This paper presents an assessment of the Soil and Water Assessment Tool(SWAT) on a glaciated(Qugaqie) and a non-glaciated(Niyaqu) subbasin of the Nam Co Lake. The Nam Co Lake is located in the southern Tibetan Plateau...This paper presents an assessment of the Soil and Water Assessment Tool(SWAT) on a glaciated(Qugaqie) and a non-glaciated(Niyaqu) subbasin of the Nam Co Lake. The Nam Co Lake is located in the southern Tibetan Plateau, two subbasins having catchment areas of 59 km^2 and 388 km^2, respectively. The scores of examined evaluation indices(i.e., R^2, NSE, and PBIAS) established that the performance of the SWAT model was better on the monthly scale compared to the daily scale. The respective monthly values of R^2, NSE, and PBIAS were 0.94, 0.97, and 0.50 for the calibration period while 0.92, 0.88, and -8.80 for the validation period. Glacier melt contribution in the study domain was simulated by using the SWAT model in conjunction with the Degree Day Melt(DDM) approach. The conjunction of DDM with the SWAT Model ensued improved results during both calibration(R^2=0.96, NSE=0.95, and PBIAS=-13.49) and validation (R^2=0.97, NSE=0.96, and PBIAS=-2.87) periods on the monthly time scale. Average contribution(in percentage) of water balance components to the total streamflow of Niyaqu and Qugaqie subbasins was evaluated. We found that the major portion(99.45%) of the streamflow in the Niyaqu subbasin was generated by snowmelt or rainfall surface runoff(SURF_Q), followed by groundwater(GW_Q, 0.47%), and lateral(LAT_Q, 0.06%) flows. Conversely, in the Qugaqie subbasin, major contributor to the streamflow(79.63%) was glacier melt(GLC_Q), followed by SURF_Q(20.14%), GW_Q(0.13%), and LAT_Q(0.089%). The contribution of GLC_Q was the highest(86.79%) in July and lowest(69.95%) in September. This study concludes that the performance of the SWAT model in glaciated catchment is weak without considering glacier component in modeling; however, it performs reasonably well in non-glaciated catchment. Furthermore, the temperature index approach with elevation bands is viable in those catchments where streamflows are driven by snowmelt. Therefore, it is recommended to use the SWAT Model in conjunction with DDM or energy base model to simulate the glacier melt contribution to the total streamflow. This study might be helpful in quantification and better management of water resources in data scarce glaciated regions.展开更多
基金supported by National Natural Science Foundation of China (41671067 and 41630754)State Key Laboratory of Cryosphere Science (SKLCS-ZZ-2015)
文摘This paper presents an assessment of the Soil and Water Assessment Tool(SWAT) on a glaciated(Qugaqie) and a non-glaciated(Niyaqu) subbasin of the Nam Co Lake. The Nam Co Lake is located in the southern Tibetan Plateau, two subbasins having catchment areas of 59 km^2 and 388 km^2, respectively. The scores of examined evaluation indices(i.e., R^2, NSE, and PBIAS) established that the performance of the SWAT model was better on the monthly scale compared to the daily scale. The respective monthly values of R^2, NSE, and PBIAS were 0.94, 0.97, and 0.50 for the calibration period while 0.92, 0.88, and -8.80 for the validation period. Glacier melt contribution in the study domain was simulated by using the SWAT model in conjunction with the Degree Day Melt(DDM) approach. The conjunction of DDM with the SWAT Model ensued improved results during both calibration(R^2=0.96, NSE=0.95, and PBIAS=-13.49) and validation (R^2=0.97, NSE=0.96, and PBIAS=-2.87) periods on the monthly time scale. Average contribution(in percentage) of water balance components to the total streamflow of Niyaqu and Qugaqie subbasins was evaluated. We found that the major portion(99.45%) of the streamflow in the Niyaqu subbasin was generated by snowmelt or rainfall surface runoff(SURF_Q), followed by groundwater(GW_Q, 0.47%), and lateral(LAT_Q, 0.06%) flows. Conversely, in the Qugaqie subbasin, major contributor to the streamflow(79.63%) was glacier melt(GLC_Q), followed by SURF_Q(20.14%), GW_Q(0.13%), and LAT_Q(0.089%). The contribution of GLC_Q was the highest(86.79%) in July and lowest(69.95%) in September. This study concludes that the performance of the SWAT model in glaciated catchment is weak without considering glacier component in modeling; however, it performs reasonably well in non-glaciated catchment. Furthermore, the temperature index approach with elevation bands is viable in those catchments where streamflows are driven by snowmelt. Therefore, it is recommended to use the SWAT Model in conjunction with DDM or energy base model to simulate the glacier melt contribution to the total streamflow. This study might be helpful in quantification and better management of water resources in data scarce glaciated regions.
