Study On Flood Forecasting Methods In The Areas With Shortage Data In Eastern Liaoning

The special geographical location,landform and unique climate conditions determine that Liaoning Province is a province with frequent floods and floods.In recent years,,The flooding of small and medium-sized rivers is becoming more and more serious with the increase of strong typhoons and heavy precipitation events.However,due to more new hydrological stations in small and medium-sized rivers,there are no long series of observations.Flood forecasting faces difficulties in modeling such as data shortage.Combining hydrological model,remote sensing technology and other methods,retrieval of parameters by searching for hydrologic similarities between basins is a commonly used method for solving hydrological simulations in areas where there is a shortage of data.This article takes 7 medium and small river basins in eastern Liaoning as the research object,first,study the uncertainty of Xinanjiang model parameters.Second,using Arc GIS to extract the characteristics of the research basin,based on principal component analysis method to reduce the dimension of watershed's attribute index.Third,using the main and objective weighting methods to calculate the similarity of the river basin,determine the reference watershed.Last parameter migration,at the same time,the paper analyzes and studies the correction method for the regression coefficient CS of the river network in the Xin'anjiang model.The main research contents and papers of the thesis are as follows:Firstly,using ArcGIS tools to extract topographic features and surface coverage characteristics of the research basin based on DEM data and remote sensing data,which provide data support for the quantification of watershed similarity.Including watershed area,watershed length,form factor,mean river basin elevation,mean river basin slope,main river length,river ratio drop,total river network length,river network density,average height difference,forest coverage,cultivated land coverage,grass coverage.Secondly,based on the multi-objective GLUE method,the uncertainties of Xinanjiang model parameters were analyzed,and the sensitivity of each parameter was obtained.The posterior probability distribution range of each parameter was used as the initial range of parameter determination,and the parameter rates were calculated for 7 research basins.It has improved parameter optimization efficiency and laid the foundation for parameter migration.Thirdly,firstly,four principal components are extracted as the indicators of similarity evaluation based on principal component analysis;then the main and objective weighting methods weight the principal components.The similarity calculation results of the two methods are similar,which verifies that the principal components can be represented more comprehensively.The information of the original data shows that the eastern region of Liaoning Province can directly use the method of objectively empowering to evaluate the similarity of the river basin;finally,the parameters of transplanted target basins and reference watersheds are determined.Fourthly,direct transplantation of parameters found that the effect of transplantation of production parameters was acceptable,but the simulation effect of flood peaks was generally poor.Combined with the characteristics of medium and small-scale watershed forecasting,the revised method of the coefficient of extinction of river network in the Xin'anjiang model was studied and generalized.The linear relationship between the network regression coefficient CS and the watershed form factor and the average slope of the river basin shows that the simulation results of the modified flood peak are satisfactory,providing a reference for the study of flood forecasting in the data-free areas.