Abstract: Flood coincidence probability analysis is essential to flood control and disaster reduction. Previous studies on multi-source flood coincidence are limited to asymmetric Archimedean Copula that has the defects of complicated formula, more parameters, and poor expansibility. This paper adopts an elliptic Copula function to construct numerical models for the coincidence of occurrence dates, dates-magnitudes, and runoff processes of multi-source floods in the Dongting Lake basin. The probability of flood coincidence is calculated for multi-type and multi-scenario cases. We reveal time delay in the coincidence process of the floods from three tributaries and compare the model calculations with the statistical data of synchronous field observation. The calculations are consistent with the natural patterns of flood occurrence and basically agree with the field observations, showing the reliability of our model. Thus, the proposed method helps improve river basin flood control and project safety against floods.

Key words: Dongting Lake basin, multi-source flood coincidence, Copula function, multi-scenarios, time delay pattern