CAJ | 학술논문

依据高关水库流域1971～2011年主汛期实测降雨量资料，应用均值标准差法建立5级分级标准。针对降雨量为相依随机变量的特点，以各阶自相关系数为权重，运用马尔可夫链模型预测未来一年的时段降雨量状态，由于采用此方法得到的是一个区间值而不是一个确定的数值，因此，不但扩大了预测的范围，最重要的是能够很好的指导现实的工作。该方法具有直观、预测准确、计算简便的优点，为区域降雨量的中短期预测提供了新的分析途径。
의거고관수고류역1971～2011년주신기실측강우량자료，응용균치표준차법건립5급분급표준。침대강우량위상의수궤변량적특점，이각계자상관계수위권중，운용마이가부련모형예측미래일년적시단강우량상태，유우채용차방법득도적시일개구간치이불시일개학정적수치，인차，불단확대료예측적범위，최중요적시능구흔호적지도현실적공작。해방법구유직관、예측준학、계산간편적우점，위구역강우량적중단기예측제공료신적분석도경。
On the basis of the rainfall precipitation data of Gaoguan reservoir basin in main flood season from 1971 to 2011, the mean standard deviation method is used to set up 5-grade classification standard. According to the characteristics of rainfall for dependent random variables, with the weight of each order autocorrelation coefficient, the Markov chain model is used to forecast the rainfall condition next year. As the result of this method is an interval value rather than a determined value, therefore, it not only expands the range, the most important it can guide the work of the reality very well. The method is intuitive, its prediction is accurate, and its calculation is simple, so it provides a new method of the middle or short term prediction for regional rainfall.