Key points are not available for this paper at this time.
Flooding is a recurrent hazard in Malawi, particularly in the Shire River Basin (SRB), where low-lying terrain, rapid population growth, and reliance on rain-fed agriculture increase flood exposure. While previous studies have examined rainfall variability and flood hazards, few have integrated long-term rainfall analysis, flood mapping, and spatial exposure assessment in their analytical frameworks. This study combined 34 years (1990–2024) of daily CHIRPS rainfall data with Synthetic Aperture Radar (SAR) imagery to derive rainfall thresholds associated with flood onset and map inundation extent of four major flood events (2015,2019,2022 and 2023). Rainfall thresholds were evaluated using cumulative rainfall windows (1–14 day), ROC and precision-recall analyses, and cost-weighted optimization. The 7-day rainfall accumulation produced the ideal operational performance (ROC-AUC = 0.987; 95% CI: 0.980–0.993), with an optimized operational trigger threshold of 101.04 mm. Flood inundation ranged from 33,950 ha to 58,172 ha. Persistent hotspots (15,553 ha) were identified. The 2023 event produced the highest estimated population exposure (157,509 people). Scenario analysis indicated that present-day exposure conditions would substantially increase the estimated population and agricultural exposure under historical flood extents. These findings provide a consistent quantitative basis for combining short-term anticipatory flood early warning, with long-term exposure management in SRB.
Kapichi et al. (Mon,) studied this question.