摘要
近年来,干旱、洪涝、雪灾等极端气象灾害频发,强烈冲击农业生产和粮食安全,科学估算农业自然灾害给粮食产出带来的风险价值,对预警农业自然灾害、确保粮食安全具有重要意义。文章针对农业自然灾害大灾损失的低频高损、数据稀少的特点,采用极值理论中的POT模型对历史观测值超阀值数据进行建模,运用GPD模型对农业自然大灾损失进行拟合,模拟计算农业自然灾害VAR值,为农业自然灾害预警和粮食安全储备提供科学依据。
In recent years, such extreme weather as the drought, floodwater, blizzard and so on occurs frequently, which bring a strong impact on agrieultural production and grain reserve security. It, s important for early warning natural calamity and ensuring grain reserve to estimate the risk value of natural disasters scientifically. Because agricultural calamity of extreme weather is high loss and low frequency, the article adapts the model of Peaks Over Threshold and Generalized Pareto Distribution, according the theory of extreme value, to simulate the loss due to natural calamity, which can be an important basis for natural disasters early warning and safe grain reserve.
出处
《统计研究》
CSSCI
北大核心
2011年第7期79-83,共5页
Statistical Research
基金
国家社会科学基金项目"农业发展与组织结构创新研究"(10XJY032)资助
关键词
POT—GPD模型
农业自然灾害
风险价值
Peaks Over Threshold & Generalized Pareto Distribution
Agricultural Calamity
VAR