Flood risk in India's arid Ramsar wetlands: Integrating extreme rainfall, fuzzy risk assessment, and CMIP6 projections

Abstract

Arid Ramsar wetlands in India face escalating flood risk as anthropogenic climate change intensifies extreme rainfall events and amplifies the influence of large-scale climate oscillations. Here we present an integrated assessment of flood hazard, risk, and future climate trends across 17 arid Ramsar wetlands using three complementary analytical frameworks. First, a nonstationary Generalized Extreme Value (GEV) analysis was applied to daily rainfall data from the India Meteorological Department (IMD) spanning 1951–2022, incorporating climate oscillations as time-varying covariates across 56 linear model combinations. Return levels of four extreme precipitation indices, including daily maximum rainfall (Rx1), Simple Precipitation Intensity Index (SDII), heavy rainfall day count (R10), and Consecutive Wet Days (CWD), were estimated using Bayesian inference with 95% credible intervals. Khijadia Wildlife Sanctuary recorded the highest return levels for SDII and R10, while Nalsarovar led for Rx1. Second, a fuzzy logic risk framework integrating hazard, vulnerability, and exposure classified six wetlands, such as Sarsai Nawar Jheel, Beas Conservation Reserve, Chitrangudi, Kanjirankulam, Ranganathittu, and Saman Bird Sanctuaries, as being at very high flood risk. Critically, the highest return levels and the highest risk rankings do not coincide, underscoring the indispensable role of vulnerability and exposure in determining composite risk. Third, Modified Mann-Kendall trend analysis applied to 13 CMIP6 model projections under SSP245 and SSP585 confirmed statistically significant increasing trends in precipitation and temperature across nearly all sites. These findings provide an evidence-based foundation for prioritizing adaptive flood management at India's most vulnerable arid wetlands. © 2026 Elsevier Ltd.