A Statistical Framework for Spatial Boundary Estimation and Change Detection: Application to the Sahel Sahara Climate Transition
The framework uses heteroskedastic Gaussian process regression and a scaled Maximum Absolute Difference test to estimate and track spatial climate boundaries over time. Applied to the Sahel-Sahara transition zone, it found no decade-scale changes but successfully detected localized boundary shifts during extreme drought years in 1983 and 1984, confirming its ability to find year-specific anomalies.