在分析北疆牧区畜牧业生产与雪灾分布特征的基础上,利用RS、GIS和地面监测资料,从草地抗灾力、家畜承灾体和积雪致灾力3个子系统中选择9个因素作为预警参评因子,用家畜死亡率作为风险评估因子,构建了一个在完全放牧状态下的牧区雪灾预警与风险评估体系和模式.采用多层次综合法和目标线性加权函数法,建立了雪灾预警分级模型、判别模型和风险评估模型,据此可对草地和家畜抵御雪灾的强度和雪灾的风险期望损失进行动态预警与评估.北疆牧区雪灾预警与风险评估模型的精度分别达85%和72%.
Based on the analysis of animal husbandry production and of distribution characteristics of snow disaster in northern Xinjiang, and by using RS and GIS techniques and field survey data, 9 early warning factors were selected from the three subsystems of grassland' s disaster-resistant capability, livestock' s disaster-bearing capacity, and disaster-causing potential, and the death rate of livestock caused by snow disaster was used as a factor of risk assessment. An index system of snow disaster's early warning and risk assessment for completely grazing grassland was established, and the early warning model of snow disaster, its distinguishing model, and risk assessment model were built by using multi-hierarchical synthetic and multi-objective linear weight function methods to predict the resistant capability of grassland and livestock against snow disaster, and to assess the potential risk loss from snow disaster in northern Xinjiang. The accuracy of the early warning model and risk assessment model was 85% and 72%, respectively.