube 19 sliding during stress increases and stick-slip motion during the stress release. Ins ula tors and Conductors 1- radio emission The theory rests can be mechanically stressed that an insulator conductor or semiconductor with to become a subsequent charge migration conducting rocky material from the principle on the insulating rock to the equivalent electric field generated by a shock during brittle fracture of the rock. wave Experimental evidence from the Ballistic Eesearch Laboratory, Aberdeen Proving Ground, Maryland showed that shock-induced polarizaunder the tion is an increasing function of pressure [Ref. 7]. low At pressures shock-induced associated with level of both plexiglas and Shock-induced polarization polystyrene was observed at lower pressures, but the noise experiments prevented the (Ps) to approximately is easily measured down becomes small but 25 Kbar- polarization accurate measurement. At high pressures data indicated transitions Ps more rapid increase, began a state where the suggesting transitions to a more easily plexiglas and polystyrene were polarized by the shock wave. The dielectric constant of the shocked medium, was also an increasing function of pressure. K' were characterized by appreciable scatter, K* The values for but indicated gradual increase up to the transition pressure where a more rapid increase was noted. The relaxation time, 0^ , was a characterized by appreciable scatter, but remained between 0.5 and 1.5 usee pressure, up to the transition also where a rapid decrease began. zation is caused by the motion then the sudden more If it is freely decrease in in the transitions are to a high more ' i of atoms or suggest that pressure phase, fluid state. 20 assumed that polari- atomic groups, motion occurs and that the Greater freedom of motion would result in a larger polarization of the material forces in aggreement with the phase K» at Ps and by both mechanical and electrical the observed increase in both 212 kiloBarx 187 kilobori ^ 112 kiMort 0.2 0.3 0,4 0,S TIMC -psec Shock- Induced Polarization Signals Figure 2.5 of Plexiglas Below Transition Pressure Figures 2.5, 2.6, 2.7, and 2.8 show that over the entire pressure range of the measurements, the polarizatransition. by a physical model involving tion signals can be described three variables, Ps, the computed K' curves are in an d^ It should be noted that better agreement 21 with measured 33 3.2 20 MEASURfO SIGNAL — o--C0M/»U7T0 POINTS 297 (iilobors a. E o I i 1.0 u 0.3 2 0.4. 0.5 Shock- Induced Polarization Signals Figure 2.6 of Plexiglas Above Transition Pressure signals at low pressures than at high pressures where the ends of the measured signals are more distorted by curvature and obliquity of the shock front. 2. Contact El ec t rificat ion Electrification of insulators oil. Slow sliding conductors by experiments conducted by Schnurmann two metals separated by an insulating jerks can proceed by a series of impact are described in and Warlow-Davies of and 22 0\4 Za7 012 Vitoosr^ 'COM(»ur£D POINTS 1^ 010 kitabari I 009 a I »- = 006 <r U kidbarr o \ ac< ^e: Shock-Induced Polarization Signals Figure 2.7 of Polystyrene Above and Below Transition Pressure exhibiting two characteristics, tion. The first that of relaxation oscilla- is a fixed amplitude and the period determined by second is a relaxation time for the building up of a limiting value. The experiments described the amplitude of the jerJcs was observed to decrease when the dielectric a breakdown strength of diminished, or when, of sliding was electrification show a lubricant was with the same lubricant, the velocity the material used as increased. Observations that an electrostatic 23 of contact force resisting 30 10 20 o O-PtEVlGLAS I a -polysty<jcne < 4 O a. I u iGO ;to cc-o 300 PRCSSURE - Xbor Shock-Indaced Polarization Signals of Polystyrene and Plexiglas as a Function of Shock Pressure Figure 2-8 sliding must exist if the boundary layer consists of dielecand this electrostatic compo- tric or semiconducting matter nent is an appreciable part of the force of sliding friction [Ref. 8]. Active fault lines move by relaxation oscillations and can be described by the nonlinear stick-slip motion that is generated than the stress, when the static force of friction is larger kinetic frictional value. During the increasing the motion of the fault proceeds uniformly by creep 24 insulating converts the stress the large and rock to a conductor or semiconductor. On reaching the brittle fracture stress, the material unloads and the stress is reduced and results in shock wave retreating a shock vave has The with free electrons conductor ahead of it a material. in the insulator conduction band behind the that can fall into the retreating shock front. The charges can accumulate by succesive creep-slip motions to cause the electronic charge to produce an electrical field beyond the breakdown value in an arc with the attendant causing the air or rock surface, radio emissions. Fracture of Quartz-bearing Rocks 3- geological research The believed Laboratory were Scientific group the Los at to Alamos first be the to detect radio frequencies when guartz-bearing rocks and other hard piezoelectric experiments. Using placed about 5 to a standard RF coil with 50 CM from the recorder it was found that small transients with shown produce single crystals maline in a fractured in materials were Figure 2-9 Rock . ferrite core a sample and transient a cracks in quartz and tour- exponentially decaying constant of about time laboratory 10~5 samples were seconds as fractured applying uniaxial load between two parallel steel by pressing a small sample. RF by anvils or steel ball against a flat surface of the The signal developed on the coil was amplified by a wideband amplifier and fed into one channel of the transient recorder. All measurements were performed under ambient conditions. Also the samples were not treated in any way except for cutting or coring [Eef. 9]. The fact that the fracture related EM observed only in guartz-bearing quartz rocks (frictional electricity) and rules emission was tourmaline crystals out triboelectric and in effects as the source of this emission and indicates that it is related to piezoelectricity. 