Application of Remote Sensing Technology to Geothermal Exploration Wendy M. Calvin, Mark Coolbaugh, Chris Kratt, and R.

Greg Vaughan", Great Basin Center for Geothermal Energy (GBCGE)

and the Department of Geological Sciences

University of Nevada -- Reno

1 Now at Jet Propulsion Laboratory 4800 Oak Grove Drive

Pasadena, CA 91109

Correspondence Author:

Wendy Calvin

Geological Sciences, MS 172 UNR, Reno, NV 89557

wcalvin@unr.edu 775-784-1785

Running Title: Remote Sensing for Geothermal Exploration

Abstract

Recent significant improvements in the wavelength coverage, spectral resolution and quality of remote sensing imagery have lead to the extensive application of these data sets in exploration and site characterization. Traditional techniques in spectral and spatial analysis of imagery, coupled with new, high signal to noise data, allow their direct application to problems in geothermal energy exploration and development. Our work seeks to define surface indicators of geothermal resources at known source regions and to apply these tools to recognizing geothermal potential in other areas. Geothermal indicators include sinter, tufa (carbonate), hydrothermal alteration (clays, sulfates) and thermal anomalies. Our work responds to the need to identify new resources and to map the geology of these areas. In addition, we hope to bring down costs by helping to focus new development in existing resource areas. Our work demonstrates that airborne and spaceborne imagery can be used to map mineralogy and hot spots over broad and inaccessible regions. We have clearly associated indicator mineralogy with

present and past geothermal activity. Remote sensing data have been able to identify fault extensions not previously mapped by field geologists. Future work will test these methods on new sites not yet producing power to confirm their predictive capabilities for new resources.

Key Words: Remote Sensing, Mineral Mapping, Thermal Anomalies, Steamboat Springs, Bradys Hot Springs

Introduction

Renewable energy exploration and development is currently enjoying new interest as the nation seeks to lessen its dependence on foreign oil sources. Geothermal energy is competitive in areas where high heat flow and crustal water circulation combine to provide a readily accessible resource. Nevada in particular has abundant potential for energy development. The effort at present within the Department of Energy is focused on reducing the cost per kilowatt hour (Kwh) by enhancing energy production at existing sites and bringing new resources on line.

Our work seeks to define surface indicators of geothermal resources through analysis of characteristic mineral, vegetation, and thermal properties at known source regions and to establish markers of potential in other areas. Geothermal indicators amenable to remote sensing include mineral spectral signatures as well as thermal anomalies mapped in both nighttime and corrected day/night imagery. For our analysis we use a variety of data sets from both airborne and spaceborne platforms.

Our research has concentrated on two established geothermal sites - Steamboat Springs and Bradys Hot Springs in Nevada. We have also examined data of the Dixie Valley and Buffalo Valley geothermal areas. We have used two primary methods of analysis: 1) combined day and nighttime imagery for thermal anomalies and 2) spectral analysis to determine diagnostic surface mineral absorption features to create mineral maps. These two methods have been merged using GIS techniques at Steamboat and Bradys.

Instruments and Methodology

We use a combination of optical, near-infrared and thermal infrared imagery that provides a multifaceted approach to identify geothermal resources and map geologic structures based on their surface expression. The data are both multispectral (several wavelength channels) and hyperspectral (hundreds of wavelengths). These data sets are sensitive to surface mineralogy and lithology as well as surface temperature. Derived image products include spatial maps of minerals and surface alteration, locations of thermal anomalies, and synthesis of remotely sensed thematic maps with field-based geologic maps. The data are geo-rectified to standard projections so that they can be readily included in regional and site specific databases.

Airborne and spaceborne sensors used include ASTER, MASTER, AVIRIS, HyMap, and SEBASS. Wavelength coverage, number of channels, and ground spatial resolution for these instruments are summarized in Table 1. The first two are multispectral systems and the last three are hyperspectral. All except SEBASS measure in the reflected solar portion of the spectrum (0.4 to 2.5 urn, the combined optical and near-infrared) and ASTER, MASTER and SEBASS observe

Calvin et al.

Remote Sensing

Page 1

surfaces at thermally emitted infrared wavelengths (7 to 15 um), ASTER is the only spaceborne data set used. While we often acquire Landsat coverage over our target sites, we typically do not spend a great deal of time on image processing and analysis of that data because ASTER offers improved spatial resolution in the visible and more spectral channels in the near-infrared over Landsat.

(Table 1 near here.)

