Abstract: Robust calibration of physically based hydrological models is essential for the quantification of runoff components in snow and glacier melt runoff fed basins. This study evaluates a hydrograph partitioning curve (HPC) based calibration technique utilizing seasonality of precipitation and temperature in the Yarlung Tsangpo River basin. The HPC indicate the separating periods when various runoff components, including baseflow, snowmelt and glacier melt runoff, and rainfall direct runoff, dominate the basin hydrograph. Parameters of the THREW distributed hydrological model are grouped into four categories according to their controls on the runoff processes, and subsequently calibrated in a stepwise procedure using extracted HPC. The HPC-based calibration method is evaluated against traditional methods on the basis of robustness, performance of discharge simulations in sub-basins, and estimates of snow water equivalent (SWE) across the whole basin. Results show that:① The HPC-based calibration method provides results comparable to traditional methods in the calibration period while improving the discharge simulation for the evaluation period relative to single objective calibration methods; ② The HPC-based calibration method shows superiority in producing robust sub-basin discharge and tends to estimate smaller bias for the basin SWE; ③ The HPC-based calibration method estimates the contributions of snowmelt, glacier-melt, and rainfall direct runoff to discharge during 2001—2015 as 20%, 14% and 66%, respectively, while the traditional calibration methods yield a higher contribution for glacier-melt runoff and lower contribution for snowmelt runoff. Our findings indicate the potential of the HPC-based calibration method as a tool to quantify the contribution of runoff components, thus improving the modeling of hydrological behaviors under changing climate conditions for similar basins.