25 The most o to >- q: t- a or < a 3 a. > < FREQUENCY (MMj) Spectral content of fracture-related Fig. 2. EM enisKion from (mean grain size shown in a. coarse-yrain granite (15 mn) b. brackets): ; granodiorite (2 ran); c. fme-prain cranite (from Westerly, R. I., 0.5 mm); d. fine-grain quartzite (0.1 rcr.) e. a 2 x 1 x 0.5 cm toui-maline single crystal cracked by applying The curves show pressure along the c axis. averaRCS of 10 spectra for each sample. ; Figure 2.9 Spectral Content of Fracture-Belated EM Emission 26 for this emission probable mechanism the piezoelectric occurs as shown in Figure hypothesis is supported of the since regardless experiments) and independent of load, distribution within the higher than the average (much the ampli- the local stress exceeds a sample fracture occurs only when specific value This . signal is the stress of 2.9 by the observation that resulting RF drop in release of field accompanying the sudden stress when fracture tude is the rapid therefore drop in stress stress in the is essentially independent of load. compares the spectral content Figure 2.10 of frac- several samples, obtained by averaging 10 different spectra for each sample. It is clear ture related EM emission from that a shift to higher frequency occurs as the grain size of quart z- bearing rocks decrease. Detection and analysis signals was very difficult at With regard to RF emissions outside of substantial increase fracture-related RF distances greater than 50 cm. conclusions could be made from a of the laboratory, no the study concerning whether in the size of the fracture results corresponding increase in the emitted RF energy, thereby permitting detection at larger distances. also in a 4. Porous Rock Electrokinet ic phenomera induced by ground water flow associated with earthquakes are believed to provide a possible means of earthquake prediction. Detectable variations of earth currents, electric potential and geomagnetic field may be caused by diffusion of fluid into a dilatant (wider or larger) focal region- The electrokinetic phenomena controlled by the processes of mineralized water migration through microcapillary rock channels are an source of charge separation and relaxation. 27 independent The charge S M**C 2 11 («c i^My>v-^ 500u»tc Fig. 1. Examples of EM transients acco~panyinf, fracture of piezoelectric spcci.-ier..s (tl'.e digital sampling interval is given in brackets): a. small crack in a quartz single crystal (50 ns); b. cr^ll crack in a tourmaline single crystal (10 ns); c. failure of a specinen of fine grain sandstone (200 ns) d. failure of a qu-irtzite specinen, showing also the accoinpanyin?; crop load (2C00 ns, see text). Note the different time scale for each transient. ; m Figure 2.10 Exao] lies of EH Transients Accompanying Fracture oj Piezoelectric Specimens ; 28 separation is due to the selective absorbtion the solutions on the fracture of ions from surfaces and the formation of double electric layers. -Solid-Liquid Interface Electric Ooubte Layer Figure 2. 1 1 Electric Double Layer and Velocity Profile in a Capillary The double layer is made firmly held to the solid, up of a layer of ions and a more which are diffuse mobile layer extending into the liquid phase- The resultant charge of the diffuse layer is equal in magnitude but of opposite sign to that of the firmly held A layer. simplified schematic diagram of the structure of the double layer and profiles of the electric potential and velocity of liquid in a capillary shown in Figure 2.11 . Because of the electrical charges, there is a difference of electrical potential between the is 29 solid-liquid interface and the bulk of the liquid. This is When called the electrokinetic potential or zeta potential. electromotive force (EMF) is applied to an electric an double layer, there will be a displacement of the oppositely If the solid cannot charged layers relative to one another. double layer also is unable to move, the fixed part of the move; the application of an EMP will result in movement of the diffuse layered ions and liquid [Ref. 10]. Applied EMF and viscous force is given by: j = ie"^ where j //| ) grad (2.1) is the fluid flow flux the viscosity of the through a capillary (volume) and per unit area, per unit time is E "^ fluid, is H the zeta potential, and ^ is the dielectric constant of the fluid. The streaming potential difference when potential (E) liquid a is the production of a is forced through a porous medium or a capillary tube. It is given by grad where E = - and (^^ P /PjCT) grad (2.2) E are the electrical conductivity and the pres- sure of the fluid, respectively. Since all electrokinetic processes are irreversible, thermodynamics of an irreversible process can be applied. most general relations between the electric The current i and the fluid flow j, and the forces of grad E and the grad P are -i = 111 grad E + L12 grad P (2.3) -j = L21 grad E + L22 grad P (2.4) 30 where Lij are constants. Explicit expressions of Lij's for a capillary tube are: Cr L11 = (2.5) L12 = 121 = L22 = k* (f^ "^ /A (2.6) {2.1) /f\ where k* is the specific permeatility of the capillary tube. For porous media k is given by k = ^k* where (2.8) is the porosity of the porous media.. The basic equa- <^ tions of fluid flow in the porous medium are -I = <)> -J = - cr gr ad E - {(^e ^ / ) {^C^/f) grad E grad P ) (k /.I + ) (2.9) grad P that the electrical conductivity It is assumed substance is much smaller than ^cT (2. 