Abstract: The change of annual maximum water level at Cuntan section in the Yangtze River was analyzed. It was found that an obvious decrease since 1970. Grey Markov Model was developed and used to analyze the change of annual maximum water level in the basin. And it was concluded that the predictions were coincident with the actual case. So the model can be spread and applied in the long-term hydrological predictions

Key words: grey model, Markov model, annual maximum water level, climate change, long-term hydrological predicting

摘要： 寸滩水文站是长江上游的主要控制站，分析可知，从2o世纪70年代开始，由于年降水量的减少和气温的下 降，年最高水位也呈逐渐下降的趋势。本文将灰色系统理论与马尔科夫链理论相结合，以194o年至2003年寸滩站 年最高水位序列为研究对象，建立灰色马尔科夫预测模型。从预测结果看出，所建灰色模型大体反映了年最高水位 的下降趋势，而经过马尔科夫模型对预测误差进行修正后，模型预测精度提高，计算值与实际值吻合良好。对其他 站年最高水位的模拟分析表明，模型精度能满足年最高水位预报要求。说明灰色马尔科夫模型可以应用于长期预 报中

关键词: 灰色模型, 马尔科夫模型, 年最高水位, 气候变化, 长期水文预报