成都经济区极端降水广义帕累托分布模型研究

Study on Generalized Pareto Distribution Model of Annual Extreme Precipitation in Chengdu Economic Zone

为研究成都经济区极端降水的概率分布特征,利用成都经济区(绵阳、德阳、遂宁、成都、资阳、雅安、眉山、乐山)8个站点1960-2018年的逐日降水量资料,选取第99个百分位值为极端降水量的國值,采用广义帕累托分布模型(GPD)对成都经济区极端降水事件进行拟合并计算其重现期降水量。将极端降水的理论累积概率与实测概率进行了对比,发现成都经济区8个测站极端降水事件符合GPD分布。研究结果表明,成都经济区极端降水在不同重现期内空间分布特征基本-致,只是数值有所不同;具体表现为乐山、雅安和绵阳极端降水量最大,成都、德阳、眉山和遂宁次之,资阳市最小,总体呈现向东南逐渐减小的趋势。

To study the spatial distribution characteristie of the probability of extreme precipitation in the Chengdu Eco-nomic Zone(CEZ),we analyze the daily precipitation data from 1960 to 2018 which come from eight stations(Mian-yang,Deyang,Suining,Chengdu,Ziyang,Ya'an,Meisharl and Leshan)in the CEZ.The 99th percentile value is de-fined as the extreme precipitation threshold and we choose the generalized Pareto distribution(GPD)model to fit ex-treme precipitation events and calculate precipitation return period in the CEZ.Comparing the theoretical cumulative probability of extreme precipitation with the measured probability,it is found that the extreme precipitation events from eight stations conform to the GPD.The results indicate it's much the same for the spatial distribution characteristics of extreme precipitation in different return periods in the CEZ,only the values of precipitation are different.Leshan,Ya'an and Mianyang have the highest extreme precipitation,followed by Chengdu,Deyang,Meishan and Suining,Ziy-ang is the smallest one.On the whole,it shows a gradual decrease precipitation pattern from other directions to the southeast in the CEZ.