Feature, Label, Prediction, and Zone Data for Great Basin Zone Classification and Hydrothermal Favorability Modeling Using Supervised Machine Learning

These datasets were compiled to support the publication Mordensky et al. (2026) that discusses how supervised machine learning can be used to delineate zones that vary according to feature data and predict which locations in a given region are more favorable for geothermal activity. The datasets contained herein are the point data for the 53 publicly available input features organized by labeled (i.e., well data in which modeled departures from conductive heat flow serve as labels in supervised machine learning) and unlabeled examples (i.e., grid data without modeled departures from conductive heat flow) used to produce the hydrothermal favorability model for the region delineated as the INnovative Geothermal Exploration through Novel Investigations Of Undiscovered Systems project (INGENIOUS; Ayling et al. [2022]), and the median hydrothermal favorability map (prediction percentile = P50) of the 120 realizations from the model presented in Mordensky et al. (2026). Each favorability map depicts an estimate of relative favorability with respect to the other locations (i.e., cells). Corresponding geoTiff files are located in the subdirectory. Note that the grid_features.csv (input_data/CSVs/), cluster_and_zone_assignments.csv (output_data/CSVs/), and median_predictions_hydrothermal_favorability.csv (output_data/CSVs/) files have too many examples to open in MS Excel. We suggest opening this file using Notepad or Notepad++. As of ArcGIS Pro 3.5.5, ArcGIS Pro could import and display the data in gridFeatureData.csv file on a computer with at least 64 GB of RAM.