Random Forest Assessment of Drought Mitigation from Bush Encroachment Control Using Standardized Runoff and Soil Moisture Indices (1982–2100)

This repository contains two Jupyter notebooks implementing a Random Forest–based assessment of drought mitigation potential from bush encroachment control. Hydrological drought was quantified using the Standardized Runoff Index (SRI) derived from ERA5-Land runoff, and agricultural drought using the Standardized Soil Moisture Index (SSI) derived from FLDAS soil moisture (1982–2100). Climate predictors were generated from an ensemble of 10 CMIP6 GDDP models aggregated to monthly means/sums. The workflow integrates Google Earth Engine with Python for data preprocessing, index calculation, and machine-learning analysis to evaluate how changes in vegetation structure influence runoff and soil-moisture–related drought dynamics.