10) of a matrix . The maximum magnitude of the current density associ- ated with the water flow is obtained by setting grad -I = {(^6^ / 1) ) grad P E = ; (2. 11) This shows the electric current is proportional to the pore pressure gradient (the pore pressure is assumed to change by 10 - 1000 bars following the in the dilatant earthquake) . If focal region the mean preceding or gradient of pore pressure is approximated to be grad P = 1 '^ 100 bar/Km (2.12) 31 the electric current induced is = 0.7 X lli (10^ - A/m2 (2.13) flow exists, of detecting the water If the capability direction 10^) of water flow can te consequently and known, the The earth current earthquake. predict the epicenter of the due to electrokinetic effects yields variations of the geomagnetic field on the surface. horizontal a in the current flows uniformly If the magnetic field caused by circular cylinder of radius R, the current is simply obtained by Biot and Savart's law, H = / 2Ttr) (I • TfrRS (2. 14) the distance from the center of where r is the cylinder to the point at which the magnetic field is observed. r = R = L / 2 If (2. 15) then the field strength becomes H = 0.2 ^ 200 (2. 16) y where the field strength is independent of R and L for this Interestingly enough most observed magnitudes of magnetic anomalies associated with earthquakes are assumed geometry. almost all in the range estimated above. C. SODBCES OF ELECTBCMAGNETIC lEECORSOR EVENTS '^ • Electric At F ie lds and Atmcspheric Elect ric Potential the earthquake moment and just before it the effects due to disturbances of natural electromagnetic earth 32 field were observed (NEMEF) These repeatedly. effects quasi-stationary disturbances of the atmospheric at distances up to electric potential (AEP) up to 1000 V/m, some hundred kilometers from the epicenter. In the course of observations the AEP perturbations were noted some field involve or two hours earthquake moment within one hours before the period [Bef . 11]. AEP disturbances charged surface estimated for the are: ^ = AEf C-/ 2 (2.17) 6k6^ where Cr = q N (2.18) § Under the assumptions that the linear charge density q, surface density N, and the dislocation path length are: q = N = 10-1* - 10-10 109 - 10*1 AEP disturbances c/m (2.19) m2 of the (2. 20) 100 - V/m can be expected 1000 in the final stage of the earthquake precursor period. 2. Magnetic Fields and Aniital Behayoir The amplitude and time duration of the most charac- teristic time varying geomagnetic disturbances occurring at the earth's surface are listed in Table presenting the same data is in total magnetic 2.12 [Bef. 12], plotted as a field. Such a a . Another way of plot of the spectrum of the spectrum is shown where skin depth function of frequency 33 I (Z = f 500 in Figure (c-f)"^'^) and conductivity cr . is It TABLE I Time Variations in the Geomagnetic Field Typical Magnitude Description of Phenomena Characteristic Period (gamma) (sees) 10*2 Sub-audio frequency fluctuations Micro-pulsations 10° 0.0 sec~^ 0.1 Giant pulsations Sudden commencements 102 1 10 20 Crochets 20 Bay Phenomena 100 10* Diurnal variation 30 106 Storm Time effects 10 Twenty-faeven day recurrence effects 8 Annual variation 10^ 10 Solar cycle secular variations is evident that semi-diurnal, sun) the most dominant signals 25 year"* are the diurnal, the rotation of the twenty-seven day (due to and high frequency changes that are, in general, nons- tationary, that is net frequency stable. This spectrum could represent also a crude measure of discrimination limits for detection by animals. It is, of course, possible that animals can somehow understand these variations well enough to detect changes well Schemes for doing this are below the background variations. necessary also for field experi- ments presently in operation on the San Andreas Fault where information on the large scale of these distrubances is used to enhance detection capability for 34 local magnetic field SKIN DEPTH (km) ^ u o >- o UJ IS s a S (yna^ Figure 2.12 ^-mns) asMOd Spectrum of the Geomagnetic Field 35 If animals can detect effects occur- variations [Ref- 13]. an earthquake, ring days before systematic charge effects involve spectrum over a possible that these it is which would enhance the critical frequency range. There are a number of physical mechanisms that might be considered sources as earthquake-related for animal TABLE II Order of Magnitude Estimates of the Various Possible Physical Hechanisms for Earthquake-Related Field Anomalies L. MAGNETO HYDRODYNAMIC ^^XNtvcEt, ~ Bo£r>v (cr* = conductivity, D = length scale) /•V d ^= B. - 0.1 gamma x 10"'', ^ v = velocity, TELLURIC CURRENTS ^^XNt^'cec^' cr^ / ^ ~ 0.01 gamma (cr = conductivity, E = gecelectric field, D = depth of current flow) C. PIEZO MAGNETIC EFFECT AB = 40 S I - gamma stress change, 1= magnetization) (S = D. 1 LOCALIZED CRACKS AND FRACTURES AB behavior. = ? Table II lists several order of magnitude estimates of of these together with their amplitudes. It would appear that for these effects at least, it is difficult to generate source amplitudes of more than a few tens of gammas. The last process suggested which involves this 36 concerns effects of the relatively unkno^jn local surface electrical