典型黑土区坡耕地次产流径流曲线数计算方法

A method for calculating curve number of each runoff event on a sloping cropland in the Typical Black Soil Region

径流曲线数模型原理清晰简单、对参数和输入变量要求低,在地表径流预测中已得到广泛应用。为了提高该模型预报典型黑土区地表径流的精度,利用黑龙江省嫩江市鹤北小流域7个坡面径流小区共138场降雨-产流资料,通过量化次降雨过程中雨量在时间上集中分布程度,分析其对地表产流影响,提出次产流径流曲线数CNt的计算方法。CNt/CN与降雨过程中最大时段Xmin内的降雨量和次雨量比值(PX/P)之间呈极显著正相关;当10≤X≤40时,两者相关程度更高。(P10/P)是对本区CNt影响最显著的降雨过程特征因子。可通过建立(CNt/CN)与(P10/P)之间的函数关系计算CNt,并提出该方程在不同耕作方式下的参数取值。使用径流曲线数模型时,若采用本文提出的计算方法,将CNt作为模型参数计算地表径流量,可显著提高模型在本区的预报精度。

[Background] Accurate prediction of surface runoff is crucial for the simulation of soil erosion and non-point source pollution. The soil conservation service curve number( SCS-CN) model has a simple structure and fewer required parameters,and is thus widely used to predict surface runoff. In various storm events,even for the unchanged land use and management,the parameter-curve number( CN) varies over a wide range,and the lack of understanding of rainfall processes and their influence on runoff may cause the simulated deviations. Moreover,the application of SCS-CN model for runoff prediction of the Typical Black Soil Region in China is limited. [Methods] Therefore,our study intended to optimize the SCS-CN model simulation for the Typical Black Soil Region by proposing a SCS-CN calculation method for each runoff event( CNt) based on observation data for 138 rainfall and runoff events from 7 plots in the Hebei watershed. And the rainfall intensity degree over time during the rainfall process was applied to calculated CNt. Nash-Sutcliffe efficiency( Ef),correlation coefficient( R),and mean relative error( Re) were used in the examination of simulation results that were calculated by CNt. [Results]The ratio for the maximum to minimum CNtvalue for different plots varied from 1. 38-2. 24,with an average of 1. 81. This study elucidated a significant positive correlation between the ratio of CNtand the average SCS-CN( CN),as well as the ratio of the maximum X-minute rainfall amount( PX) to the total rainfall amount for each rainfall event( P). The ratio of curve number for each rainfall event to the annual mean curve number( CNt/CN) significantly increased with the ratio of maximum X-minute rainfall to total rainfall( PX/P) for the events. When 10 ≤ X ≤40,the correlation coefficient between( CNt/CN) and( PX/P) was higher,and( P10/P) was the most significant factor which represented the character of rainfall process that influenced the variation of CNt. The equation for calculating CNtwas established based on the logarithmic function between( CNt/CN) and( P10/P).Moreover,the present study provided the parameters for cropland and bare land. For cropland with longitudinal ridge tillage and bare land,rills were easily formed on exposed surfaces during rainfall events,which facilitated the formation of a more stable convection path. For the calculated runoff depth by SCS-CN model with the parameter of CNt,Efwas 0. 47,R was 0. 77,and Rewas-18. 9%. The Ef for the SCS-CN model with the parameter of CN was only-1. 63. [Conclusions]Adoption of a SCS-CN calculation method for each runoff event( CNt) may improve the prediction performance for SCS-CN model in the Typical Black Soil Region in China.