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

A geographic information system (GIS) and spatial statistical techniques were used to quantify relationships in Nevada between high-temperature (≥160°C) geothermal systems and young northeast-striking faults, young (≤1.5 Ma) volcanic rocks, high concentrations of boron and lithium in groundwater, and areas with shallow groundwater tables. Shearing caused by the Sierra Nevada plate as it moves past Nevada in a northwest direction (documented by global positioning system (GPS) measurements) is believed to preferentially open some northeast-striking structures, allowing meteoric waters to penetrate deeply into the crust and be heated to high temperatures. Areas with deep groundwater tables may conceal undiscovered geothermal resources.; Using the above-mentioned correlations, regional predictive maps for high-temperature geothermal systems were built using weights-of-evidence (WofE), logistic regression, and a “density function”, a hybrid model with features of both WofE and fuzzy logic. A density function is defined as the fraction of the total training sites occurring within a given histogram bin of an evidence layers data distribution, divided by the fraction of the total study area within that bin. The density function model has predictive capabilities similar to binary WofE, but with a finer resolution of favorability rankings that makes it easier to discern regional trends and relate them to known geological structures.; To facilitate more detailed searches for geothermal systems within favorable areas, refined methods of directly detecting surface geothermal temperature anomalies were developed using the airborne Thermal Infrared Multispectral Scanner (TIMS) and the satellite-based Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER). Surface temperature variations caused by diurnal solar radiation were minimized using day and night TIR images to reduce thermal inertia effects, visible and near-infrared bands to correct for albedo effects, and a digital elevation model to compensate for the effect of topographic slope orientation. At Steamboat Springs, Nevada the processing algorithms greatly expanded the size of a geothermal temperature anomaly on a sinter terrace that was not anomalous in unprocessed pre-dawn TIR imagery. At Brady's Hot Springs, Nevada, similar algorithms didn't increase the size of the known geothermal anomaly, but the magnitudes of “false” temperature anomalies unrelated to geothermal activity were significantly reduced.