Prediction Of Surface Runoff In Typical Areas Of Mountainous Cities Based On SCS-CN Model

According to the local climate characteristics and runoff data,the runoff curve number model proposed by the Soil Conservation Service has the advantages of simple calculation,less required parameters,and high simulation accuracy,and is widely used in runoff prediction at home and abroad.However,due to geographical differences,when using the SCS-CN model in mountainous urban areas with large terrain differences,the CN value calculation proposed by the Soil Conservation Service directly differs from the actual situation.It is necessary to correct the model parameters according to the actual situation of the study area.This study uses the experimental method to simulate the rainfall runoff in the typical area of the main urban area of Chongqing,and compares the CN value determination method based on the measured rainfall runoff data.The model parameters are corrected by the model parameter evaluation,the correction slope factor,the initial loss rate?optimization,etc.,so as to adapt to the actual runoff situation in the mountainous area of Chongqing,and finally select the 37 typical rainfall runoff data pair parameters of a runoff field in the Yubei area.The application of the revised SCS-CN model is verified and evaluated.The main conclusions are as follows:(1)Combined with the model evaluation parameters,the runoff prediction accuracy of different CN value calculation methods is comprehensively evaluated.It is found that the arithmetic mean method has the best runoff prediction effect for Z1,H1 and H2plots,and the corresponding E is 0.87,0.91,0.79,RE is less than 20%of the license error.The median method has better prediction accuracy for Z2 plots,and the model evaluation parameters E and RE are 0.88 and 14.82%,respectively.Therefore,it can be preliminarily determined that the CN values of the Z1,Z2,H1,and H2 plots under normal humidity conditions(AMCII)are 72.46,73.76,79.74,and 82.92,respectively.(2)Considering the influence of slope change on runoff prediction,the slope correction factor is added on the basis of the traditional SCS-CN model.By comparing the prediction results of the runoff before and after the slope correction,it is found that the traditional SCS-CN model and Williams slope formula are not effective in runoff prediction,and the corresponding E is lower,and the relative error RE exceeds 20%of the allowable error;After being corrected by the Huang slope formula,the E of each block reaches 0.85 or more,and the relative error RE is within the allowable range of20%,which satisfies the requirements of the model prediction accuracy after the slope correction.(3)By analyzing the sensitivity of the initial loss rate?to the prediction accuracy of the model,it is proved that further optimization of the initial loss rate?on the basis of the slope correction can improve the reliability of runoff prediction.Through sensitivity analysis,it is found that the calculated surface runoff Q of the Z1,Z2,H1and H2 plots is increased by 53mm,49mm,36mm and 29mm respectively,which is mainly affected by the change of rainfall P,and the influence of?can be neglected,Therefore,the prediction accuracy of small rainfall fields can be further improved by optimizing the initial loss rate?of each block.The initial loss rates?after Z1,Z2,H1 and H2 optimization are 0.24,0.22,0.15,and 0.16,respectively.The E after the initial loss rate?is optimized to 0.91,0.89,0.95,and 0.93,which is higher than the slope corrected E.5.81%,1.14%,6.74%,9.41%,and E were 9.62%,10.01%,6.53%,and 5.09%,respectively,which were further reduced after the slope correction.(4)The measured runoff data of 37 representative rainfalls in a runoff test field in Yubei District were selected to verify the runoff simulation of the standard SCS-CN model and the modified model.The traditional SCS-CN model R~2=0.968,E=0.34,~—RE=120.34%,slope correction and initial loss rate?optimized SCS-CN model R~2=0.902,E=0.83,~—RE=24.40%,considered SCS-CN model after slope correction and initial loss rate?optimization It has a good effect on the runoff prediction of typical areas in southwestern mountains.