Assessing Urban Areas’ Vulnerability to Flood Disaster Based on Text Data: A Case Study in Zhengzhou City

Abstract

Confronting the frequent flood disasters triggered by torrential downpour, the vulnerability of urban rainstorm flood disasters was analyzed with one highly popular area of research in mind: big data. Web crawler technology was used to extract text information related to floods from Internet and popular social media platforms. Combining these text data with traditional statistical data, a flood disaster vulnerability assessment model based on Analytic Hierarchy Process (AHP) was established to evaluate rainstorm and flood disaster vulnerability, and the spatial distribution characteristics of vulnerability to pluvial flooding were analyzed based on Geographic Information System (GIS). The established model was applied in Zhengzhou, a city that often suffers from heavy rainstorms. The results show that the areas located near downtown Zhengzhou were more vulnerable to rainstorm and flooding than others, and most of the city could be at moderate and high vulnerability. Finally, the waterlogging spots extracted from various sources were used to evaluate the performance of the proposed model. The results show that most of waterlogging spots were located in very-high and high risk zones, while less waterlogging spots were found in districts with low vulnerability, which demonstrates the discriminative power of the established model based on big data sources. This study overcomes limited data in flood disaster vulnerability assessment methods and provides a basis for flood control and management in cities.