Data and code for "Moisture-Gradient and Environmental Dependence of Daytime Land Surface Temperature Effects at Photovoltaic Plants in China's Drylands"

This repository contains the processed data and analysis code supporting the study “Moisture-Gradient and Environmental Dependence of Daytime Land Surface Temperature at Photovoltaic Power Plants in China's Drylands.” It includes radial temperature profiles, temperature differences (TD), thermal influence distances (DM), site-level environmental predictors, and grouped statistics for 227 photovoltaic power plants across four seasons. The archive also provides nested cross-validation results, 40 trained random forest models, held-out SHAP values for TD and DM, predictor correlations, and threshold sensitivity analyses. Python scripts support TD/DM calculation, statistical summaries, random forest training, SHAP analysis, and analysis plotting. The README describes the archive structure, dependencies, and reproduction workflow.