Processed data and code for Wet-Year Precipitation Anomalies and Geohazard Economic Losses in China: A Province-Year Sensitivity and Climate-Exposure Assessment

This dataset provides processed data and R/Python code for reproducing the analyses, figures and supplementary tables of the revised manuscript "Wet-Year Precipitation Anomalies and Geohazard Economic Losses in China: A Province-Year Sensitivity and Climate-Exposure Assessment" (IJDRR-D-26-01970).

Version 3.1.0 contains IJDRR_processed_reproduction_v3.zip. The main model and the annual GPCC benchmark in Supplementary Table S4 use the same stored precipitation predictor, outcome and 620 province-year observations. Refitting gives a coefficient of 0.758437676133168. Monthly, seasonal and daily-derived indicators follow their documented definitions. Table S3 reports the main-model increment in millimetres separately from descriptive precipitation variability.

Start with README.md for installation and execution. The complete calculation entry point is code/run_all.py; code/run_figures.py generates the figures. TABLE_MAP.csv links supplementary tables to the corresponding calculation modules and numerical outputs. README_DATA.md and the data dictionaries explain variables and units.

Focused numerical verification covers the main model, 134 comparison specifications and physical-scale calculations, with independent checks of coefficients, covariance estimates and multiple-testing adjustments. A separate matched-product execution check uses 19 draws. See validation/README.md for the verification scope. The focused entry point is code/run_R2_6_update.py.