使用积雪遥感面积数据改善山区春季融雪径流模拟精度

Using snow remote sensing data to improve the simulation accuracy of spring snowmelt runoff: take Babao River basin as an example

随着寒区水文模拟研究的深入,春季融雪径流模拟误差大这一问题越来越明显.针对此问题,以八宝河流域为研究区,利用积雪衰减曲线将MODIS积雪面积比例产品转化为雪水当量,并用其更新分布式水文模型GBHM(Geomorphology-Based Hydrological Model)中模拟的雪水当量,达到提高春季融雪径流模拟精度的目的.利用GBHM模型对八宝河流域2005-2007年进行了模型预热,参数率定,同时选择2008年进行模型检验.结果表明:GBHM模型在八宝河流域有较好的径流模拟精度,年均纳什效率系数(NSE)达到0.64.但模型对春季融雪过程的模拟效果较差,通过引入积雪遥感数据,这一问题得到很大改善.加入积雪遥感数据后,2008年3-6月径流模拟精度NSE和相对偏差Bias分别由-1.0、-0.45改进为0.58、0.06,单点雪水当量模拟精度获得提高,流域水量平衡也更加合理.

With the development of hydrologic simulation in cold regions,the large simulation error of spring snowmelt runoff has become one of the largest uncertainties.In this study,Babao River basin was selected as the study area which is regarded as the representative alpine catchment and GBHM（Geomorphology Based Hydrological Model） was established to simulation the watershed hydrological process.Then,the maneuverability of the GBHM was discussed and analyzed.On this basis,the snow cover extent（SCE） product with cloud reducing which retrieved from the Moderate Resolution Imaging Spectroradiometer（MODIS） was used to update GBHM snow water equivalent（SWE） for a more accurate description of the spring snowmelt runoff process.A simple snow depletion curve model was used for the necessary SWE-SCE inversion.And,the model＇s calibration period（2005-2007） and validation period（2008） was chosen.The results show that：the model can basically reflect the trend of the runoff and achieve good simulation accuracy with the average annual Nash-Sutcliffe coefficient（NSE） reached 0.64.However,the model doesn＇t work well during spring snowmelt process.After applying snow remote sensing data into GBHM,this situation has been improved.The NSE and Bias of spring snowmelt runoff respectively from —1.0,-0.45 become 0.58,0.06,the SWE of grid gets more accurate,and water balance is more reasonable after adding remote sensing data between March and June 2008.