Uncertainty Study Of EFDC Model In Hydrodynamic Simulation Of Poyang Lake

The seasonal differences of the hydrodynamic variation characteristics of Poyang Lake are very significant,For better study of the hydrodynamic variation characteristics of Poyang Lake,EFDC model has received wide attention in recent years.However,most of the current studies on the uncertainty of the hydrodynamic model of Poyang Lake based on EFDC model focus on the problems of input data and parameters,and less attention is paid to the model modeling approach and other issues.To research the influence of several different modeling methods on the uncertainty of simulation results of 2D hydrodynamic model of Poyang Lake,EFDC model is used as the modeling base,several different resolution orthogonal grids,different resolution advective orthogonal grids are constructed,and the roughness of lake bottom and backflow problems are modeled,using the water level and flow data of Poyang Lake from June 2011 to December 2012 for simulation.Quantitative analysis of the effect of different modeling approaches on the model uncertainty of the results.The results of the study showed that:(1)The orthogonal grid resolution is directly proportional to the accuracy of the model output,the higher the grid resolution,the higher the accuracy of the model output,the lower the grid resolution,the lower the credibility of the model results;the orthogonal grid resolution is inversely proportional to the model efficiency,the higher the grid resolution,the longer the model operation time and the more computer resources are occupied.(2)The simulation results of orthogonal advective grid and orthogonal grid are better than orthogonal grid under similar resolution,and the high resolution of grid is also proportional to the accuracy of simulation results.However,the computational efficiency of the advective grid model is lower than that of the orthogonal grid at similar resolution scales.(3)The effect of roughness on the hydrodynamic characteristics of the lake is different in different periods of the lake.The feedback of the simulation results for the different values of lake roughness taken in the high water level case is not obvious,but the influence of roughness starts to be prominent when entering the low water level period,and the closer the lake bottom roughness is to the real lake bottom condition,the closer the simulation results are to the measured water level.(4)The feedback of the backflow problem on the results is more obvious with the high-resolution grid model.The results of the low-resolution grid also fluctuate differently in the simulation due to the appearance of the backflow problem,but this effect is tolerated because of the low grid resolution,so it is not significant in the results.