青藏高原东部牧区大—暴雪过程雪灾灾情实时预评估方法的研究

利用青藏高原东部牧区26个县气象站30a的大－暴雪过程资料，以雪灾形成的主导因子－持续积雪日数作为灾情评估的等级标准，应用逐步回归方法，建立了冬半年持续积雪日数对大－暴雪过程的累积降水量、平均气温、最大积雪深度和最低气温降温4个因子的回归方程。经F检验，所有方程的回归效果达到十分显著成显著的水平，从而建立了大－暴雪过程雪灾灾情的灾时预评估方法。通过对1996－1999年资料的对比试用，表明这种预评估方法具有较高的精度和业务化的潜力，可作为雪灾情报服务的一种手段。

Study on real-time predictive assessment of snowstorm disaster in eastern pastoral area of Qinghai-Tibet Plateau

Based on the heavy snowstorm process data of 26 county stations in eastern pastoral area of Qinghai-Tibet Plateau during last 30 years, the sustained snow-cover days which is main factor of snow-disaster formation, is made as the rank standard of disaster condition. By use of stepwise regression method, a regression equation of sustained snow-cover days is established with four factors, i. e. accumulated precipitation of heavy snowstorm, average air temperature, maxiumum snow-cover depth and minimum air temperature lapse of the process. The regression efffect of the equation reaches so signifficant level via F test that it can form a snow-disaster real-time prediction assessment method of heavy snowstorm process. By means of probation from 1996 to 1999, the results show that the real-time prediction assessment method has a high assessment accuracy and operational potential, which can be made as a measure of snow-disaster information.