Published In

Geothermics

Document Type

Article

Publication Date

12-1-2025

Subjects

Geothermal resource assessment, Hydrothermal favorability, Great Basin, Data-driven -- Machine learning

Abstract

Highlights

  • • The new approach to predict hydrothermal resource favorability for the U.S. Great Basin is a synthesis of modern data-driven machine learning improvements from the last several years and predicts 85 % of power-producing systems with operating power plants in the most favorable 10 % of the total area with over half of the power-producing systems (10 of the 19 exposed power-producing systems and 5 of the 9 hidden power-producing systems) in the 99th percentile.
  • • The new hydrothermal favorability map predicts both hidden and exposed power-producing hydrothermal systems equally well.
  • • The new hydrothermal favorability map preferentially predicts systems with greater convective heat flow than systems with comparatively less convective heat flow.
  • • The new hydrothermal favorability map outperforms previous mapping approaches due to improvements in methods, improved input data/evidence layers, and a significant increase in the number of training sites.
  • • Monte Carlo cross-validation (n = 1200) is used to estimate predictive uncertainty in the favorability maps, with the final best map being the median prediction for each location.

Rights

Copyright (c) 2025 The Authors

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.

DOI

10.1016/j.geothermics.2025.103450

Persistent Identifier

https://archives.pdx.edu/ds/psu/44103

Publisher

Elsevier BV

Share

COinS