PROCEEDINGS, Twenty-Ninth Workshop on Geothermal Reservoir EngineeringStanford University, Stanford, California, January 26-28, 2004SGP-TR-175

MAGNETOTELLURIC SURVEYING AND MONITORING AT THE COSO GEOTHERMALAREA, CALIFORNIA, IN SUPPORT OF THE ENHANCED GEOTHERMAL SYSTEMS

CONCEPT: SURVEY PARAMETERS AND INITIAL RESULTS

Philip E. Wannamaker (1), Peter E. Rose (1), William M. Doerner (2),Brian C. Berard (3), Jess McCulloch (3), and Kenneth Nurse (2)

(1) University of Utah/Energy & Geoscience Institute423 Wakara Way, Suite 300Salt Lake City, UT 84108

e-mail: pewanna@egi.utah.edu(2) Quantec Geoscience Inc.(3) Coso Operating Company

ABSTRACT

Electrical resistivity may contribute to progress inenhanced geothermal systems (EGS) by imaging thegeometry, bounds and controlling structures inexisting production, and by monitoring changes inthe underground resistivity properties in the vicinityof injection due to fracture porosity enhancement. Tothese ends, we are acquiring a dense grid ofmagnetotelluric (MT) stations plus contiguous bipolearray profiling centered over the east flank of theCoso geothermal system. Acquiring good quality MTdata in producing geothermal systems is a challengedue to production related electromagnetic (EM) noiseand, in the case of Coso, due to proximity of aregional DC intertie power transmission line. Toachieve good results, a remote reference completelyoutside the influence of the dominant source of EMnoise must be established. Experimental results so farindicate that emplacing a reference a distance of 65miles from the DC intertie in Amargosa Valley, NV,is still insufficient for noise cancellation much of thetime. Even though the DC line EM fields are planarat this distance, they remain coherent with the non-planar fields in the Coso area so that remotereferencing produces incorrect responses. We havesuccessfully unwrapped and applied MT times seriesfrom the permanent observatory at Parkfield, CA, andthese appear adequate to suppress the interference ofthe artificial EM noise. The efficacy of thisobservatory is confirmed by comparison to stationstaken using an ultra-distant reference east of Socorro,NM. Operation of the latter reference was successfulby using fast ftp internet communication betweenCoso Junction and the New Mexico Institute ofMining and Technology, using the University of Utahsite as intermediary, and allowed referencing within afew hours of data downloading at Coso.

INTRODUCTION

A candidate for an Enhanced Geothermal System(EGS) possesses higher than average heat flow buthas natural permeability and/or fluid content which islimited (Robertson-Tait and Lovekin, 2000). At theeastern margin of the Coso geothermal system,measured formation temperatures exceed 300oC atdepths less than 9,000 ft, but the natural permeabilityis disappointingly low. Preliminary studies indicatethat appropriate tectonic conditions exist to createshear failure following hydraulic stimulation at theeast flank of Coso, such that reservoir stimulationtechniques may create sustained permeable fractures.In addition, the main Coso field is one of the largestand most productive liquid-dominated geothermalsystems in the U.S., and concepts on reservoircontrols are still evolving (Adams et al., 2000;Kurilovitch et al., 2003).

A successful approach to understanding anddeveloping an EGS must include technology toprovide images of subsurface structures whichcontrol geothermal fluid flow. Moreover, an EGSwill be promoted if images can obtained of thechanges in subsurface properties which may occurfollowing hydraulic stimulation. Electrical resistivityis a primary physical property of the Earth which canbe strongly affected by geothermal processes. Sincean increased fluid content due to fracturing, and thedevelopment of more conductive alteration minerals(clays, etc.), can give rise to an electrical resistivitycontrast, electromagnetic (EM) methods of probinghave been investigated and applied for many years.Consequently, we have been acquiring a reasonablydense magnetotelluric (MT) data set over the easternCoso field (Figure 1).

