东北三省农作物洪涝时空风险评估  

Spatial-temporal Risk Assessment of Crops Caused by Flood in the Three Northeastern Provinces of China

作　　者：赵思健[1,2] 张峭[1] 

机构地区：[1]中国农业科学院农业信息研究所,农业部智能化农业预警技术重点开放实验室,北京100081 [2]北京师范大学民政部/教育部减灾与应急管理研究院,北京100875

出　　处：《灾害学》2013年第3期54-60,共7页Journal of Catastrophology

基　　金：国家自然科学基金项目“自然灾害风险时空差异的多维表达及其评估技术研究”(40901274);中央级公益性科研院所基本科研业务费专项资金课题(中国农业科学院农业信息研究所)“县级尺度下的农业生产灾害风险预警技术研究-以东北三省农作物洪涝风险为例”(2013-J-014)

摘　　要：仅考虑空间差异的灾害风险评估已经满足不了灾害风险管理向精细化方向发展的要求,增加时间维度的时空风险评估分析有利于增强风险评估结果,对提升风险管理的准确性与针对性起着重要的作用。以东北三省为研究区,在气象数据、地形数据、农作物灾情数据和种植面积数据的支撑下,以县和月为单位的时空两个维度开展农作物洪涝时空风险评估研究。在方法上,采用反距离权重法(IDW),利用与县行政区最临近的3个气象站点的日降雨数据插值出县级行政区的日降雨数据;利用二元回归建立农作物洪涝受灾率与过程降雨量、县平均高程之间的农作物洪涝脆弱性函数;分县分月提取过程降雨量,构建非参数核密度的信息扩散模型拟合降雨量的概率分布;综合概率分布与脆弱性函数,计算出分县分月的农作物洪涝条件期望受灾率,实现风险时空差异表达。最后,制作出东北三省县级尺度下4至9月的农作物洪涝风险差异图,并对风险时空差异规律进行分析。Nowadays, risk assessment of natural disasters just concerning on spatial variance of risk can not meet the requirement that risk management of natural disasters would develop in higher accuracy. Risk variance as- sessment in both space and time dimensions could help to enhance the results of risk assessment, and play an im- portant role in risk management. The spatial-temporal risk assessment of crops caused by flood using county as space unit and month as time unit is performed in the three northeastern provinces of China under the support of metrological data, terrain data, flood record data and the planting area data. With inverse distance weighting （IDW） method, daily rainfall data for each county is interpolated by using daily rainfall data of its adjacent 3 mete- orological stations; with bivariate regression, vulnerability function of crop caused by flood among variables of crops＇ disaster affected area, rainfall process volume and terrain is built; by using the information diffusion model of nonparametric kernel density, the probability density function of total rainfall volume in a storm for each month of each county is fitted; The conditioned expected disaster-affected area of crops as risk measure for each month of each county is calculated on the combination of probability density function and vulnerability function; A serial of risk maps on county-level and month-level scales are produced to be used to recognize the rules of risk variance in both space and time.

关 键 词：农作物 洪涝 时空风险 评估 脆弱性 信息扩散 东北三省 

分 类 号：X43[环境科学与工程—灾害防治]