Assessing Traffic Resilience: A Spatiotemporal Analysis of the Extreme Rainstorm in Beijing, 2023

Climate change leads to frequent and intense extreme weather, posing threats to densely populated urban areas. The non-engineering measure (early warning system, school closure, work from home) is an effective strategy for future adaptation, but research in megacities of developing countries remains limited. In this context, our study focuses on Beijing, the capital of China, examining a case of the 140-year rainstorm that occurred in late July 2023. Using 30-minute interval road congestion data, our study conducts a quantitative analysis of the urban operational rhythms. We found that citizens actively responded to the disaster event, reducing traffic pressure and travel exposure on rainy days. Specifically, when the red warning was issued, road congestion experienced a slight increase with an early rush hour, reflecting the preparatory actions before the rainstorm; while in the following three days, road congestion was below normal levels, indicating the cancellation of numerous activities during the rainstorm. The complex congestion dynamics can be categorized into three modes: the intensely changing evening peak, the locally changing off peak, and the relatively stable morning peak. We then clustered the urban internal spaces based on typical mode proportion and finally classified them using multi-source variates. It was found that population characteristics, location accessibility, and built facilities can effectively explain the road congestion change. Based on these findings, policy suggestions for extreme disaster response and future resilience enhancement are proposed.