Multivariable flood risk and its dynamics considering project reasonable service life in a changing environment

Abstract Concern over the impacts of changing environments on extreme floods has made the multivariable flood risk assessment paid more attention by the whole world. Recent studies have taken the combination of flood peak, flood volume, and duration into account when assessing flood risk. Nevertheless, the dynamics of multivariable flood risk considering project reasonable service life remains unresolved. To this end, flood peak and flood volume were considered, and Copula function was used to describe their dependent structure, considering project effective service life based on the joint return period to calculate the risk rate for characterizing flood risk. Then, a sliding window of 25 years was adopted to explore its dynamics, and attribution analysis was subsequently conducted. The Pai-lung River, which belongs to the Yangtze River Basin in China, was selected as a case study. Results show that: (1) on basis of “and (∩)” joint return period, considering 50-year reasonable service life, the risks of slight, moderate, severe, and extreme floods are 1.00, 0.96, 0.66, and 0.06, respectively; (2) the flood risk dynamics vary with different levels, where the risks of slight and moderate floods tend to decrease whilst the risk of severe and extreme floods shows a significant increase; (3) the risk dynamics of severe and extreme floods are positively related to Sunspots, El Nino Southern Oscillation and Arctic Oscillation, while Pacific Decadal Oscillation has a weaker influence on flood risk dynamics. Generally, this study sheds new insights into multivariable flood risk assessment in a changing environment.