Simulation Of Runoff In Wannian Watershed Based On HEC-HMS And ELM Model

Aiming at the characteristics of less rainfall,fragile ecological environment and less surface runoff in the hilly area of Jinxi,Shanxi Province,the Nanchuan River,the first tributary of Sanchuan River in Luliang City,was selected as the object to study the characteristics of hydrological cycle and explore the applicable areas.Hydrological models and algorithms,and comparative analysis of runoff simulation.The research results can provide a basis for the rational utilization and management of local surface water resources,and provide reference for surface runoff simulation in similar areas.Based on the research status at home and abroad,this paper uses ENVI to interpret the land use information of the Wannian Basin in 1990,1995,2002 and 2009.From the overall trend,the forest land and water area are increasing,and the grassland is Decrease the trend,and the cultivated land will increase first and then decrease.The trend of rainfall?1956-2008?and runoff?1969-2008?in this basin was analyzed by linear trend method,run test method and M-K trend test.The results show that the rainfall and runoff in this basin have a downward trend.But not significant.The periodicity of rainfall and runoff in the basin was analyzed by spectrum analysis and wavelet method.The results show that the oscillations of the three periods with time scales of 6a,15a and 27a control the variation of precipitation in the whole time domain.The time scale is The oscillating period around 15a is the strongest,which is the first main period.In the 49a series,the average period of rainfall change is about10a,and it has experienced about 5 periods of abundance and dryness.The time scale of runoff period change is 6a,13a.The oscillation period around 22a and 22a is the strongest,which is the first main period.In the 40a series,the average period of runoff variation is about14a,and about three periods of abundance and dryness have been experienced.The M-K mutation method was used to analyze the abrupt change of rainfall and runoff in the basin.The time series of rainfall in the basin was 1959,1991,1994 and 2008.The time series of the runoff time series was 1972.1974,1975,and 1989.Based on the analysis of rainfall-runoff characteristics,the HEC-HMS distributed hydrological model of the study area was established.The production flow adopts the SCS curve number loss method,the slope surface convergence adopts the SCS unit line method,the base flow calculation uses the regression base flow method,and the river flood evolution adopts the Muskingum method.In the simulated 40a runoff,the qualified rates of abundant water year,flat water year and dry water year were 81.8%,75.0%,and 70.6%,respectively.The simulation results of the wet water year were the best,followed by the flat water year,and the rainfall was larger.The better the effect.It indicates that the HEC-HMS hydrological model can be applied to the surface runoff simulation in the region.The relationship between model parameters and runoff depth and maximum flow is analyzed.TheCN?RC?R?K value is positively correlated with runoff andx have negatively correlated with runoff.It is independent of runoff;CN?R?^tlag?K?x value is positively correlated with maximum flow,RCand negatively correlated with maximum flow.Relationship,regardless of maximum traffic.The order of sensitivity of the model parameters to the simulation results is:CN?29?R?29?K?29?x?29?RC?29?t_lag?29?Q.In order to further explore the simple and easy surface runoff model,the ELM algorithm is used to simulate the surface runoff of the study area 40a.The results show that in the simulated 40a runoff,the qualified rates of abundant water year,flat water year and dry water year are 81.8%,83.3%,and 82.4%,respectively.ELM model can also be used for surface runoff simulation in this area,simulation results and rainfall.The amount of relationship is not significant.The HEC-HMS model and the ELM model are compared and analyzed.Both have their own advantages and disadvantages.In practice,the two can be combined to achieve better simulation results.