Identification of Thermal Anomalies

Surface temperature depends on numerous physical parameters. These include albedo (brightness), slope, compaction and coherence (soil vs rock) as well as recent insolation history. A light colored surface will reflect more sunlight and hence be cooler during the day (a darker one is warmer). A slope facing the sun receives more direct solar energy per unit area and hence is also warmer. Soil compaction and rock coherence are combined into a physical property known as thermal inertia or resistance to change in temperature. A fine grained surface (low inertia) will heat up and cool down quickly, while a competent rock will heat up but also cool down more slowly (high inertia) so that low inertia surfaces will be warmer during the day, but cooler at night. The surface temperature also depends on the amount of solar insolation reaching the surface (has it been cloudy?) and convective transport to the atmosphere (winds) can also affect surface temperature.

Geothermal anomalies are sometimes found in nighttime images (being warmer than the surroundings). However, we desired a more robust method that would account for the various aspects contributing to surface temperature and help identify anomalously warm regions with confidence. This method employs both day and night imagery so that the varying contributions cited above can be accounted for. Daytime imagery provides information on albedo and day/night comparisons can provide temperature differences and hence estimate thermal inertia. Digital elevation data (OEM) allows correction for slope and solar incidence angle. Full details of the parametric model that was developed are described in Coolbaugh et al. (2000) and Coolbaugh (2003). Results from both Steamboat Springs and Bradys Hot Springs are described later.

Mineral Mapping

Optical and infrared spectroscopy has been used for over 40 years to identify rocks and minerals. Beginning in the 1970's, imaging spectral instruments were deployed on aircraft and multi-spectral image platforms have been in orbit since that time. The technique relies on a combination of absorption features that arise from primary rock-forming elements (especially iron), and molecules and anions such as water, hydroxyl, carbonate and sulfate. In the thermal infrared the fundamental absorptions of the Si04 tetrahedra also cause absorption features that vary with the crystallographic co-ordination of the mineral (e.g., Farmer, 1974; Hunt, 1977). Remote spectral data sets have been successfully used to identify

Calvin et al.

Remote Sensing

Page 2

and map numerous mineral species of interest in geothermal exploration including sinters, carbonates, sulfates, and native sulfur (e.g. Hellman and Ramsey, 2004; Kruse, 1999; Martini et al. 2003; Nash et al. 2004).

Sinter, carbonate and hydrothermal alteration minerals, such as clays and sulfates, are identified by their unique spectral signatures from visible through infrared wavelengths, identified in Figure 1. These materials are spatially mapped in the airborne and spaceborne imagery and identifications are validated through field and laboratory spectral measurements as well as x-ray diffraction and thin section analyses.

(Figure 1 near here.)

Results

Steamboat Springs

In a pilot study for thermal anomaly identification, Coolbaugh et al. (2000) found that the main sinter terrace did not appear warm in uncorrected day/night thermal images. However, analysis that accounted for slope, albedo, and thermal inertia provided an accurate map of the main terrace as an anomalously warm region (Fig. 2). In this study the optical and infrared images were from different sensors and seasons, limiting the temperature difference that could be classified as anomalous. For our thermal anomaly analysis at Bradys we used ASTER data where optical and infrared data are acquired simultaneously.

(Figure 2 near here.)

We performed mineral mapping at Steamboat Springs using the thermal spectral channels of both MASTER and SEBASS. MASTER was able to separate clay dominant surfaces from quartz/sinter surfaces; however the data could not discriminate between quartz and sinter. Also, MASTER mineral maps did not show unique relationships to regions of known geothermal activity. The SEBASS instrument has much higher spectral resolution that allowed detailed separation of quartz from sinter deposits as well as distinguishing the evolution of sinter deposits with age. Vaughan et al. (2003,2004) and Vaughan (2004) summarize the results and note that alunite, kaolinite, and transient hydrous sulfates are uniquely found in regions of geologically recent venting or forming around active fumaroles.

These mineral identifications have been confirmed in the field and with XRD. Synthesis of SEBASS mineral maps with thermal anomaly maps shows a clear relationship between indicator minerals and geothermal activity.

Bradys Hot Springs

We reproduced the thermal anomaly analysis at Bradys Hot Springs using ASTER with two scenes, one acquired during the day, and one at night, within a 24-hour period. In addition surface temperature data was gathered in the field during the overflight along with diurnal temperature measurements at 15 to 30 minute intervals to establish thermal inertia curves of representative surfaces. Although

Calvin et al.

