基于灾损评估的北京地区冰雹灾害风险区划

The Hail Risk Zoning in Beijing Integrated with the Result of Its Loss Assessment

北京地区冰雹灾害风险区划主要依据"风险=概率×损失"的风险评估方程来计算风险值,不单纯依靠极端天气及气候事件的概率统计来做区划。首先利用灰色关联模型将北京地区近30年(1980—2009年)的冰雹灾害历史灾情资料进行灾损评估,评估结果以关联度归一化值及高斯变换值的方式分别用于时间序列分析及风险值估算。然后采用启发式分割算法(BG算法)对灾情数值进行时间序列分析,分析发现:1997年为不平稳结点,1997年前后的灾情数值从平稳性上应分属两个不同的时间序列。以1997年后的期望偏差作为变异系数。在灾损评估结果的基础上进行关联度归一化值在0～1范围内10个等间距的雹灾频次统计,依据统计结果计算北京地区冰雹灾害的平均风险值并用该值来制定冰雹风险等级标准。最后利用风险评估方程计算历次雹灾的风险值,结合风险等级标准完成北京地区冰雹灾害风险区划。区划结果显示;北京地区冰雹灾害的高风险区域主要分布在北京城市中心地区、密云县城及平谷区等几个人口稠密地区,而山区及山前迎风坡地带尽管降雹频次高,冰雹灾害风险却相对较少。

In order to recognize the meteorological disaster risk and assess it, the task of risk zoning is started exploring the possibility and severity influenced by the disaster, and the distribution of risk area zone is determined. The study is very important for the disaster avoidance or prevention, and is widely utilized in disaster emergency and meteorological service of supporting decision regulation. Normally the risk zoning depends upon the probability of extreme weather or climate events, such as the occurrence or probability of extreme meteorological elements. However the risk of disaster focuses on the intensity or loss of the disaster besides the probability. Moreover, the index of hail risk zoning in Beijing is carried out based on the formula that risk equals probability multiplied by loss, and it can avoid the solely doing statistics on the probability of extreme weather and climate events. The loss assessment is firstly fulfilled by the Gray-correlation Model with the data of disaster occurred in Beijing during recent 30 years. Meanwhile, the BG al-gorithm is introduced here for the time-series analysis on the disaster data, and it is found that the year of 1997 in the time-series is an inconsistent standpoint, which means that the data should be separated into two series for its inconsistence. Based on this judgment, the expectations of the bias from 1997 are consid ered to be the variation modulus. On the other hand, with the result of loss assessment, the hail occurring frequency is calculated out by 10 groups divided by average intervals form 0 to 1 of the loss index value. It can be utilized in deducing normalized hail risk index of Beijing, with which the risk zoning standards can be determined. Eventually the risk index is calculated with the risk function, and the indexes are also summed up by groups. Furthermore, the risk zonings are symbolized upon the index with its zoning standards. From the result of risk zoning, it can be concluded that the relatively higher hail risk zone mainly lo- cate in urban area of Beijing, such as the centers of Miyun and Pinggu, while the risks of the mountain area and its piedmont, are relatively lower.