crac]<s and possit3le exception to this. 37 and magnetic fractures might be a III. ELECTBOHAGHETIC PROPOGATICN CRITEEIA POR EARTH^OAKES fault line becomes A geological and electromagnetic thermal relative to plates move fault shown line, Figure 3.1 in than in the source radiations each ether. compressive stress and the greater here a is of acoustic, when the tectonic The region near the in both shear and relative permittivity, (c/--] ' ^^ unstressed region having permit- tivity ^^ . The fault width of 2Yo is the region of uniform stress and the regions beyond this point are unstressed, When fault line motion commences, causing current flow across voltage is generated interface an The interface different materials- a can be between two formed by the following means: a. Forces near 100 Kbars can convert the insulating rock to a semiconductor (or conductor) in solid form [Ref, 14] so that charges can flow at the insulating-semiconductor inter- also the voltage generating the work functions differ in the two media. face which is b. Forces above produce a 100 Kbars may cause plastic rock flew to liquid-solid interface work functions mechanism since [Ref. causing current flow 15] with and charge differing storage in the solid portion having insulating properties. c. Shock waves can distort the dipole fields in the rock to produce a polarization in the material and increase the permittivity by nearly an order of magnitude [Ref. 16]. The transient charge is stored on the lower permittivity regions and accumulates with subsequent shocks. 38 E FXELt XV Ari^ ^. \ Su/?FAc£ > 1 > TO fftuLT (ZAjcrtEtJTj Figure 3.1 Cross-Section of Earthquake Fault Line 39 THE ELECTRIC FIELD EQOATIONS A. The source of electromagnetic the radiations is accumulated electrons that were discharge of the shock waves or increasing pressures on the fault. uous corona discharge can occur from the formed by A contin- nonuniform fields the rock or a sudden discharge that form on sharp points of exceeds the breakdown strength of can result when the field the rocks in uniform or nonuniform fields. If the breakdown radio waves in the megahertz region occurs in microseconds, are radiated. the This increasing pressure at a fault can be related to the strength and frequency spectrum of radiations which can be correlated to the acoustic data presently being recorded at the earthquake sites. hypocenter, shovin in The fault belov the surface and is the source radiation, shown as a point source. ably elliptically polarized depicted for clarity. Figure 3.1 is usually of the electromagnetic The emissions are prob- but only the parallel Figure 3.2 shows the field is electric field volts/meter in the incident X-Y plane and the magnetic field vector, H in amperes/meter, normal to the incident plane. The energy propagates on the transverse mode vector, (TEM) E in and the Poynting vector, P in watts/sguare meter, on the X axis. The fault location of epicenter is shown in the hypocenter at Figure 3.1 and the fault surface is the where the electromagnetic radiations are a maximum. The electromagnetic radiaticns can be obtained from the retarded potential method but the objective is to only indicate the attenuation of the field with distance which can be found from the Maxwellian equations: VxH=CrE+6aE/at (3.1) V (3.2) X E = -/^^H / ^t 40 /N X -^y Vbt-Ts Figure 3.2 /merBt Electromagnet Yectors at Fault Hypocenter 41 The parallel vector mode, E where the E field is in the incident plane, can tie obtained from the vector equations to obtain the scalar differential equations: aHz / ^x = - 3Ey / - = <3t /at e c?Ey crEy (3.3) ^^Hz /^t (3.4) Let the arc discharge have the spectrum Oj = 2 T' If E = Eo« f- / 3iEy ^x - e"^ , ^2 Ey (x,t) f1 < f < f2 and the scalar field reduces to: = (3.5) where Y .2 -j^^" = • Assume ^ p6: (3.6) scalar solution with E a = Eo« e • e "^'^ This gives the phase shift and attenuation: ^ = Kjjto)'^ = j(//^u)'5 (J 0-/^6 -1) (3.7) {1-jcr/,.,^),5 (3.8) = j(/^) 'S(1-jcr/ 2a>6+ ...) - j(//^) - j(^^)'5 '5{l-j cr/ 2^e) + (3.9) (3.10) cr'/2 ^fc (3-11) The radiating field has the form: , E = Eo e i -^ '^e .., -^ ^ , (3.12) e-" 42 SKIN DEPTH AND RADIATED FBECOENCIES B- Figure In 3.1, fault the line region which has a constant compressive and shear stress is region one, and the designated region away from fault line, unstressed region, and conductivities permittivities The two. in the two regions are: D7=Zri- (3-14) ^;,= ^^^^^ (3.15) ui (3.16) ir^= Where £,^ permittivity of free space is the are the relative permittivities. and 6r The skin depth, %, and^^ is the distance where the radiation is reduced by the factor of 1/e from the initial value. The skin depth can be obtained from equations: % = 2cjG^ /c- = 2^6f6/cr; (3.17) Using: f = SxlO'^ Hz, 6^ = 25.0, ^ e^= S Since the skin depth is lOOTx = = 106 Hz, /j; =10-5^ (3. 18) 8.8 x10"i2 (3.19) 13.9 Km (3.20) San Andreas fault is less than this value 43 about 15 Km deep, which indicates the that under the worst case of the hypccenter at the extreme depth, the electrical field is reduced slightly more than 1/e, but easily measured at the epicenter. The values and conductivity of permittivity frequency and the upper limit of propogating be estimated from the graph by Kraus Using the worst case of urban 15 Km frequency can and Carver [Bef- 17]. ground at 100 megahertz, the and if the hypocenter is halfway skin depth is about 0.8 Km down the vary with the attenuation is tectonic plates, This limits the radiation to the range X 5 10^ < f < e"^.