Figure 1. Simplified geological map of the Cosogeothermal area including interpreted reservoircompartmentalization (after Adams et al., 2000), andMT data sites. Geology map somewhat modified fromAdams et al. based on Whitmarsh (2002). Principalalteration areas Devil’s Kitchen (DK), Coso HotSprings (CHS), and Wheeler Prospect (WP) shown.

The MT technique makes use of naturally occurringEM wave fields produced by regional/globallightning activity and by solar wind/magnetosphericinteractions as sources for probing undergroundelectrical resistivity structure (Vozoff, 1991). Thesesignals are small in amplitude (e.g., 10’s of micro-V)and thus are easily contaminated by unrelatedenvironmental noises, principally man-made ones.Given that the Coso field currently produces inexcess of 250 MW or power, artificial noises can beexpected to be large. In addition, Bonneville PowerAuthority operates a DC intertie power transmissionline extending some 800 miles from The Dalles, OR,to the north Los Angeles area, with peak power loadsup to 3 GW (Keeping Current, BPA monthly, Oct.2000). In principle, as described more below, highquality MT data can be recovered in suchenvironments if a remotely located MT site whichrecords simultaneously with the main survey isestablished. The broad reach of the DC intertie fieldshas been perhaps the principle challenge to achievinggood MT results and it is the main purpose of thispaper to describe our experience in establishingeffective remote referencing for the Coso field.

COSO FIELD GEOLOGICAL SETTING

The Coso Geothermal area (Figure 1) is located in theCoso Range at the margin between the eastern flankof the Sierra Nevada and the western edge of the

Basin and Range tectonic province of southeasternCalifornia, and lies within the Walker Lane/EasternCalifornia Shear Zone (WLSZ). The Basin andRange province, an area of high heat flow andseismicity, is characterized by northerly trendingfault block mountains separated by alluvial valleysthat result from extensional tectonism. The WalkerLane/ Eastern California Shear Zone is a tectonicallyactive feature and is characterized in this region asaccommodating approximately 11 mm/yr of north-south trending, right-lateral strike-slip motionbetween “stable North America” and the SierraNevada (McClusky et al., 2000). To the west, theCoso Range is separated from the Sierra Nevada byRose Valley, the southern extension of the OwensValley. It is bounded to the North by Owens Lake, alarge saline playa. On the east, the range is boundedby Darwin (Coso) Wash and the Argus Range, andon the south by Indian Wells Valley.

The Coso Range basement is dominated by fracturedMesozoic plutonic with minor metamorphic rocks,that have been intruded and partly covered by lateCenozoic volcanic rocks. The basement complex hasbeen intruded by a large number of northwesttrending, fine-grained dikes. These dikes range incomposition from felsic to mafic and are believed tobe part of the Independence Dike swarm with asuggested Cretaceous age. The late Cenozoicvolcanic rocks consist of basalts and rhyolites. In themost recent volcanic phase, thirty-nine rhyolitedomes were emplaced in the past million years in thecentral region of the field along with a relativelysmall amount of basalt on the margins. Over the past0.6 My, the depth from which the rhyolites eruptedhas decreased, ranging from ~10 km depth for the~0.6 Ma magma, to ~5.5 km for the youngest (~0.04Ma) magma. These results are consistent with either asingle rhyolitic reservoir moving upward through thecrust, or a series of successively shallower reservoirs,in keeping also with recent Ar-Ar geochronology(Kurilovitch et al., 2003). As the reservoir hasbecome closer to the surface, eruptions have becomeboth more frequent and more voluminous (Manleyand Bacon, 2000). This partially molten magmachamber is believed to be the heat source that drivesthe geothermal system.

