多源遥感驱动的SRM模型在缺资料地区的研究及应用

以MODIS雪盖、风云静止卫星降水、GLDAS气温等多源数据,作为传统SRM模型的输入参数,构建多源遥感驱动的SRM融雪径流模型,并在缺资料地区——青藏高原的年楚河流域进行融雪过程的径流模拟。研究表明融雪后期的瞬时降雪很大程度上影响了插值后积雪覆盖率的精度,在插值的时候考虑降水和气温,排除瞬时积雪干扰,改进线性插值获得每天的积雪覆盖率,可以提高模型模拟精度;遥感驱动的SRM模型在缺资料地区年楚河适用性较好,Nash-Sutcliffe系数(NSE)达到0.681,体积差(Dv)为-0.17%,均方根误差(RMSE)为9.678,模型模拟的精度较高。研究结果可为高寒地区生态水文模型研究提供重要参考,同时可为SRM模型在其他流域尤其是缺资料地区融雪径流计算中的应用提供有效支撑。

The Snowmelt Runoff Model based on Multi-source Remote Sensing Studied and Applied in Ungauged Basins

Technology of remote sensing has been widely used to obtain hydrological variables and parameters.With snow|cover of MODIS,precipitation of FY meteorological satellites,air temperature of GLDAS remote sensing data are as the main input variables,the snowmelt runoff model based on multi source remote sensing was established and was used to simulate the snowmelt process in ungauged Nyangqu River Basins in Tibetan Plateau.When obtaived the input variable of snow by snow|cover of MODIS,as for the instantaneous snowfall largely affects the precision of calculative snow cover,considering the temperature and precipitation,removing the effects of instantaneous snow,daily coverage of snow and ice were obtained by improved linear interpolation which can improve model simulation accuracy.The simulation fit indexes of Nash|Sutcliffe coefficient(NSE)is 0.681,the volume difference(Dv)is -0.17% and RMSE error is 9.678.The snowmelt runoff model based on remote sensing is applicable in the Nyangqu River Basin.The precision of the simulation model is higher.This study can provide important reference for the research on Ecohydrological models in the cold regions and effective support for application of other river basin especially lack of data area.