^. lO^ Hz. THE PEOPOGATION CBITEEION C. source at the hypocenter Figure 3.1 shows the radiation and the parallel to electric field, the incident plane, makes an incident angle 9i with the normal to the fault. The reflected stressed wave is angle at an region has a width Or with and 2Yo regions are insulating and lossless, ^r the normal. > ^r • ^f The "tiie the reflection coeffi- cient is given by: f*= Er / Ei (3.21) rr The reflected wave, when the (3.22) 1 radical in the Er, equals the incident wave, numerator vanishes. Ei, The critical value of the incident angle for complete reflection is: sin Gic = (^2/6,) This relation '5 = (3.23) (6;3/^^^),5 obtains for either perpendicular or parallel polarization of the electric field and for all <c > ^ic. Using ^r ~ 25.0 for the stressed region and 6 r^ = 5.0 for the unstressed portion, gives Oi = sin-i (0.2) .5 = sin "MO- Since the critical angle is ^5) -'" 26 small, (3.24) the majority of the radio emissions are reflected from the unstressed region for hypocenters deep in the fault. emission may account for the This wide region abnormal human of radio and animal behavior distant from the fault line but within the stressed fault region. epicenter When the exceeds confirms the field in the 30 Kv/cm , reports of lightning earthquakes. 45 the air space air breaks before and above the down and during large IV, A- DESCRIPTION OF EESEARCH DESIGN GOALS One of the most important problems associated with using electromagnetic radiation as a precursor of earth quakes is select the point of observation. Placement of the how to system far from industrial areas, but near potential earthquake activity was potential site desired. A preliminary field locations was conducted under survey of the direction Geological Survey. Site location along the San Andreas fault was chosen of the United States Department of the Interior, historically and recently it being known because of The chcsen observation site is 1000 active earthquake area. feet from the San Andreas Rift 23.5' and 36° 45.5' W as an N Zone which is located at 121° three miles east of San. (fault) Juan Bautista and seven miles scuth of Hollister. It became very important to the system cies for have any United States identified the Western project interference from The Army Frequency Coordinator for so not to transmitted radio waves. secure spot window frequen- bands VHF frequencies, (38.45 MHz with a 50 KHz and and assigned 150.75 bandwidth, Mhz) to this . These have been allocated with exclusive use for a period of one year. For observation of signal intensity and direction in the VHF range, a sjstem was positioned with its data recording directional sensors (antennas) fault as shown in Figure 4.1 detect both horizontally and looking both up and down the . vertically polarized emissions at both reserved frequencies. (two) are erected The antennas (eight Yagi's) The omnidirectional antennas for detection at both with vertical polarization. connected to a detector frequencies, Each of the ten antennas are (communications receiver) 46 and A,B,D, and E have 2 antennas each. One is vertically polarized and the other is horizontally polarized. There are two omni antennas vertically polarized at C. Figure 4.1 Position of Antennas Along the San Andreas Fault 47 I\N\ Tx.\sA^^ 1 Figure 4.2 Block Diagram cf Data Reception Systea 43 Low cost modularized receivers were procured and modi- assigned frequencies. The output of noise level the receiver is offset to nullify any ambient collect any and is fed into slow speed chart recorder to detected emissions as a function of time. A block diagram of fied to receive at the is illustrated in Figure the measurement system 4.2 . The noise measurement system can operate from 12 volt DC storage batteries or a power supply. Electromagnetic batteries have to be replaced periodically. emission data collected through this to seismic guake provided data and storage The chart paper system will be related the by Onited States Geological Survey Office, Palo Alto, Ca. B. SYSTEM COMPONENTS gec eiv ers 1 . The VHF Converter Module is convert the specific frequency to converter is supplied with (IF) transformer for a designed to amplify and the HF range. 10.7 MHz intermediate frequency the mixer output circuit- output circuit consists of a The slug tuned coil, The mixer a capacitive voltage divider, and an RCA jack to provide a 50 ohm output. Figures 4.3,4.4 show schematic diagrams of the VHF Converter Modules. The IF Audio Module comprises a highly sensitive and selective IF amplifier, squelch system for use audio amplifier, with the converter to make a communications receiver. The selectivity of the unit tailored to operate in the noise measurement system. unit has automatic gain control (AGC) of both IF and AM detector, with the EF AGC being delayed f cr best response. log detector provides drive fcr the Eustrak recorder) output. 49 "S" meter and VHF was The EF, An integral (signal to u. -^yi> t3. Q 41- -^-\w '^ ^^ **. -'SN SI 'Mnf-t> Vi Figure 4.3 VHP Converter Schematic for 38.45 MHz 50 2s U + ^ -1^ 41- [-?