Assessing compound flood drivers in Peninsular India: Multivariate copula-based approach

Floods are among the most devastating natural disasters, and their frequency and intensity have been exacerbated by climate change. Compound flooding, driven by multiple drivers has been largely unexplored in India. In Peninsular India, flood events are primarily driven by precipitation, runoff, and antecedent soil moisture conditions. Traditional univariate analyses fail to account for the interdependence of these factors, potentially underestimating flood risk. This study employs an advanced multivariate analytical framework using a copula-based approach to assess compound flooding in six major river basins: Narmada, Tapi, Mahanadi, Godavari, Krishna, and Cauvery. Using daily gridded precipitation data from IMD, runoff data from MERRA-2, and soil moisture data from GLEAM (1980-2023), the study analyzes extreme events at 90th, 95th and 99th percentile thresholds. This study is the first in Peninsular India to comprehensively analyze compound flooding using copula-based techniques. Multivariate copula models are used to examine joint dependencies between precipitation, runoff, and soil moisture. Exceedance probability, conditional probability, and joint return period analyses provide insights into the likelihood of extreme flood events occurring under different scenarios. Bivariate results indicate that runoff and antecedent soil moisture significantly influence compound flooding, particularly in Krishna and Cauvery basins, where prolonged wet soil conditions exacerbate flood risks. The trivariate analysis reveals that Mahanadi basin is the most prone to extreme compound flood scenarios. Findings emphasize the need for improved flood risk assessment and adaptive management strategies, particularly given the increasing impacts of climate change. The results can aid in flood forecasting, infrastructure planning, and policy formulation to mitigate future flood damage in India's river basins.