Geothermal Detection Study Using Remote Sensing Data by Combining Machine Learning and Deep Learning: A Case Study of Huanggang City

Abstract

Energy science has significantly advanced societal progress, and the use of renewable energy has become a universal consensus. Among these, geothermal energy offers the advantages of being green, environmentally friendly, efficient, stable, and highly utilized. However, detecting geothermal resources involves significant uncertainty. Remote sensing (RS) data and artificial intelligence (AI) have shown immense potential in overcoming these challenges. To achieve geothermal detection, this study designs a geothermal detection method based on RS and AI, taking into account various geothermal factors, including land surface temperature, magnetic anomaly, gravity anomaly, distance to faults and rivers, nighttime light, land use type, landform, lithology, and more. The detection process is divided into two stages: coarse detection using machine learning (ML) methods such as the Information Model (IM), Artificial Neural Network (ANN), Logistic Regression (LR), One-Class Support Vector Machine (OCSVM), Support Vector Machine (SVM), and Random Forest (RF). Then, the coarse geothermal detection results are combined with fine-grained detection using a multi-channel U-shaped deep learning network (MUnet) to achieve high-quality detection. Taking Huanggang City as the research area, the results show that (1) The Precision and Recall of these models are as follows: IM (47.76% and 72.73%), ANN (55.22% and 90.24%), LR (76.92% and 85.11%), OCSVM (67.69% and 93.62%), SVM (83.02% and 91.67%), and RF (84.62% and 93.62%). Among them, the RF model performed the best, but its GDA is still 24.43%, indicating that 4,262 km2 of the study area is classified as geothermal units, which is quite large. (2) The fine-grained detection model MUnet achieved a Precision of 86.54%, a Recall of 95.74%, an F1 score of 90.91%, and a GDA of 2.82%, highlighting its exceptional sensitivity in capturing geothermal regions. The MUnet model's series of feature extraction and progressive decoding processes enables it to generate precise geothermal detection results from multi-channel inputs. (3) The combined RF_M method outperformed both the RF and MUnet models individually, achieving a precision of 93.48%, a recall of 91.49%, an F1 score of 92.47%, and a GDA of 1.94%. This demonstrates that the RF_M model effectively balances precision and recall, providing the most detailed and reliable geothermal detection results while minimizing false positives.

Keywords: Remote sensing data, Machine Learning, Deep learning, Geothermal detection