><n, lJ:s^ -nnm_^ ^^-i> ^ Ml- ^)-^ '^Ob^iA^ < '-iKu^ —\)—^-l> I II L © ^UJUL/ ^aM> -J Figure 4,4 7HF Converter Schematic for 150.75 MHz 51 amplified three fcy converted to signal is 10.7 MHz The stages. transistors biased from an 455 KHz Each stage consists and of several integral zener regulated source. coupled three stage IF filter provides selecFigure 4.5 tivity against adjacent channel interference. shows a schematic diagram of the IF Audio Module. Appendix An optimally A Table IV and Table contains 1, signal input voltage at both output voltages for a given RF VHF assigned receiver S-meter 2 • frequencies. were determined receiver signal data of output capabilities The from this information of the on the (output signal to the chart recorder). Recorders The Rustrak DC recorder prints through the impinging action of its stylus pressure-sensitive series of driven ty chart paper. dots appearing as a speed varies with motor speed. depend on the ratio of couples the paper the chopper bar against presentation is a continuous line. Writing Its Chart speed and trace density the interchangeable gear drive to the motor. chart speed and trace density. Table Figure 4.6 is box which III shows the schematic of a the Model 2146 dual trace chart recorder and Figure 4.7 is a schematic of the same model with a signal processor included as shown. Both versions of Model 2146 had to be modified for single channel use. The single channel modification involved bypassing a switch (S-2) by putting a small jumper wire between the normally closed (N/C) (N/0) terminals recorders have as shown in Figure and the normally open 4.8 . Three of signal processors to desensitize input to one volt per division. demodified by rearranging the the Jones plug making non-bias (S-MOD) the the signal The desensitizer board was bias stepper connections at stepper connections as shown in Figure 4.9 There were no modifications necessary for . the Model 388 chart recorder. 52 i iTi^^ go Q-~ — --> ih-^ :s^ I ^^. I T -xD-y-t> 41— u. 3 K> -ff- -1 IhqIt-J J >o Figure 4.5 IF Audio Module Schematic 53 TABLE III Chart Speed and Trace Density Model Drive Writing Motor Speed 2 6 30:1 10; 1 Gear Train No. Gear Ratio (RPM) 2194 1 1 4 388 2 1 2 strike/ seconds strike/ seconds Chart Speed 76 mm/hr Chart Speed 252 hrs 2 Paper Duration Chart Speed Chart Speed Paper Duration in/hr 50 mm/hr 378 hrs Trace Density 3. in/hr 3 900 300 Strikes/Inch Interf ace Amplifiers By design the receiver output voltages were approxi- mately between 280 mV to 320 m7 with RF inputs from approximately 10 uV to 100 uV, Ideally this 40 mV range would drive the recorder from zero to full amp voltage offset scale, only 80 % scale but because of an op readings were achieved. interface amplifier was designed to set full scale minimum and maximum needle movement in the chart recorder. Calibration of the meter movement was accomplished by variable resistor settings as shown in the schematic of Figure An 4.10 . 4 . Timer To compare seismological disturbance data with the noise measurement site recordings accurate time needed to be 54 Figure 4.6 Bastrak Mcdel 2146 Schematic 55 Figure 4-7 Rustrak Model 2146 with Signal Processor Schematic 56 si^^xxc-H [jNr^^oc^TFrE:b. f-TO ^To TanjeS. PXAi &fi vj^I-'/O •- . WH//«£0-mx — TO i± t*± « i SW N/C ^H/fLe^ n/o j t^fc 4^-^ COG AS PAPE^ N/ /yXohZFTEh \ '^yrrrc'r\ r T^ SW 4 — TO Toues 8^4. Ptat i*4- wH/«ECi v^H/»t£^-*-ei.>c (X T^O GrAL'/O Figure t».8 Single Channel Modificatioa (S-MOD) 57 t/?/^i^ dm^ sra?o&Q^ XJJPor I II5^ '"^ •••. *\^ y ,u — ^ -• y >- ^!.'^ r-i >•' r-\ i_ "1 "" I -1^ Figure 4.9 Bias Stepper Demodification 58 CO _!' Figure 4.10 Interface Amplifier with Voltage Generator 59 Because of indicated on the chart paper. ultra-low dissipation and operation a CMOS 21 stage conjunction with counters (RCA counter timer CMOS 8 CD40103B) (BCA requirement for a using a battery source, CD4045A) was used in stage presettable synchronous down CMOS dual J-K master-slave and a flip-flop (ECA CDa027B) to obtain the time. The 2.097152 MHz crystal establishes the one second clock pulse as shown in Figure 4.11 . are set using The time and the length of the pulse duration 8-bit down counters. When the timer pulses, fifteen minute pulse will last for six seconds. A one hour pulse will last twenty-four hour pulse will last for twelve seconds and a Time can be initialized as necessary for eighteen seconds. the length of the pulse indicates the time. by the reset switch. 60 A Figure 4.11 Time Tick Generator 61 V. PEHFORHANCE OF SYSTEM The informational parameter of is variations in the level a possible seismic event of the electromagnetic radiation envelope or in the number of radiation impulses whose ampliexceeds a tude noise discrimination threshold. pre-chosen measurement system range of 0.1 uV to 100 uV is designed EF input. with a sensitivity The system was given a the roof of Spanagel Hall "shakedown test" on The at the Naval Postgraduate School to see if it was sensitive to variations in the level of electromagnetic radiation based on a 24 hour period. Along the San Andreas Fault, it is expected that varia- tions in signal level will occur due to seasonal changes. during autumn and winter the highest signal For instance, level summer, the 2U falls into the sunless part the day. Spring, autumn have two maxima during and the beginning of hour period, of the nightly and the post-noon ones. Figures 5.1, 5.2, and 5.5 are recordings which 5.3, 5.4, show the noise levels over a 24 hour period. 62 I I ^f; ;— --r-/ .1— i w> , Ifh, ; ! :t:\/ -9^ PS" > \ I i r T I h y- : 1 I ; 1 I "" ! I T I I I l -!-; I "rr'T- 15--:q < I I 1;^ I ! j I ^^ M )M: )i 41 I s.-:^ ,_..-.^ I 4- I r i---^--^ i I ! I ! I I n: ^^ !' i c 1 ! I I ! U-. I \ til ill I I 5^ -T-T-p-rTl 1 H ! ! I I R > \ »t I — n -> I i Xr T— •I t- A ! i i" 'i -If- r t I I -r—ti -r- ! ^ : r^ -J, :t Figure 5-1 i V n -f-9'c u "Y" t I ! I Signal Level Data from Noise Measurement System 63 Figure 5.2 Signal Level Data from Noise Measureaent System 64 1 1 ' ! 1 ! ! n^: 1 •r H i 1 ! 1 ! 1 1 Figure 5.3 Signal Level Data from Noise Measurement System 65 "l ^1- V ^ ._r^ —p-^T-: M ' . ' ' 'I- 'I cy —• i~~: ^%i ^ 1^ ' - I ' I ! I J I : ' I r: I o i :-:-..: ^ ' — 1--or I ! I 4 --r- :r^ I I .; ---] t -Tl 1 .L:.4..-.i.