Application of similarity analysis to flood forecasting

摘要：

　　Objective There are plenty of useful information in hydrological observations.Predicting future flood on the basis of similarity information in historical records is an effective and promising approach.Method In this paper,a multi-measure similarity analysis method of rainstorms is developed based on “quantity,“type similarity indicators,the earth movers distance(EMD)and the rainstorm distribution similarity indicator.Search the similar rainstorm and its corresponding typical flood in historical library and then scale the typical flood process according to the ratio of rainfall amounts to achieve flood forecasting.Result The method is applied to a case study in Xinmiao station of Kuye River.The results show that with the accelerating information of rainstorm and flood process,the forecasted flood process is updated continuously,and the prediction accuracy is gradually increasing.Conclusion The proposed similarity analysis method is effective and applicable to flood forecasting.

作者: Xiao Zhangling Liang Zhongmin Li Binquan Zhou Yan Wang Yanlan

作者单位: College of Hydrology and Water Resources,Hohai University,Nanjing 210098,China College of Hydrology and Water Resources,Hohai University,Nanjing 210098,China;State Key Laboratory

会议类型: 国际会议

会议名称: 2018 International Symposium on Water System Operations(ISWRSO 2018)(2018年水资源系统及调度国际研讨会)

会议地点: 北京

会议语种:英文

页码: 1-6

在线出版日期: 2018-10-12（万方平台首次上网日期，不代表论文的发表时间）