As the core element triggering urban waterlogging, disaster-causing rainfall is difficult to predict accurately due to its randomness and uncertainty. Identifying disaster-causing rainfall and clarifying its evolutionary characteristics and development patterns are of great significance for flood disaster reduction. The research takes the High-tech Zone of Zhengzhou City in China as the study area, constructs an SWMM model for urban waterlogging simulation, and identify the disaster-causing rainfall events based on waterlogging risk process. Then, the copula function was employed to construct multidimensional joint distributions of the key characteristics of disaster-causing rainfall to assess their behaviors. The results are as follows: (1) A total of 48 disaster-causing rainfall events were identified in the study area, with event-scale precipitation ranging from 32.5 to 86 mm. (2) The joint probability tends to increase with the magnitude of rainfall characteristics, and the event return periods mainly concentrates between 1 and 10 years. Two-dimensional event return periods may overestimate disaster risk compared to three-dimensional ones. (3) The synchronous encounter probability among total rainfall, rainfall intensity, and peak rainfall was found to be substantially greater than the asynchronous probability, indicating a strong positive correlation among these variables.
Li et al. (Fri,) studied this question.