Stresses that control the faulting and fracturepermeability of the reservoir rocks are believed to bethe result of the location of the Coso Range in thetransitional zone between Basin and Rangeextensional tectonics to the east and strike-sliptectonics to the west (Roquemore, 1980; McClusky etal, 2000). Two major fault orientations havetraditionally been recognized to control thegeothermal system. The first set of faults strikeWNW, have a vertical dip, and have strike-slipearthquake solutions, while the other stronglydeveloped system of faults strike NNE and dip to the

east. The NNE striking fault zones have beensuccessfully targeted in development of the Cosogeothermal field, in particular in the east flank areawhere wells drilled with a steep westerly dip havebeen the most productive (e.g., Sheridan et al., 2003).Permeability is high within the individual Cosoreservoirs (i.e., Figure 1) but low in most of thesurrounding rock. This limits recharge to thereservoir and makes reinjection important forsustained productivity. A dominantly ESE-WNW tonearly E-W extensional direction is confirmed byrecent focal mechanism studies (Unruh et al., 2002).However, this extension is interpreted in the contextof distributed dextral shear involving important faultzones of truly SE-NW orientation (op. cit.), somesubsidiaries of which are drawn in Figure 1.

MAGNETOTELLURIC DATA ACQUISITION

A very simplified cartoon of an MT site deploymentas used at Coso is shown in Figure 2. The electric andmagnetic field components of the EM waves aremeasured with two types of sensors. The electric fieldprinciple is simple, with voltage differences takenover a bipole span of nominally 100 m, and thevoltages divided by distance to get E-field.Contacting endpoints of the bipoles were cold-rolledsteel plates in holes ~20 cm deep with ~1 liter wateradded to improve contact. The magnetic fields areobtained using high-sensitivity solenoids (coils) withpreamplifiers built in. These are buried a similardepth for thermal and mechanical stability. Due to theremote nature of the sources and the high index ofrefraction of the earth relative to the air, the sourcefields are assumed to be planar and to propagatevertically downward. However, two axes of E andthree axes of H are measured because the scatteringof EM waves by subsurface structure can be arbitraryin polarization, necessitating a tensor description.

Figure 2. Simplified view of an MT station asdeployed at the Coso geothermal field.

Broadband EM times series are recorded by thesedevices, and they are decomposed into individualfrequency spectra through fourier transformation.Through band averaging and ratioing, we arrive at thefundamental MT quantity which is interpreted forgeological structure, namely the tensor impedance ofthe earth to vertically incident, planar EM wavepropagation. This is expressed as:

E = [Z]H ,

where Z is a two-rank tensor. Individual elements ofthe impedance are subject to simple arithmetic toobtain an apparent resistivity (Rho) and impedancephase, which are more intuitive to inspect andinterpret (Vozoff, 1991). The nominal frequencyrange recorded was 250 Hz to 0.01 Hz, which spans adepth range of several 10’s of m to >10 km.

The assumption of a planar geometry to the MTfields is crucial and artificial EM sources nearby caninvalidate it. An obvious source could be the high-voltage 60 Hz production of the field, which may beso strong under transmission lines or next togeneration plants as to saturate MT recordingelectronics. Strictly speaking, however, the loss of avery narrow band of results around 60 Hz is generallynot serious since the impedances are a smoothlyvarying function of frequency. More problematic arebroadband noises whose causes are often obscure butcan include load fluctuations (either within the Cosofield or from the DC intertie), rotating machinery,and vibration. These are especially onerous in the 3-0.1 Hz frequency band where the MT fields areparticularly weak.

An example of MT time series taken at the Coso area~0.5 km west of the Navy II power plant is presentedin Figure 3. These are compared to series acquiredsimultaneously just east of Socorro, NM, some 600miles to the east of Coso. Most obvious in the firstportion of the plot at the Coso site is sinusoidalvariation with a frequency of ~6 Hz which is absentat Socorro. The cause of this fluctuation is unknown.In the quieter portion of the plot, higher energy burstsof MT signal are seen at both sites (e.g., time 4:06:14and 4:06:22; note that amplitude range of plotsdiffers). However, the signal burst at 4:06:05 atSocorro is swamped by the noise source at Coso atthe same time. During weaker signal times even withnoise sources less obvious, the visible correlationsbetween the two sites is often obscure, evidence ofbroad band noise of a lower level but still competitivewith the signal. Note that the vertical magnetic fieldis completely dependent on lateral variations inresistivity structure, so that visible correlations evenwithout noise are naturally more obscure. Numerousother noise examples could be plotted, such as verylarge spikes with accompanying decaying transients.

