Modelling Storage‐Discharge Relationships in an Alpine Basin in the Canadian Rockies

ABSTRACT A physically based hydrological model was set up in the Cold Regions Hydrological Modelling platform (CRHM) for a well‐instrumented alpine headwater basin in the Canadian Rockies in order to diagnose relationships between storage and discharge. The hydrological model was constructed and parameterised from the field research results in Fortress Mountain Research Basin (~5.9 km 2 ) and elsewhere and includes relevant streamflow generation processes such as blowing and gravitational snow redistribution, snow interception and sublimation, energy budget snowmelt, evapotranspiration, infiltration to frozen and unfrozen soils, overland, subsurface and groundwater flow along hillslopes, and streamflow routing. Uncalibrated simulations during 11 hydrological years (2013–2024) matched the observations of snow accumulation, snowmelt, and liquid soil moisture well. Uncalibrated streamflow simulations achieved Nash‐Sutcliffe efficiency (NSE) of 0.62 and 0.67 and modified Kling‐Gupta efficiency (KGE″) of 0.63 and 0.75 for Bonsai Creek and Fortress Creek, with model bias within 23%, and normalised root mean square error (NRMSE) no higher than 63%. The model was then used to examine the relationship between storage and discharge in the basin. The results show significant positive correlations between storage and discharge for the smaller Bonsai Creek ( r = 0.86) and larger Fortress Creek ( r = 0.83) and a nonlinear, hysteretic storage‐discharge relationship for both creeks. A “catchment sensitivity function” was combined with CRHM‐modelled water balance components over 11 seasonal recession periods to derive exponential storage‐discharge functions that can be used to estimate discharge. Using these functions to predict future discharge using CRHM storage state variables showed only marginal or no improvement over the CRHM‐simulated discharge. Using earlier storage state variables as initial conditions for the exponential storage‐discharge functions degraded their ability to estimate discharge. This suggests that simple exponential storage‐discharge functions are not sufficient to capture the complex hydrological dynamics during hydrograph recession in such an alpine basin.