气候变暖下太湖极端洪水的归因探讨

Probe on attributions of extreme floods responding to the climate changes in Lake Taihu catchment

全球增温引起的降水变化是否引起极端洪水的增加,发生在不同气候背景的极端洪水事件可提供不同参照系;而不同驱动因子下气候、水文数值模拟为认识洪水发生和归因提供了有效途径.本文结合机理数值模拟和随机统计模拟两种途径,针对1990s和1880s的太湖流域特大洪水,通过GCM气候模拟驱动的流域水文模拟和不确定性的阈值模拟,分析19世纪末和20世纪末极端洪水的发生强度和频率的变化,从而论证极端洪水发生的风险系数.结果表明,1990s的极端洪水流量(0.1%的极端洪水流量(Q0.1%)为2929～3601 m3/s,0.5%的极端洪水流量(Q0.5%)为1842～1893 m3/s)比工业革命前大气温室气体状况下(Q0.1%为2069～3119 m3/s,Q0.5%为1436～1561 m3/s)显著增大.与19世纪末相比,由于太湖流域人类活动改变的流域下垫面在1999年特大洪水中引起最大增量占35%,本文模拟和分析的20世纪末气候下的洪水最大增量占60%.去除人类活动影响的下垫面变化,估计特大洪水风险的最大增量为25%,因此认为20世纪末气候变化引起的太湖极端洪水风险在增加;这将为认识与全球增温相关联的洪水灾害预测预警提供科学依据.

Extreme floods under different climate systems, the pre-industrial time, and the end of the 20th century can provide good comparisons to examine differences among different mechanisms, while numerical models of the climate and hydrology could provide effective ways to simulate processes of the extreme flood changes and analyze the attributions. The paper applied the numerical modeling and statistical analyses to simulate extreme floods of the 1990s and the 1880s in Lake Taihu, driving by different scenari- os-climate simulations from four GCMs, and analyzed the flood frequency differences between the end of the 19th and 20th centu- ries. Results show that extreme flood discharges in 1990s （Q0.1% 2929-3601 m3/s and Q0.5% 1842-1893 m3/s） have exceed statistically those during pre-industrial time （ Q0.1%, 2069 -3119 m3/s and Q0.5%, 1436 -1561 m3/s）. Comparing these floods with those at the end of the 19th century, the highest extreme-flood risks produced by land surface changes due to human activities were 35% increases, and the total highest risks were 60% increases in the end of the 20th century. Thus the highest extreme-flood risk would increase 25% risks attributed from the 20th century climate changes due to the increase of greenhouse gases ; this could pro- vide scientific basis for understanding and predicting the flood disasters responding to the global warming.