Remote Sensing

Page 3

the spatial footprint of ASTER is large, gOm/pixel, the data set easily identified the main structures and vents along the primary fault, extending over 4 km in length (Fig 3). Outlier patches of warm ground were confirmed in the field and there is a precise correspondence between the remotely derived hotspot map and fumaroles and hotspots identified and located using GPS in the field. Several anomaly outliers pointed to potential complexity in the fault system that was subsequently examined at higher spatial resolution with mineral mapping techniques.

(Figure 3 near here.)

We obtained optical and near infrared spectral coverage using the HyMap instrument (Fig. 4). Kratt et al. (2003, 2004) and Kratt (2005) summarize the results of mapping both sinter and carbonate tufa in the area. Sinter was mapped using a diagnostic absorption feature at 2.25 11m and identification was confirmed with laboratory analysis and XRD. Large areas of silicified diatomite were also mapped using this technique and significant field validation was performed in order to optimize detection thresholds and localize true sinter deposits. Figure 5 shows large deposits of sinter, mapped in red, that mark an extension of the fault line south of the main plant that was not previously recognized from field mapping. Structurally controlled tufa mounds trending from the main fault were also identified.

(Figure 4 and 5 near here.)

Conclusions

Our work demonstrates that spaceborne imagery can be used to map mineralogy and hot spots over broad and inaccessible regions. We have successfully mapped thermal anomalies at known sources - including uncovering "hidden" features at Steamboat Springs. Hyperspectral thermal infrared data can differentiate sinter deposits of opaline silica from massive chalcedony, thus separating recent from older deposits. More detailed mineral mapping shows clay alteration and sulfates are uniquely mapped in regions with current or geologically recent geothermal activity.

At the Bradys location, detailed thermal anomaly and mineral mapping found extensions of the main fault not previously mapped by field geology and suggest more complicated underlying structure at the site. Here the SWIR spectral region was able to locate sinter and structurally controlled sub-lacustrine tufa associated with current and historical geothermal activity.

Ongoing work is testing these methods at Pyramid Lake, in regions without any current power production, to analyze their effectiveness in a predictive mode for drilling and site placement of new power plants.

Calvin et al.

Remote Sensing

Page 4

Acknowledgements

This work is supported by DOE Grants DE-FG 07-021 D14311 and DE-FG36- 021 D14311 to the Great Basin Center for Geothermal Energy with partial contributions from the NASA Graduate Student Research Program (NG5-50362), UNR's Graduate Student Association and Nevada Space Grant.

References

Coolbaugh, M. F. (2003) The prediction and detection of geothermal systems at regional and local scales in Nevada using a geographic information system spatial statistics, and thermal infrared imagery. University of Nevada - Reno Ph. D. Thesis.

Coolbaugh, M. F., Taranik, J. V., and Kruse, F. A. (2000). Mapping of surface geothermal anomalies at Steamboat Springs, NV. using NASA Thermal Infrared Multispectral Scanner (TIMS) and Advanced Visible and Infrared Imaging Spectrometer (AVIRIS) data. In Proceedings, 14th Thematic Conference, Applied Geologic Remote Sensing, Environmental Research Institute of Michigan (ERIM), Ann Arbor, MI., p. 623-630.

Farmer, V.C. Ed. (1974). The Infrared Spectra of Minerals. Mineralogical Society Monograph 4, London.

Hunt, G. (1977). Spectral signatures of particulate minerals in the visible and nearinfrared. Geophysics, v. 42 (3), p. 501-513.

Hellman, M. J. and M.S. Ramsey (2004). Analysis of hot springs and associated depostis in Yellowstone National Park using ASTER and AVIRIS remote sensing. J. Volcan. Geotherm. Res., v. 135, p. 195-219.

Kratt, C. (2005) Geothermal Exploration With Remote Sensing from 0.45 - 2.5 IJm Over Brady-Desert Peak, Churchill County, Nevada. University of Nevada - Reno, M.S. Thesis.

Kratt, C., M. Coolbaugh, and W. Calvin (2003). Possible extension of Brady's fault identified using remote mapping techniques. Geothermal Resource Council Transacilons,v.27.

Kratt, C., W. Calvin, and M. Coolbaugh (2004). Geothermal exploration with hyperspectral data in the Carson Sink, Nevada. Proceedings Geoscience and Remote Sensing Symposium, 2004. IGARSS '04. IEEE International, v. 2,

p. 1304 - 1306.

Kruse, F. (1999). Mapping hot spring deposits with AVIRIS at Steamboat Springs, Nevada. AVIRIS Airborne Geoscience Workshop Proceedings 1999. JPL Publications, Pasadena, p. 1-7.

