A Precise Prediction Method for Subsurface Temperatures Based on the Rock Resistivity–Temperature Coupling Model

The accuracy of deep temperature predictions is critical to the precision of geothermal resource exploration, assessment, and the effectiveness of their development and utilization. However, the existing methods encounter significant challenges in predicting the distribution characteristics of deep temperature fields with both efficiency and accuracy. Many of these methods rely on empirical formulas to approximate the relationship between geophysical parameters and temperature. Unfortunately, such approximations often introduce substantial errors, undermining the reliability and precision of the predictions. We present an advanced prediction methodology for deep temperature fields based on the rock resistivity–temperature coupling model (RRTCM). By converting the fixed parameters in the empirical formulas to variables dependent on the formation depth, we establish a dynamic model that correlates rock resistivity with temperature on the basis of limited constrained borehole data. We then input the 2D magnetotelluric inversion results into the model, and the subsurface temperature distribution can be predicted indirectly with high precision on the basis of the resistivity–temperature coupling relationship. We validated this method in the Xiong’an New Area, China, and the determination coefficient (R2) of maximum temperature prediction reached 98.88%. The sensitivity analysis indicates that the prediction accuracy is positively correlated with the number and depth of the constrained boreholes and negatively correlated with the sampling interval of the well logging data. This method robustly supports geothermal resource development and enhances the understanding of geothermal field formation mechanisms.