中文摘要：

水文组合预报方法是对多种预报模型的预报结果进行组合分析的一种预报方法．针对历史洪水数据不丰富的流域难于利用多种水文组合预报模型进行洪水预报的现状，构建了以贝叶斯分析为基础，同时结合专家经验、马尔科夫蒙特卡洛模拟、Gibbs抽样法，并引入实时校正的组合预报模型．以嫩江流域为实例，对基于贝叶斯分析的组合预报模型的精度进行了验证．验证结果表明，该模型可行而且实用，预报精度明显高于单个模型的预报精度．

英文摘要：

Hydrological combined forecasting method can offer synthetical analysis to prediction results received from several separately forecasting models. Remarkablely, as it showed the impression of weakness when facing absence of historical flood data for some drainage areas. A new combined forecasting in this paper was introduced, which was on the basis of Bayesian analysis, Experts ＇ experience, MCMC, Gibbs sample and real-time emendation function. Finally,the precision of the Bayesian combined analysis model was validated by taking the drainage area of Nen River for instance. The result indicates the established model was feasible and practical, and its precision overwhelms any single model＇ s precision.