基于SCS模型的新疆博尔塔拉河流域径流模拟

Runoff simulation based on SCS mode in Bortala River Basin in Xinjiang

为探索SCS(soil conservation service)模型在干旱区地表径流模拟中的适用性,以新疆博尔塔拉河上游温泉流域为例,利用2013年融雪期18场次降雨量、日径流量及卫星同步观测基础数据,基于度日模型、土壤水分吸收平衡原理及地表温度-植被指数特征空间反演土壤水分等方法,探索适用于流域尺度耦合降雨、积雪融水混合补给径流的SCS模型参数改进算法,采用国产"高分一号"16m分辨率遥感影像和landsat8 OLI 30m多光谱遥影像为模型提供面状数据。分别利用参数算法改进后SCS模型与原SCS模型进行径流模拟,前者模拟值较后者更接近实测值,二者模型验证期Nash效率系数分别为0.66和0.38,相对误差系数分别17%,27%;研究结果表明,利用遥感反演地表参数,结合土壤饱和含水量计算SCS模型参数的方法在面积大且数据缺乏的温泉流域具有可行性和实用性。

Rainfall and snowmelt are important ways to produce runoff in Boertala River. The characteristics of arid basin are large area and snowmelt in the surrounding mountains. But the water resource is the most important factor for socio-economic development especially in arid regions. Calculating the total water resources can provide reference for government in formulating strategies. The soil conservation service curve number (SCS-CN) model developed by National Resources Conversion Service (NRCS), Department of Agriculture, the United States of America is the most popular and widely applied for direct runoff estimation. This method is modified by accounting for the static portion of infiltration and the antecedent moisture. It has passed through a great deal of discussion among scientists and hydrologists. The model is based on the water balance equation, in which the curve number is derived from the tables given in the National Engineering Handbook for catchment characteristics, such as soil type, land use, hydrologic condition, and initial soil moisture condition. The variability of these components can cause the instability of model parameters. The method provides the relationship among rainfall depth, retention parameter, curve number, and initial abstraction. In practice, the variability of retention parameter and curve number is observed, as a result of seasonal and storm morphology changes. To describe this variability and assess the errors in the estimation of curve number, many statistical studies were performed. The aim of the paper was to assess the variability of retention parameter and contribute to a better understanding and estimation of retention parameter for its practical application in water resource management. Applications of the original SCS-CN model for runoff estimation without the calibration of retention parameter value tended to give inaccurate results in arid area. The existing universal rainfall-runoff model is widely used in humid areas of small watershed; but for arid and semi-arid large watershed, we need find the good method to improve the model. In this paper, we focus on Wenquan watershed in source regions of Boertala River, and analyze the change trend about the temperature, rainfall and runoff in nearly 50 years. This paper used the 16 m resolution remote sensing image from high-definition earth observation satellite “Gaofen-1”, which provided the land use classification data for original SCS-CN model to determine the curve number. This paper used Landsat8 OLI 30 m resolution remote sensing image to inverse the land surface temperature and vegetation index. And the surface temperature/vegetation index (TS/VI) was used to inverse the soil moisture. Use the principle of soil moisture absorption balance to estimate the model’s retention parameter. It is the innovation in this paper. Use the degree-day model to calculate the equivalent water of snowmelt.Precipitation may need to be revised as the sum of snowmelt and rainfall. At last we used both original model and modified model to simulate the runoff of the 18 rainfall events from March to October in 2013 so as to explore the suitable modified SCS-CN model to simulate the runoff for the large basin with the mixed supply of rainfall and snowmelt. Results showed that the improved model’s simulation value was close to the original model very well. In the validation periods, their correlation coefficients were 0.66 and 038, respectively. And relative errors were 17% and 27%, respectively. It further demonstrated that we can use remote sensing to inverse the land surface parameters for SCS-CN model, and use modified algorithm of SCS-CN model parameters to improve the accuracy of the simulation results. Improved model’s simulation results show that it has feasibility and practicability for Wenquan watershed. From the improved algorithm of SCS-CN model parameters, it is evident that there is still some room for the original SCS-CN model to be modified and replaced by other relationships for more reliable runoff estimation.