Derived datasets for: "Rating Curves Dominate Flood Exposure Uncertainty in Data-Scarce Transboundary Basins"

This dataset contains all derived outputs from the study "Rating Curves Dominate Flood Exposure Uncertainty in Data-Scarce Transboundary Basins" (Ismail and Yamashiki, submitted to Natural Hazards). Files include the 47-year Annual Maximum Series (1979–2025) for five Pakistan drainage basins derived from GloFAS ERA5 reanalysis using spatial averaging (BSFMI method), fitted Generalized Extreme Value (GEV) parameters and design discharge return levels (T50–T500), HAND-based inundation areas, GHS-POP 2025 population exposure and GHS-BUILT built-up area exposure at T50–T500, Manning n=0.055 upper bound exposure estimates, rating curve sensitivity grid outputs (3×3 parameter combinations), first-order uncertainty decomposition results, and 90% combined probabilistic exposure ranges. These files reproduce all numerical results reported in Tables 1–4 and Supplementary Materials S1, S9, and S10 of the associated manuscript. File What it contains AMS_5basins.csv 47-year Annual Maximum Series for all 5 basins GEV_parameters_47yr.csv μ, σ, ξ for all 5 basins GEV_return_levels_47yr.csv T50–T500 discharges and depths inundation_summary_47yr.csv HAND inundation areas for T50–T500 exposure_summary_47yr.csv GHS-POP population exposure for T50–T500 n055_exposure_47yr.csv Manning n = 0.055 upper bound exposure sensitivity_full_47yr.csv Full 3×3 rating curve sensitivity grid sensitivity_summary_47yr.csv Min / baseline / max / range per basin sensitivity_decomposition_47yr.csv % contribution per uncertainty source probabilistic_exposure_ranges_47yr.csv 90% confidence interval bounds per basin