Code for "Hidden degrees of freedom in weather-based risk indices: a methodological sensitivity analysis with a mycotoxin case study"

Analysis and figure-generation code accompanying the study “Hidden degrees of freedom in weather-based risk indices: a methodological sensitivity analysis with a mycotoxin case study”. The study evaluates the methodological robustness of a rule-based weather favourability index for Fusarium head blight (FHB) and deoxynivalenol (DON) in Estonian cereals. The FHB/DON index is used as a case study to examine how defensible methodological choices affect individual classifications and aggregate conclusions. The archive contains Python code for one-at-a-time sensitivity analyses of baseline period, threshold-derivation method, wet-day definition and weather-station network; a full factorial ensemble of 36 methodological configurations; a diagnostic analysis of wet-spell threshold derivation; and scripts for reproducing manuscript Figures 1–4. Source meteorological observations are not redistributed in this archive. They are publicly available from the Estonian Environment Agency. The included README provides data-source information, the required input-data structure, dependencies and instructions for reproducing the analyses and figures.