Scenario-driven data fusion for compound coastal flood risk and exposure assessment using night-time lights in Jeddah

Coastal cities are increasingly vulnerable to compound flooding caused by the combined effects of sea level rise, storm surge, tides, rainfall, and atmospheric pressure. However, integrated and scenario-based assessments remain limited, particularly in rapidly urbanizing and data-scarce regions. This study develops a rainfall-adjusted, scenario-driven bathtub flood model for Jeddah, Saudi Arabia, to project future flood extent, depth, and urban exposure in 2030, 2050, and 2100 under the shared socioeconomic pathways (SSPs). The model integrates open-access earth observation datasets with scenario-specific flood drivers, including sea level rise, storm surge, wind speed, and atmospheric pressure, to simulate total water levels (TWL). Urban exposure is estimated using night-time light (NTL) intensity as a proxy for population and infrastructure density. Model performance is evaluated against the November 2022 Jeddah flood event, achieving strong predictive skill (AUC = 0.872). Under the high emission SSP5-8.5 scenario, flooded areas increase from 62.38 km² in 2030 to 233.70 km² by 2100, with mean water depths exceeding 2.5 m in densely populated zones. The proposed rainfall-adjusted bathtub approach offers a computationally efficient and physically plausible framework for compound flood assessment. Its spatially explicit outputs support climate adaptation planning, exposure mapping, and nature-based solution design in data-limited coastal environments.