Savanna fire regimes across sub-Saharan Africa have undergone documented shifts over recent decades, yet the relative contributions of climate forcing, land use restructuring, and demographic change remain poorly quantified at national scale. This study presents a 22-year burned area analysis for Zimbabwe (2001–2022) using MODIS MCD64A1 v061 at 500 m resolution, processed entirely in Google Earth Engine (project: ee-muhambatawanda2024). Mann-Kendall trend analysis reveals no monotonic decline (p = 0.274), but a Welch two-sample t-test identifies a significant structural step change of −27.6% between Period 1 (2001–2011, mean = 3.85 Mha/yr) and Period 2 (2012–2022, mean = 2.79 Mha/yr), at p = 0.010. Partial correlation analysis attributes this regime shift primarily to agricultural land expansion (r = −0.59, p = 0.004) and population growth (r = −0.49, p = 0.020), with direct ENSO forcing proving statistically negligible (r = −0.05, p = 0.820). Spatially, a persistent fire-suppressed Central Plateau contrasts with high-frequency fire zones in the Zambezi Valley (18–20 of 22 years burned) and Hwange National Park (15+ years). A four-module process-based human–fire interaction model framework is proposed — comprising fuel dynamics, ignition probability, fire extent, and human suppression modules — directly addressing the INFLAMES WP6.3 objective of replacing population-density proxies with explicit land-use and suppression parameterisations in JSBACH-SPITFIRE. Prepared as part of PhD interview materials for the INFLAMES project, Faculty ITC, University of Twente, May 2026. Data sources: MODIS MCD64A1 v061 · NOAA ONI · CHIRPS v2.0 · WorldPop · FAOSTAT · GADM v4.1Analysis platform: Google Earth Engine (ee-muhambatawanda2024)