Urban rainstorm flood rapid simulation in plain river network area based on cellular automata

Choosing a flood simulation model with acceptable accuracy may be the first step towards achieving urban flood management in plain river network area. Hydrologic-hydrodynamic coupling models require various data that are difficult to obtain, and simplified models such as cellular automata (CA) appear. The purpose of this research is to examine the performance and uncertainty of CA model for simulating flood in plain river network region, and to explore the management of urban waterlogging by scenarios analysis. The results show that: (1) The CA model’s performance is all within an acceptable range, with average correlation coefficient (R) and Nash–Sutcliffe efficiency coefficient of 0.88 and 0.87; (2) Elevation had the greatest influence on the model simulation result, followed by roughness; (3) The model climate change hydrological response analysis revealed that, the suburbs of Yinzhou’s southeast plain were more susceptible to flooding than the urbans; (4) According to the urbanization hydrological response study, maintaining the land permeability rate and the river network water surface rate both have a positive effect on reducing the degree of flooding. In light rain, the influence of river network change (1.37%) is stronger than land use change (0.18%). While in heavy rain, the effect of both is comparatively weaker, less than 1%. (5) Low-impact development measures are effective in controlling urban flooding. The shorter the storm recurrence period, the greater the weakening effect. In this paper, the result and modeling process of CA model are investigated to provide a new option for flood assessment and management in plain river network area.