Figure 3. A 30 s segment of MT time series taken atsite E34 in the Coso field (top five traces) comparedto series taken near Socorro, NM. Ordinates are full-scale plots, so relative amplitude factors may differ.

“Local” Remote Reference Processing AttemptsThe remote reference method (reviewed by Vozoff,1991) is designed to overcome environmental noisethrough coherent detection utilizing simultaneous,remotely recording sensors completely outside theinfluence of the noise sources. The reference site ofcourse may have its own noises but as long as theseare uncorrelated with the local site then the detectionprinciple remains valid. Of course, the remote siteshould be of high fidelity or the final result will bedegraded or a greater length of averaging time will beneeded to achieve equivalent results (e.g., Egbert,1997). In the Coso data campaign so far, referencesites were attempted at five locations of varyingdistance from the geothermal field (Figure 4). Thenearer three sites which measured fields using thesame equipment of the MT contractor (QuantecGeoscience Inc.) include the Centennial Flat areaapproximately 15 miles north of the Coso field,Panamint Valley about 30 miles distant, andAmargosa Valley NV about 60 miles distant. Timeseries from these sites were downloaded to field PCby the local Coso collection crew for processing atessentially the same time as those in the Coso field,which is normal survey procedure. For the moredistant references we discuss later, more elaboratecommunications procedures were required. Anincreasingly great distance for the reference wasrecognized as the survey proceeded.

Figure 4. Location of “remote” MT references in thevicinity of the Coso geothermal field. These areCentennial Flat (CEN), Panamint Valley (PAN),Amargosa Valley (AMG) and Parkfield (PKD). Notshown is Socorro, NM, some 600 miles to the east.Also shown is trace of BPA DC intertie line (red)which passes within a few miles of the Coso field.Urban centers include San Francisco (SF), LosAngeles (LA), Las Vegas (LV), and Carson City (CC).DC intertie trace provided by Jim Lovekin.

The MT apparent resistivities corresponding toimpedance elements Zxy and Zyx for the reference

site at Centennial Flat and in Panamint Valley areshown in Figure 5. Characteristic of non-MTartificial effects are apparent resistivities which risewith decreasing frequency at an anomalously steeprate in the weak MT middle band, and then fallalmost discontinuously around 0.1 Hz where the MTfields become quite strong again due to solar windenergy. This is apparent at the Centennial Flat site, asis the case for soundings generally in the Coso areaprocessed using these nearer references. The effect isanalogous to that seen in controlled-source (CSAMT)surveys when the transmitter is to close to the surveyarea (Zonge and Hughes, 1991; Wannamaker, 1997).The distortion propagates to higher frequencies to anunknown degree so that structural images in thepertinent depth range of several km may beunreliable. Essentially only the yx component iseffected, which corresponds to an E-field directed N-S. This behavior implicates the Bonneville DCintertie, which runs in this direction, and whichgenerates E-fields parallel to itself.

Figure 5. Apparent resistivity soundings taken atCentennial Flat about 15 miles north of Coso field,and in eastern Panamint Valley, about 30 milesnortheast of the field. Note the abrupt change inRhoYX around 0.1 Hz, indicative of a near-field (NFor non-MT) effect, in the former (closer) site, but itsabsence in the further site.

In an attempt to escape the influence of the DCintertie, references were tried in Panamint Valley and

Amargosa Valley, about 30 and 65 miles distant fromthe intertie respectively. The sounding from Panamintis shown in Figure 5 also. This good-quality soundingdoes not show the cusp-like behavior near 0.1 Hzseen in Centennial Flat, and the site at AmargosaValley was even more clean. Initially, we concludedfrom this that we were essentially free from influenceof the DC intertie at these distances and that theywould constitute sufficient references. However,soundings in the Coso area processed using thesereferences still showed strong cusp-like behavior inthis frequency range (Figure 7, site taken ~1/2 kmeast of well 34-9), although the results using the moredistant Amargosa site were better. This wasdisappointing as 65 miles of often winding road wasreaching the limit in terms of practical, on-sitereference retrieval.

