GIS-based multi-criteria predictive modelling for geothermal energy exploration
Year: 2025
Type: article
Abstract: Renewable energy resources, including geothermal, are crucial for sustainable environmental management and climate change mitigation, offering clean, reliable, and low-emission alternatives to fossil fuels that reduce greenhouse gases and support ecological balance. In this study, geographic information system (GIS) predictive analysis was employed to explore geothermal prospects, promoting environmental sustainability by reducing the dependence on fossil energy resources. Spatial and statistical analysis including the attribute correlation analysis was used to evaluate the relationship between exploration data and geothermal energy resources represented by hot springs. The weighted sum model was then used to develop geothermal predictive maps while the accuracy of prediction was determined using the receiver operating characteristic/area under curve (ROC/AUC) analysis. Based on the attribute correlation analysis, exploration data relating to geological structures, host rock (Asu River Group) and sedimentary contacts were the most critical parameters for mapping geothermal resources. These parameters were characterized by a statistical association of 0.52, 0.48, and 0.46 with the known geothermal occurrences. Spatial data integration reveals the central part of the study location as the most prospective zone for geothermal occurrences. This zone occupies 14.76 % of the study location. Accuracy assessment using the ROC/AUC analysis suggests an efficiency of 81.5 % for the weight sum model. GIS-based multi-criteria analysis improves the identification and evaluation of geothermal resources, leading to better decision-making. • Integrated multiple spatial datasets including attribute correlation analysis for prediction. • Discretized geothermal predictive map for potential zones at an accuracy of 81.5 %. • Valid and reliable GIS-based predictive models for geothermal exploration.
Source: Energy Geoscience
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4409204968