Calvin et al.

Remote Sensing

Page 5

Martini, B.A., E. A. Silver, W. L. Pickles, and P. A. Cocks, 2003. Hyperspectral mineral mapping in support of geothermal exploration: examples from Long Valley Caldera, CA and Dixie Valley, NV. Geothermal Resources Council Trans. v.27, p.657-662.

Nash, G.D., G. W. Johnson, and S. Johnson (2004). Hyperspectral detection of geothermal system-related soil mineralogy anomalies in Dixie Valley, Nevada: a tool for exploration. Geothermics, v. 33, 695-711.

Vaughan R.G. (2004). Surface Mineral Mapping at Virginia City and Steamboat Springs, Nevada with Multi-Wavelength Infrared Remote Sensing Image Data. University of Nevada Reno, Reno, NV. Ph. D. Dissertation.

Vaughan, R.G., Calvin, W.M., and Taranik, sv. (2003). SEBASS Hyperspectral Thermal Infrared Data: Surface Emissivity Measurement and Mineral Mapping. Remote Sensing of Environment, v. 85, 48-63.

Vaughan, R.G., Hook, S.J., Calvin, W.M., and Taranik, J.V. (2004). Surface Mineral Mapping at Steamboat Springs, Nevada with Multi-Wavelength Thermal Infrared Images, Remote Sensing of Environment, ASTER Special Issue, In Press.

Calvin et al.

Remote Sensing

Page 6

Tables:

T bl 1 S tid S t' I P

rt'

fR

t I

D t S t

a e ioec ra an ipa ra rope res 0 emoe mage aa es
Data Sets Used Wavelength Coverage Channels Spatial Footprint
ASTER: Advanced Spaceborne 0.5 to 0.9 urn 3 15 m
Thermal Emission and Reflection 1.6 to 2.45 urn 6 30 m
Radiometer s to tz um 5 90 m
MASTER: MODIS and ASTER 0.46 to 2.39 urn 25
Airborne Simulator 7.76 to 12.88 urn 10 5m
HyMap: 0.45 to 2.5 urn 126 3m
AVIRIS: Airborne Visible and 0.45 to 2.5 urn 224 3m
Infrared Imaging Spectrometer
SEBASS: Spatially Enhanced 7.5 to 13.5 urn 128 2m
Broad Array Spectrograph
System Figure Captions:

ow.

G'eo:.tr,er:m 0.11 n dotcDt~r °Mi'n Nels

o 1;..5 0 0 • 0 2'.b; .

WQveleng'th

Figure1: Visible and infrared spectral properties of common geothermal minerals (from top to bottom in the plot): carbonate (calcite), sulfate (alunite), clay (montmorillonite) and sinter (opal). Band center position, band shape, and width are diagnostic of specific minerals and can be mapped in hyperspectral data sets.

Calvin et al.

Remote Sensing

Page 7

Before Enhancement-

After Enhancement

Figure 2: Thermal anomaly map before and after correcting for surface albedo, slope and thermal inertia. Before enhancement (left) anomalies are dark grey and primarily identify the powerplant and sun facing slopes just above the North arrow in the image. After corrections (right), anomalies are identified by sharp gradients from black to white and identify numerous small features and a large anomaly that covers the main terrace as well as the commercial hot springs south of highway 395. The corrections allow for the identification of the main sinter terrace, previously hidden due to it's high albedo and high inertia.

Calvin et al.

Remote Sensing

Page 8

Figure 3: Map of thermal anomalies at the Brady's sight compared with warm ground and fumaroles identified and located in the field using GPS. Warm colors indicate warmer ground derived from the imagery at gOm/pixel. Small black and brown points indicate field located steam vents or surface hot spots. Fault offset at the southern end was first noted in the remotely derived anomaly map.

Calvin et al.

Remote Sensing

Page 9

Figure 4: Flightline coverage of HyMap data at Bradys shown over topography. Actual lines are marked by flags at their north and south ends. The right 5 lines were offset south in order to cover the enhanced geothermal plant at Desert Peak, a geothermal resource with no obvious surface expression.

Calvin et al.

Remote Sensing

Page 10

Figure 5. Sinter map at Bradys Hot Springs. Portions of 1-80 cross the image from left to upper right. The main plant is identified by small white rectangular features below the highway. Sinter outcrops are mapped in red. Small outcrops in the lowest portion of the image identified relic fault structures not previously recognized in field mapping.

Calvin et al.

Remote Sensing

Page 11