Understanding the influencing factors and controls of rainstorm-induced floods, which have caused tremendous losses of human lives and national economy, is a pressing need for flood risk management in China. Based on the meteorological disaster census data of counties in China, hourly precipitation data at 2420 stations, statistical yearbook, terrain data and other data, the authors (1) investigated the spatiotemporal pattern of flood impacts in China over the period from 1984 to 2007 using trend analysis techniques and (2) explored the driving factors of the spatiotemporal pattern by adopting the geospatial statistical analysis tool (Geodetector). This study considered the spatiotemporal patterns and their interplays among county-level flood impacts (i.e., flood-induced mortality rate, proportion of population affected, and economic loss in percentage), disaster-formative environmental factors (i.e., population density, urban population percentages, average elevation, river density, average slope, and average distance to the seashore), and extreme precipitation characteristics (i.e., annual average volume and duration of extreme rainfall). The results show that: (1) there were no consistent temporal trends of extreme rainfall characteristics over the study period across China. (2) The frequencies of flood disasters in the Yangtze and Pearl rivers and southeast coastal areas increased significantly, but the casualties over these regions decreased. (3) Flood-induced casualties, proportion of population affected and economic loss in percentage increased in Northwest China; and meteorological factors, disaster-formative environment factors such as geographical conditions and social economy, and geographical conditions contribute mostly to the proportion of population affected, flood-induced death and economic loss in percentage. These results indicate that more attention should be paid to improving the flood control capacity of small or medium-sized cities in the inland river basins, especially in Northwest China, and we should recognize the important roles that disaster-formative environment plays in triggering flood losses.

Keywords： extreme rainfall ; rainstorm-induced flood ; flood disaster ; spatiotemporal pattern ; Geodetector ; influencing factor ; China