Formation and modeling of flood inundation in a plain metropolis: lessons from the 7.20 Zhengzhou extreme rainstorm

Abstract

An extreme rainstorm induced by the joint effect of northward jump of the subtropical high, Typhoon In-fa and the northwest terrain wreaked havoc in Henan in late July 2021. The metropolis Zhengzhou bore the brunt, where 201.9 mm hourly rainfall was observed, with 297 people killed, 47 people missing, and direct economic loss of more than 53 billion RMB. Despite numerous weather forecasts beforehand, there lacked an efficient and accurate early warning system for flood inundation; the characteristics of the millennium event are also worthy of detailed investigation. In this light, we perform numerical simulations of the event by a full 2D hydrodynamic model. The model is calibrated and validated based on the inundation observations derived from multiple data sources from remote sensing and social media. Our study aims to reveal 1) How and how fast the inundation was shaped particularly in some vulnerable locations (e.g. hospitals, tunnels, and subway channels); 2) How long the leading time can be with an essential accuracy of the inundation area; 3) If and how the vulnerable region is sensitive to spatial and temporal patterns of rainstorms. Considering the greening measures as well as sponge city projects in Zhengzhou in recent years, the simulation result will also be able to illustrate the dynamic contribution of various city elements (streets, sewer system, gardens, canals, etc.) in holding the stormwater. The findings will provide a valuable understanding of flood inundation processes in the plain metropolis. Besides, the study can foster awareness of the necessity of inundation early warning systems particularly for vulnerable sites, which also provides an essential reference for the emergency response of local governments.