Diagnosing the regional sensitivity of a process-based hydrologic model

Abstract:

Despite commonly being able to match streamflow dynamics (following calibration), hydrologic models often fail to produce simulations that are consistent with internal watershed processes. The Catchment Connectivity Model (CCM) was developed collaboratively by physical and computational hydrologists, following a dominant process conceptualization based on hillslope hydrologic connectivity at the Tenderfoot Creek Experimental Forest (TCEF; Montana, USA). The three-parameter CCM is a spatially explicit model structure that has been validated against extensive field observations of hillslope hydrologic connectivity (an internal process simulated by the model) to demonstrate its internal consistency, in addition to a traditional assessment to the external streamflow dynamics. In this study, we sought to examine the regional sensitivity of the CCM parameters. Specifically, we explored the catchment scale variation in model parameterization across seven TCEF watersheds. A diagnostic model calibration analysis was considered where the role of a priori parameter ranges was explored in relation to both external (streamflow) and internal (hydrologic connectivity) model performance. This investigation allowed us to develop a more in-depth understanding of the model structure, its flexibility, its regional sensitivity, and its transferability.