Quantitative water resources assessment of Qinghai Lake basin using Snowmelt Runoff Model (SRM).

Summary A Snowmelt Runoff Model (SRM) is used to simulate streamflow from snowmelt in the Buha watershed within the Qinghai Lake basin for the hydrologic years 2003–2009 (from September to August). Two different precipitation products are used as inputs to examine the model and their performances: station measured precipitation and gridded APHRODITE-precipitation, with two different snow cover products, namely, flexible multiday combined MODIS Terra–Aqua (MODISMC) maps, and the standard MODIS/Terra 8-day composite (MOD10A2) product. Results show that the MODISMC product with a higher temporal resolution of 2.2 days obtains slightly small bias in runoff volumes than MOD10A2 does for basin- and zone-wide simulations. The average Nash–Sutcliffe coefficient of determination ( R 2 ) is 0.73 (0.70–0.75) for basin- and zone-wide simulations. The simulated runoff with APHRODITE-precipitation is well correlated with measured runoffs ( R 2 = 0.76). The multivariate regression analysis indicates that lake level changes are significantly associated (95% level) with the following single variables: measured runoff discharge from the Buha watershed ( R m ), simulated runoff ( R s ), simulated runoff fraction from rainfall ( R s -rainfall), the snow cover area at zone B of the Buha watershed, or with a combination of multivariates: R s -rainfall + P L (precipitation over the lake) and R m + P L + Evaporation. These results suggest that precipitation plays a dominant role in lake level variations. Snowmelt contributed runoff shows no significant contribution to lake level variation (13% of the total runoff). This study suggests that APHRODITE-precipitation data can be used for snowmelt simulations and water resource management on the Tibetan Plateau where precipitation gauges are limited, and that the SRM is an effective tool for estimating snowmelt discharge and water resource management in high mountain regions. [ABSTRACT FROM AUTHOR]