^- ! •! I I : -I i -^-r •> ; o I I.' I -1 d ^ "^ i I ^ I ^ I I T «. --4- t~-r — .r I I I > % -y-V ..- ,~>i ._^ V N O I J, ... ^ r i I <- Figure 5.4 '^ 5 "VM I i - —-,- !-" -r-i-ri'-'' -T'T-^i _u_j_4-'— -1-4.-1- ^ n n Signal Level Data from Noise Measurement System 66 Figure 5.5 Signal Level Data from Noise Measurement System 67 VI. A. CONCLDSIONS AND BECQMflENDATIQMS CONCIOSIONS The design, construction, and testing of an RF noise measurement and recording system in the 30 MHz range has been described. Test results show that background noise levels can be received and microvolts volts. Theory has to recorded from and 150 MHz tenths of shown that anomalous elec- radiation levels from earthquakes can occur. only detect the The noise measurement system data will not presence of increased background noise levels at the desigtromagnetic nated frequencies, but it may also give an indication of the direction of reception relative to the San Andreas Fault due to the polarization of the antennas. Besides physically erecting fault, the most difficult problem along the was designing a receiver/ the antennas recorder system that would be sensitive to slight changes in background noise level at frequencies near those of citizen band and amateur radio operators. The signal output of the receiver was sent through an interface amplifier and the chart recorders modified to accomplish thisAppendix B presents electromagnetic emissions and other observations possibly related to earthquake precusor activity. Measurements of electromagnetic radiation from earthquakes were detected by changes in background noise levels, but in the very low, low, and high frequency ranges. The results radiation of this research suggest measurements has potential prediction tool. 68 that electromagnetic as an earthquake B. EICOMMENDATIONS acquired from Data should be noise taken for the noise measurement V EF system observing background Since seismological data can be at least level variatiocs. a year acquired from the Geological Survey Office, any significant earthquake in the area of the noise measurement site should be immediately compared with the recorded data. Otherwise periodic basis. studies were presented data can be examined on a concerning the mechanism by which rocks emit radiowave energy when they are crushed, further investigation is needed in this area and Although several the following: 1. The relation between the intensity of the background noise in the epicenter. The VHF range and antennas of location of the the noise measurement the system are looking up and down the fault which makes location probable. 2- The relation between the VHF radio 3. intensity of the precursor noise and the earthquake's magnitude and depth of focus. The data obtained may show similar variations in background noise levels at 38 MHz or 150 MHz or both for earthquakes of similar magnitude and depth of focus, Long term data collection is needed obtain to through order-of-magnitude calculations, of VHF radio noise radiated 69 the amount in these events. APPENDIX A EECEIVEB SENSITIVITY DATA AT VHF FBEQDENCIES 70 TABLE 17 Frequency 3 8.45 MHz BF Input in uV ana Signal Output in mV RF S RF S RF S RF- S IN OUT IN OUT IN OUT IN OUT 0.0 0.268 1.8 0.227 3 0.306 300 0.327 0. 1 0.220 2.0 0.228 40 0.311 350 0.328 0.2 0.220 2.5 0.233 50 0.315 400 0.329 0.3 0.220 3.0 0.237 60 0.317 450 0.330 0.4 0.221 3.5 0.241 70 0.318 500 0.331 0.5 0.221 4.0 0.246 80 0.319 600 0.332 0.6 0.221 4.5 0.249 90 0.320 700 0.333 0.7 0.221 5.0 0.253 100 0.320 800 0.333 0.8 0.221 6.0 0.262 120 0.321 900 0.333 0.9 0. 221 7.0 0.267 140 0.321 1 .0 mV 0.334 1.0 0.222 8.0 0.271 160 0.322 1 .4 mV 0.334 .2 0.223 9.0 0.275 180 0.323 1 .6 mV 0.334 1.4 0.224 10 0.279 200 0.323 2 .0 mV 0.334 1.6 0.225 20 0.296 250 0.326 MAX 0.335 1 71 TABLE 7 Frequency 150.75 MHz HF Input in a7 ana Signal Output in mV RF S RF S RF S RF- S IN OUT IN OUT IN OUT IN OUT 0.0 0.29 1 .8 0.296 30 0.322 300 0.342 0. 1 0.226 2.0 0.298 40 0.324 350 0.344 0.2 0.237 2.5 0.304 50 0.326 400 0.345 0.3 0.246 3.0 0.307 60 0.327 450 0.346 0.4 0.255 3.5 0.310 70 0.329 500 0.347 0.5 0.263 4.0 0.311 80 0,330 600 0.348 0.6 0.269 4.5 0.312 90 0.322 700 0.349 0.7 C.274 5.0 0.313 100 0.322 800 0.350 0.8 0.279 6.0 0.314 120 0.334 900 0.351 0.9 0.280 7.0 0.315 140 0.335 1 .0 mV 0.352 .0 0.284 8.0 0.315 160 0.336 1 .4 mV 0.355 . 6 1 : ' : ! • 1.2 0.287 9.0 0.316 180 0.337 1 mV 0.355 1.4 0.290 10 0.317 200 0.338 2.0 mV 0.356 .6 0.293 20 0.319 250 0.340 1 - MAX 0.358 J 72 APPENDIX B ELECTBOMAGNETIC EMISSIONS FROM EARTHQUAKES A. COALINGA ionospheric perturbation An Coalinga earthquake of network of California. These by a northern and 285 km as shown in Figure by a ground-motion-induced acoustic in Figure B.2, that propagated to ionos- observations appear to be ionospheric disturbances earthquake epicenter in the regions of all five from the epicenter, were all affected as shown detected at distances between 160 paths, pheric heights. B. links The ionospheric refraction (horizontal range) pulse, was 1983, radio frequency high HF propagation B.I, May, 2 produced by that was to be [fief. reported this the first close to an 18]. RUSSIA Figure B.3 illustrates one example of the anomalous electromagnetic radiation possibly related to earthquake activity on September 16, 1978. In this case, the epicenter of earthquake was located in central Iran, and the receiving equipment was set up in a tunnel 50 m below the increase in ground surface at on-set time Caucasus, USSR, 1200 Km of electromagnetic radiation at distant. The frequencies of KHz and 1.63 MHz was about 30 minutes before the main shock with magnitude of 7.4, as shown in Figure B.3 27 [Ref. 2]. 