Using Official IDF Benchmarks to Calibrate GPM IMERG Rainfall for Drainage Infrastructure Assessment in Data-Scarce Coastal Cambodia

Reliable rainfall intensity-duration-frequency (IDF) information is essential for drainage design, yet such information is often difficult to access in rapidly developing coastal regions. In Sihanoukville, Cambodia, official design-standard IDF data are available, but raw rain gauge records and station coordinates are not publicly disclosed. This study evaluates the applicability of GPM IMERG rainfall estimates against the official Sihanoukville IDF benchmark and develops a practical correction framework for drainage infrastructure assessment in a data-scarce environment. The official IDF table issued by the Department of Hydrology, derived from rainfall records for 1999–2019, was adopted as the engineering benchmark. IMERG half-hourly rainfall data for 2000–2025 were extracted in Google Earth Engine for the grid cell covering the study area. IMERG-derived rainfall intensities for the overlapping period 2000–2019 were compared with the official benchmark across selected durations and return periods to derive duration-specific adjustment factors. The 2020–2025 IMERG subset was then used for temporal consistency assessment. DEM-based watershed delineation identified a 3.43 km² external catchment draining to a representative perimeter road crossing adjacent to a downstream development area. Results show that IMERG severely underestimates short-duration design rainfall, by approximately 63–74% at 30–60 min durations, with the bias decreasing as storm duration increases. The derived duration-specific correction factors were 3.65 for 30 min, 2.90 for 1 h, 2.08 for 2 h, 1.69 for 4 h, and 1.24 for 24 h. For an illustrative 1 h, 10-year runoff scenario under present-condition catchment runoff (C = 0.55), peak discharge at the representative road crossing increases from 19.83 m³/s using raw IMERG rainfall to 57.56 m³/s after benchmark correction, indicating that direct use of uncorrected satellite rainfall would substantially underestimate hydraulic demand. The proposed benchmark-based calibration approach offers a practical pathway for improving drainage resilience in rapidly urbanizing coastal settings where official design information is incomplete and ground observations are not publicly accessible.