Work

A multi-scenario framework for quantifying flood hazard and exposure accounting for runoff-driven uncertainty in global flood models

Year: 2025

Type: article

Abstract: With growing populations and an increasing frequency of flood events, large-scale flood hazard assessment (LSFHA) and exposure analyses have become critically important. Global Flood Models (GFMs) significantly contribute to these efforts by simulating flood dynamics based on runoff inputs from Land Surface Models (LSMs), Global Hydrological Models (GHMs), or Reanalysis datasets. However, GFM outputs remain highly sensitive to runoff input choice, leading to substantial uncertainty in LSFHA. To address this challenge, we develop a multi-scenario framework integrating diverse runoff forcings into the CaMa-Flood GFM to capture a plausible range of hazard and exposure outcome scenarios, ranging from optimistic to conservative. The framework defines optimistic and conservative scenarios as the minimum and maximum simulated flood depth among all simulations, while the best-case is derived from the most efficient and validated simulation. ERA5-Reanalysis runoff forced streamflow simulation emerges as the most efficient, achieving Nash–Sutcliffe efficiency greater than 0.5 at approximately 50 % of analysed gauge stations, thus representing the best-case scenario. For a precise estimation of the return period flood depth, multiple parametric and non-parametric distributions are employed in frequency analysis. Hazard analysis for a 1-in-100-year flood event reveals around 48 % (range: 20 % - 60 %) of India's land area falls in the disastrous hazard class, approximating 1.6 (range: 0.65–1.95) million km 2 . Exposure analysis, aligning closely with government estimates and previous studies, indicates approximately 690 million people (range: 335–786 million) are exposed to significant flood risk. Validated using India as a case study, this scalable framework offers an open-source tool enabling national-scale risk-informed decision-making, addressing runoff-driven uncertainty critical for both immediate and long-term flood risk management.

Source: The Science of The Total Environment

DOI: 10.1016/j.scitotenv.2025.180381

Authors: Jayesh Parmar, Subhankar Karmakar

Institution: Indian Institute of Technology Bombay

Language: English

Topic: Flood Risk Assessment and Management

Subfield: Global and Planetary Change

Field: Environmental Science

Domain: Physical Sciences

SDG: Climate action

Open Access status: hybrid

Estimated APC paid: $4,250