73 Figure B.I Earthquake Epicenter and HP Path Locations 74 souu3|v;s(ri !.- I «- ^^^'A s 2C;2 V-_1T^- f\... . ^v*Vy <<^^??v^ "/; "' r'-.vir-';'— ''-- > 'v^ •*'--'-2 -^-'-i-i /J-:/''' >^'V Mi r Figure B.2 Pseudo 3-D Plot 75 cf Doppler Disturbance Path CAUCASUS SEPTEMBER II. H7« J7KHI l.llMHzt ^^^^rvn/^u- H SISKHzli- ± 28 5 ^^=1 SEISMO-h GRAPHtU:*SZ is.'Daz ii:zsz "•"2 ,^ = ,.4 1»:»»2 ^ UT t Electromagnetic Noise Levels at Caucasus Figure B-3 C. JAPAH On March 31, Standard Time 1980, (JST) , an earthquake occurred at The magnitude was 7, 1633 Japan and the epicenter was located at 35.40 n and 135-30 E in the Kyoto prefectureThe focus of this earthquake was very deep, at approximately U80 Km. The distance between the location of the station and epicenter was around 250 Km. about half an hour before As shown in Figure B.4 [Ref. 2], reception system recorded an KHz data the 81 the shock, 76 II : IKHl M^» SUGAOAIR* >••?. >1 1*- ==5= 1'««> 17 00 5f»TMs«tO<a Electromagnetic Noise Levels at Sugadaira Figure B.4 anomalous increase in the intensity. the normal level approximately 8 by more than uV. 15 It can also The amplitude exceeded dB. The normal level was be seen in Figure B.4 that the intensity dropped sharply back, to exactly at the moment of the shock. the previous level Another event occurred at 1247 JST on January 28, 1981, as shown in Figure B.5 [Hef, 2]. The location of the epicenter was in the western Ibaragi side of Tokyo^ prefecture on and the distance between 77 the northeastern the epicenter and II.IKHt li:}« Figure B.5 g.ir..>i4MI lANUARY. M. It Jl i:.ii I l»T J»flO» 10« M» JH IJM 'I *> "r H»1 JJT SO Electromagnetic Noise Levels at Suginaai the observation point was approximately 50 Km. focus was 60 Km and the magnitude was 5.0. at 81 KHz increased to about The depth of The noise lovel 12 dB above the threshold level about 45 minutes Such a feature is before the main shock. quite similar to the typical events presented in Figure B.U, 78 LIST OF EEPEEENCES 1. Friedlander, G.D., "Earthquake Prediction: Sp ect rum, IEEE, pp. 46-57, September 1973. 2. Morgounov, V.A., Yoshino, T., and Measurement "Experimental of Possibly to Emissions Related Electromagnetic Japan," Journal of Geophysical Earthquakes in 78 27=7828'; TU NoB^, "Pages Research, Vol. 87, SepTemSer 19 82. Gokhberg, Tomizawa, M. B. A New Art," , I., 3- Review of "Earthquake Lights: A Derr, J. S. , Theories," Observations and Present Bull . Vol. Seism olo qical Society of America, 63, pp. 2177=2TH77"T^7 31 4- "Radio Stoker, C. , and Meyer, T., Warwick, J., With Rock Fracture: Possible Emission Associated Application to the Great Chilean Earthquake of May 22, 1950," Journal of Geophysical Research, Vol. No. 87, E4, AprII"T?F7. 5. Boore, "The Kotion of the Ground D.M., Earthquakes," Sci entific American, Vol. No. 237, pp. 68-7 8, in 6, 197T: 6. ed., Proceedinqs of Dickinson, W.R., and Grantz. A., Conference on Geologic Problems of San ^nHreas Faul^ System, ^fanZord Univ. PuEIicafion, Vol. TT, T^65. 7. "Shock-Induced Polarization in Plastics. II. Experimental Study of Plexiglas and Polystyrene," J our nal of Applied Physics, Vol. 36, No. 7, July 1965. 8. " Schnurmann, liarlow-Davies, The and E. , E., Electrostatic Component of the Sliding Force of Friction," Proceedings, Physical Society of London, Vol. 54, Sep Vernier "Tg^l. 9. Nitsan, "Electromagnetic Emission Accompanying U., Fracture of Quartz-Bearing Rocks," Ge oph y sica l Research Letters, Vol. 4, No. 8, August 19 77. 10. Mizutani, H. , Ishido^ T. , Yokokura, T., and Ohnishi, S,, "Electrokinetic Phenomena Associated With Earthquakes," G eop hysical Research Letters, Vol. 3, No. 7, July 197^7^ 11. Gokhberg, M.B., Gufeld, I.L., and Dobrovolsky, I. P. "Sources of Electromagnetic Earthquake Precursors," Academy of Scien ces of the USSR, Schmidt Institute of PHysics ol fEe~IarTh7~VoTT 75D"7 No. 2, 19 80. Hauver, G.E., 79 12. Bleaney, B.I. ^ and Bleaney, B., Electricity Magnetism, p, bl5, Oxford, London, 19 57. 13. M. , Smith, B. E., and Mueller, Johnston, R. "Tectono-magnetic Experiaents and Observations in Western U.S.A.," Geomaq. Geoelec. pp. 28, 85-97 1976. 14. C.W., and Frank, Electronic Drickamer, H-E. , Pressure Chemistry an3 Transitions and the High VEys ics ol SolIHs, CEapman and Hall, lond'on, 1'973. 15. Burke, J. J., Shock Haves and the Mexhanical Prope rtie s of Solids, Syracuse Univ. Press, T5TT. 16. Allen, F.E., "Shock Induced Polarization in Plastics," Journal of Applied Physics, Vol. 36, No. 2111, 7, p. T^S57 17. and Carver, Kraus, J.D., McGraw Hill, p. 402, 1973. 18. Los Alamos National Laboratory Technical Report 83-3120, Obser vat ions of an Ionospheric Perturbation Arising from"THe CoaTInqa EartEgualce of"^ ^ay T?g3, by J. FT TIolcoTr, and U.J. Slnions7 t^E3. 80 K.R., and Electromagnetics, INITIAL DISTEIEUTION LIST No. 1. Copies Superintendent Library. Code 0142 Naval Postgraduate School Monterey, California 93943 2 Richard W. Adler Code 62 Naval Postgraduate School Monterey, California 93943 6 Attn: 2- Dr. 3. Dr. Stephen Jauregui, Jr. Code 62JA Naval Postgraduate School Monterey, California 93943 4. Defense Technical Information Center Cameron Station Alexandria, Virginia 22314 2 5. LT Mickey V. Ross Ross circle North Little Rock, Arkansas 72114 2 CAPT Carl Mortensen 2 1 12 6. MS 977 U. S. Geological Survey 345 Middlefield Menlo Park, California 94025 81 /ff'V/y —to xiwiae meaM surement and recording J7ste«. to predict the intensity and location Of electromagnetic Thesis RTT568 c.l Ross Design of a freestanding noise measurement and recording system to predict the intensity and location of electromagnetic radiation from earthquakes .