Figure 6. Apparent resistivity data from site E13 inthe Coso geothermal field processed using sensors inPanamint Valley (upper) and in Amargosa Valley(lower) as remote references. Some near-field effectappears to remain near 0.1 Hz.

It is evident that a site giving good quality plane-wave (MT) results does not necessarily serve as agood remote reference for soundings taken in a noisyenvironment. Clearly, EM fields which are correlatedwith the DC intertie are persisting at least as far awayas the Amargosa Valley reference. This is occurringeven though such fields have become planar by thispoint and only serve to improve the local MTresponses at Panamint and Amargosa Valleys (e.g.,

as in Figure 5). However, the powerline fields in theCoso field area, which are quite non-planar, remaincorrelated with the Panamint and Amargosa sites andthus are not removed by remote reference processing.A reference must be established which is completelyoutside the domain of the noise source in a survey,not just put where the noise has become planar.

Distant Remote References: Parkfield CaliforniaObservatory and Socorro New MexicoBy this point (May/03), >50 MT soundings had beentaken at Coso, most of which showed such non-MTcontamination to some degree. To obtain high qualityresults at these locations, we were faced with acomplete reoccupation using a yet more distantreference (expensive) or else locate a reference site ofopportunity which fortuitously happened to berunning while the Coso survey was underway.Luckily, such a reference site exists in the way of theParkfield, CA, permanent MT observatory, run by theUniversity of California at Berkeley (for info, seehttp://quake.geo.berkeley.edu/bdsn/em.overview.htmlof the Berkeley seismic network) (PKD in Figure 4).Due to powerlaw falloff and dissipation of EM fieldsfrom a line-source, the DC intertie fields at PKDshould be weaker than those at Amargosa by about afactor of 5, making the attempt to use PKDworthwhile. The EM times series are available at nocharge essentially in real time through an ftp request.They are sampled at a rate with allows their use as areference for frequencies up to 20 Hz, which coversthe contaminated band at Coso. The contractorQuantec successfully unwrapped the native SEEDformat of the time series and automated their use asreferences in their processing software. A plot of theresults for site E13 appears in Figure 7 and indicatesthat the near-field effect has been corrected. Theprincipal frequency range of distinction between theParkfield-processed and the previous results is 1-0.1Hz. Similarly good results were obtained forreprocessing the other ~50 sites of the Coso area.

This outcome is very fortunate, but there is concernthat Parkfield is barely adequate and that duringtimes of exceptionally low natural MT field activitythere may be some noise remaining. Hence, for theresumption of MT surveying at Coso which tookplace Nov.-Dec./03, a remote reference was set upnear Socorro, NM (SOC), >600 miles to the east ofthe Coso field. Because it is important to recognizesurveying problems as they may occur, it is importantto have the Coso and the reference time seriesbrought together as quickly as possible. To achievethis, crew at Coso uploaded acquired time series tothe U. Utah/EGI ftp site using the Caithness Energyhigh-speed computer facilities at Coso Junction. Thedata processor/reference operator in Socorro down-loaded these series from Utah using the high-speed

computer facilities of the New Mexico Institute ofMining and Technology. Site, and reference timeseries thereby were combined several hours aftertheir acquisition at Coso.

Figure 7. Apparent resistivity data from site E13 inthe Coso geothermal field processed using theParkfield MT observatory as a remote reference.

