云贵高原农业致灾过程雨量阈值估算模型研究  

A model to estimate consecutive rainfall thresholds for triggering agricultural disasters caused by floods and its application over Yunnan-Guizhou Plateau

作　　者：梁宏[1,2] 霍治国[3] 刘玲[3] LIANG Hong;HUO Zhiguo;LIU Ling(Meteorological Observation Center of China Meteorological Administration, Beijing 100081, China;State Key Laboratory of Disaster Weather, Chinese Academy of Meteorological Science, Beijing 100081, China;Chinese Academy of Meteorological Science, Beijing 100081, China)

机构地区：[1]中国气象局气象探测中心,北京100081 [2]中国气象科学研究院灾害天气国家重点实验室,北京100081 [3]中国气象科学研究院,北京100081

出　　处：《自然灾害学报》2019年第3期175-186,共12页Journal of Natural Disasters

基　　金：国家科技支撑计划(2012BAD20B00,2012BAD20B02);灾害天气国家重点实验室开放项目(2016LASW-B15);国家留学基金资助项目(20160533006)~~

摘　　要：基于广义Gamma分布函数首次构建了区域农业致灾过程雨量阈值估算模型,初步应用于云贵高原农业洪涝灾害特征研究。结果表明,广义Gamma分布能很好地拟合云贵高原站点过程雨量的统计分布特征,尤其对极端过程降水频率的拟合。贵州东南部和云南西部广义Gamma分布尺度参数约45~72,过程雨量的变化较剧烈,贵州中部和云南中部广义Gamma分布尺度参数约15,过程雨量的变化相对较小。云贵高原几乎每站均出现极端降水过程。基于模型估算的区域农业洪涝指标能较好反映农作物受灾率和成灾率的年际变化特征。在贵州模型计算的区域农业洪涝指标与农作物年受灾率和成灾率的相关系数分别为0.67和0.65,在云南两者分别为0.69和0.63。模型估计结果与实际农业洪涝灾害较吻合。农业轻涝、中涝和重涝对应的过程雨量阈值中值在贵州分别为55 mm、110 mm和225 mm,在云南分为60 mm、125 mm和250 mm。本文构建的农业致灾过程雨量阈值估算模型可为区域农业洪涝灾害风险管理提供新的技术手段。A model to estimate consecutive rainfall thresholds for triggering agricultural disasters caused by floods in regional scale (MERTAD) is developed based on the generalized Gamma probability distribution function (GGPDF). This model is firstly used to investigate the characteristics of agricultural disasters caused by floods over Yunnan-Guizhou Plateau. The results show that the GGPDF perform very well on fitting the processing rainfall distribution at site level, especially on describing the extreme processing rainfall distribution. The consecutive rainfall variations with a scale parameter value from 54 to 72 in southeast Guizhou province and west Yunnan province are much greater than that with a approximate scale parameter value of 15 in middle regions of Guizhou and Yunnan provinces. The distribution of consecutive rainfall has right heavy tail and extreme heavy rain event has taken place at almost every site. The modeled index of agricultural disasters caused by floods (IADF) reflect reasonably the annual variations of crops disaster rate and hazard rates over Yunnan-Guizhou Plateau. The correlation coefficients between IADF and disaster causing effect rate and disaster causing loss rates over Yunnan-Guizhou Plateau are 0.67 and 0.69 respectively, and those between IADF and those rates are 0.65 and 0.63 respectively. The consecutive rainfall thresholds for triggering agricultural disasters caused by floods over Yunnan-Guizhou Plateau are determined by the model. The results show that the mean consecutive rainfall threshold values for light, moderate and severe agricultural disasters caused by floods in Guizhou province are about 55 mm, 110 mm and 225 mm respectively, and those in Yunnan province are about 60 mm, 125 mm and 250 mm respectively. These processing rainfall threshold values could be considered as the useful reference indices to manage the agricultural flood risks. The MERTAD model could be used as a new method to manage the agricultural disasters caused by floods in regional scale.

关 键 词：农业洪涝 过程雨量 广义 GAMMA分布 模型 云贵高原 

分 类 号：S166[农业科学—农业气象学] X43[农业科学—农业基础科学]