基于统计-FloodArea模型的平原水网区致灾临界雨量研究

Research of Critical Rainfalls in Wet-net Plains Based on Statistical Method and FloodArea

论文基于江苏老濉河流域降水、水位和流量观测数据,利用统计-Flood Area模型相结合的方法计算了平原水网区致灾临界雨量。从水位变幅角度构建了流域不同时段累计面雨量-洪水涨幅统计模型,并计算出致灾临界雨量。在修订河道高程的基础上,利用Flood Area模型模拟典型暴雨洪涝事件的动态淹没过程,建立不同时段累计面雨量与模拟水位涨幅之间的幂函数关系,计算出另一组致灾临界雨量。对比发现两组致灾临界雨量基本一致,24 h三个等级致灾临界雨量对应泗洪气象站167 a、17 a和2 a一遇降水量。统计模型能较好地捕捉短时降水与洪水涨幅的关系,Flood Area模型则更好地反映了长历时暴雨洪涝事件的雨洪关系,整合两种方法,可得到更科学的平原水网区暴雨洪涝致灾临界雨量。

Accompanying with developing economy, wet-net plains locating at the down basin of large rivers are becoming more vulnerable to rainstorm events than before. Combining DEM, terrain data and meteorology-hydrology characteristics of Jiangsu wet-net plain, we handled the mechanism of rainstorm flood through analyzing the meteorological and hydrological data. The water level of Laosui River changed with the quantity and intensity of the rainfall. Based on the correlation of water level increment and accumulative rainfalls, we modeled the linear relations during 2007-2011 and got a series of critical rainfalls. Then the heavy precipitation event occurred in 2010 was chosen to testify the statistical rainfalls.Inputting the hourly precipitation into Flood Area（FA）, the water level dynamics were grasped which correlated with the input precipitation powerfully. Compared with the statistical results,the thresholds calculated by the FA model showed smaller value at higher risk level and greater value at lower risk level, but the variability is almost less than 5%. The recurrence periods of critical rainfalls in 24 h are 167 a, 17 a and 2 a, respectively. With observation data, the relationships of precipitation-water level can be analyzed adequately, especially in short duration, while FA method has advantage in evaluating long duration correlation. The data from statistical method and FA simulation can be integrated and used as indices in flood disaster warning. Selection of proper indice is a key point of risk warning services.