To verify that the MT time series at Socorro are wellcorrelated with those at Coso, and Parkfield for thatmatter, we show in Figure 8 a 2 min segment takensimultaneously from site E29, PKD and SOC. This isan unusually quiet segment of time at E29 andexcellent correlations are seen both between theimpulsive, high-frequency data due to regional-scalelightning energy, and the low-frequency magneto-spheric signals, at all three sites. Site E13 was notreoccupied during the second deployment using theSocorro reference, but we can compare results fromsite E29 processed with PKD with those processedusing SOC (Figure 9). The sounding curves areessentially identical indicating that Parkfield was anadequate reference in this case, although we viewSocorro with more assurance as a quiet site outsidethe influence of the DC intertie.

The effort to which we have gone to establish a quietremote reference free from both local and broadscaleartificial EM interference has been completelynecessary. The zones of producing fractures at Cosoreside in the 1-3 km depth range (Adams et al., 2000)in plutonic rocks mantled by a variable layer ofaltered overburden. This situation determines the firstorder character of the soundings shown thus far,namely apparent resistivities in the 10-30 ohm-mrange for frequencies > 3-10 Hz rising to values near100 ohm-m at lower frequencies. It will be subtlevariations in the upward slope of the apparentresistivity, plus corresponding impedance phaseresponses, which will provide the second-orderevidence for bedrock structure of potentialgeothermal significance.

Figure 8. Comparison of all five channels of MT timeseries at site E29 with the two horizontal magneticfields used as references at PKD and at SOC.

Figure 9. Sounding E29 processed using theParkfield observatory as a reference compared toE29 processed using the Socorro site as a reference.

Conclusions and Plans

High quality MT data have been acquired in the Cosogeothermal field following adequate efforts toestablish a clean remote reference. It is not sufficientto use a reference site which appears to give goodplane-wave results locally. This is because there maybe EM noise fields (albeit planar) at the reference sitethat are correlated with those of the survey area,which are non-planar. In this survey, we found thatMT time series at the permanent Parkfieldobservatory in California were adequate as referencefields to the sites at Coso. In the course of doing this,technology has been established by the contractor toutilize the Parkfield fields in MT surveying generally,thereby improving survey quality for otherapplications. Noise sources of scales as broad as thatof the BPA DC intertie are rare, but this particularone may be a factor in exploration of numerous othergeothermal systems in western Nevada andeasternmost California due to its proximity.

At this writing, all but about 10 MT sites in thenorthernmost portion of the Coso area have beenacquired, and this should be completed in the latewinter of 2004. The need for high quality results isespecially important in this field because thebasement structures being sought will have a second-order influence on the data, which are dominated bythe response of variable conductive overburden uponbedrock. Following basic quality control assessmentof the site ensemble, principle structural trends in thebasement will be estimated using modern galvanicdecomposition techniques (e.g., Sodergren, 2002).This should resolve whether primarily north-south

average fault trends or primarily northwest-southeastfault trends dominate the conductivity (andpresumably permeability) grain of the basement.Ultimately, 3-D MT inversion will be applied to thedata, which we are developing building on the workof Sasaki (1999, 2001) and Wannamaker (1991).

ACKNOWLEDGEMENTS

This work was supported under U.S. Dept. of Energycontract DE-PS07-00ID13913 and Dept. of the Navycontract N68936-03-P-0303. Energy & GeoscienceInstitute researchers thank Coso Operating Companyfor access to the field and to the high-speedcomputing facilities which made the ultra-distantremote referencing possible. Similarly, Harold Tobinand Rick Aster of New Mexico Institute of Miningand Technology are thanked for access to land andhigh-speed ftp for the Socorro remote reference. Wealso thank Frank Monastero and Allan Katzenstein ofthe U.S. Navy Geothermal program office for supportand encouragement, and for funding all archeologicalsite clearances. Finally, the competence and diligenceof the field crew of Quantec Geoscience, principallyJon Powell, Joel Cross and Claudia Moraga, maderesults of this quality possible. Valuable discussionson MT data processing and use of the Parkfieldfacility were held with Gary Egbert.

