Spatiotemporal patterns and driving factors of flood disaster in China

Abstract. Flood is one of the most disastrous disasters in the world inflicting massive economic losses and deaths on human society, and it is particularly true for China which is the home to the largest population in the world. However, no comprehensive and thorough investigations have been done so far addressing spatiotemporal properties and relevant driving factors of flood disasters in China. Here we investigated changes of flood disasters in both space and time and their driving factors behind using statistical data of the meteorological disasters from Statistical Yearbooks and also hourly rainfall data at 2420 stations covering a period of 1984–2007. GeoDetector method was used to analyze potential driving factors behind flood disasters. We found no consistent extreme rainfall trend across China with exceptions of some sporadic areas. However, recent years witnessed increased frequency of rainstorm-induced flood disasters within China and significant increase in the frequency of flood disasters in the Yangtze River, Pearl River and Southeastern coasts. Meanwhile, reduced flood-related death rates in the regions with increased flood frequency indicated enhanced flood-mitigation infrastructure and facilities. However, increased flood-induced affected rates and direct economic losses per capita were found in the northwestern China. In addition, contributions of influencing factors to the spatio-temporal distribution of flood disasters analyzed by GeoDetector are shifting from one region to another. While we found that rainfall changes play the overwhelming role in driving occurrences of flood disasters, other factors also have considerable impacts on flood disasters and flood disaster-induced losses such as topographical features and spatial patterns of socio-economy. Wherein, topography acts as the key factor behind the characteristics of spatial distribution of flood disasters in China.