马尔可夫概型分析在春季10mm降水预测中的应用

根据锦州地区1960～2007年4月下旬～5月上旬日降水资料,应用马尔可夫概型分析方法,建立春季10mm以上有利春播降水预测模型,通过对4月下旬、5月上旬日降水资料将降水分为有和无2种状态,计算各状态转移概率并对未来时刻的降水状态进行预测。结果表明,从1997～2008年预测效果检验来看,其正确率达58%。马尔可夫概型分析应用在春季10mm降水预测中是一种简单而有效的方法。

Application of Markov Probability Model Analysis in the Prediction of 10 mm Precipitation in Spring

Based on the daily precipitation data in Jinzhou area from late April to early May during 1960-2007,using Markov probability model analysis method,the prediction model of over 10 mm precipitation in spring favorable for spring sowing was established.According to the precipitation data in late April and early May,the precipitation state was divided into two type.i.e."Yes" and "No".The transition probability of each state was calculated and the precipitation state of the future time was predicted.The results showed that the accuracy rate of precipitation prediction during 1997-2008 reached 58%.Applying of Markov probability model analysis in the prediction of 10 mm precipitation in spring is a simple and effective method.