REFERENCES

Adams, M. C., Moore, J. N., Bjornstad, S., Norman,D. I., 2000, Geologic History of the Coso GeothermalSystem, Geothermal Resources CouncilTransactions, 24, 205-209.

Egbert, G. D., 1997, Robust multiple stationmagnetotelluric data processing, Geophysical JournalInternational, 130, 475-496.

Kurilovitch, L., Norman, D., Heizler, M., Moore, J.,McCulloch, J., 2003, 40Ar/39Ar Thermal History ofthe Coso Geothermal Field, Proc. 28th Workshop onGeothermal Reservoir Engineering, StanfordUniversity, Stanford, CA, Jan. 27-29, SGP-TR-173.

Manley, C. R., and Bacon, C. R., 2000, Rhyolitethermobarometry and the shallowing of the magmareservoir, Coso Volcanic Field, California, Journal ofPetrology, 41, 149-174.

Robertson-Tait, A., and Lovekin, J., 2000, Potentialsites and experiments for Enhanced GeothermalSystems in the western United States: GeothermalResources Council Transactions, 24, 169-174.

Roquemore, G., 1980, Structure, tectonics, and stressfield of the Coso Range, Inyo County, California,Journal of Geophysical Research, 85, 2434-2440.

Sasaki, Y., 1999, 3-D inversion of electrical andelectromagnetic data on PC's, in Wannamaker, P. E.,and M. S. Zhdanov, eds., Three-dimensionalelectromagnetics, Proc. 2nd Internat. Gerald W.Hohmann Symp., University of Utah, Salt Lake City,128-131.

Sasaki, Y., 2001, Full 3-D inversion of electromag-netic data on PC, J. Appl. Geophys., 46, 45-54.

Sheridan, J., Kovac, K., Rose, P., Barton, C.,McCulloch, J., Berard, B., Moore, J., Petty, S., andSpielman, P., 2003, In situ stress, fracture and fluidflow analysis – east flank of the Coso geothermalsystem, Proceedings 28th Workshop on GeothermalReservoir Engineering, Stanford University,Stanford, CA, Jan. 27-29, SGP-TR-173.

Sodergren, T. L., 2002, Anisotropy and thermal stateof the lower crust and upper mantle of the centralGreat Basin province, Nevada, U.S.A., frommagnetotelluric data, M.S. thesis, University of Utah.

Unruh, J. R., E. Hauksson, E., Monastero, F. C.,Twiss, R. J., and Lewis, J. C., 2002, Seismotectonicsof the Coso Range-Indian Wells Valley region,California: transtensional deformation along thesoutheastern margin of the Sierran microplate, inGeologic evolution of the Mojave Desert andSouthwestern Basin and Range, ed. by Glazner, A.F., Walker, J. D., and Bartley, J. M., Geol. Soc.Amer. Mem. 195, 277-294.

Vozoff, K., 1991, The magnetotelluric method, inElectromagnetic methods in applied geophysics, ed.by Nabighian, M. N., v. 2B, Soc. Explor. Geophys.,Tulsa, p. 641-711.

Wannamaker, P. E., 1991, Advances in three-dimensional magnetotelluric modeling using integralequations: Geophysics, 1716-1728.

Wannamaker, P. E., 1997, Tensor CSAMT survey ofthe Sulphur Springs thermal area, Valles Caldera,New Mexico, Parts I&II: implications for structure ofthe western caldera and for CSAMT methodology:Geophysics, 62, 451-476.

Whitmarsh, R. S., 2002, Geological map of theCactus Peak 7.5’ quadrangle; Inyo County,California, CD-ROM map, in Geologic evolution ofthe Mojave Desert and Southwestern Basin andRange, ed. by Glazner, A. F., Walker, J. D., andBartley, J. M., Geol. Soc. Amer. Mem. 195.

Zonge, K. L., and Hughes, L. J., 1991, Controlledsource audiomagnetotellurics, in Electromagneticmethods in applied geophysics, ed. by M. N.Nabighian, 2B, Soc. Expl. Geophys., Tulsa, 713-809.