| Abstract |
The INnovative Geothermal Exploration through Novel Investigations Of Undiscovered Systems (INGENIOUS) project aims to facilitate the discovery of new, economically viable hidden geothermal systems in the Great Basin region of the western United States by building on previous play fairway analysis and machine learning work. A key objective of this project is to develop an exploration workflow to reduce exploration risks, particularly for hidden geothermal systems. The workflow was based on the statistical assessment and integration of regional geological, geophysical, and geochemical datasets. This led to the generation of new predictive geothermal fairway maps for the INGENIOUS study area, which encompasses most of Nevada, western Utah, southern Idaho, southeastern Oregon, and easternmost California. The work presented here updates the preliminary INGENIOUS workflow by incorporating multiple versions of negative training sites and by implementing a systematic evaluation of logistic-regression model performance. In order to more directly compare the resulting models, a prior correction method was applied to adjust model intercepts and align predicted probabilities with more realistic prevalence. To assess logistic regression modeling error, a residual-analysis framework enabling bin-based comparison of predicted and empirical probabilities was applied. Results show that all model versions exhibit stable, near-zero residuals across well-supported Z ranges, with only one version displaying a small but systematic calibration trend. These findings demonstrate that geothermal favorability models generally behave consistently and predictably across the Z-space. |