Learning from Hydrologic Model Failure: a Case from Phase 1 OF the NELSON-MiP Project
Abstract
Intercomparison studies are important for assessing the performance of different hydrologic models and highlighting their advantages and limitations. However, such studies have not been implemented in complex regions such as Western Canada, which encompasses major areas of the prairie non-contributing region, commonly referred to as the "graveyard of hydrologic models". The Nelson-MiP project strategically seeks to better understand why models fail in this region by investigating the performance of eight hydrologic and land surface models within the transboundary Nelson Churchill River Basin (NCRB) in Southwestern Canada and Northwestern United States. The models are built using a standardized input datasets and are implemented over different unregulated watersheds within NCRB (Phase 1). Results show that all models are unable to simulate the flow of prairie watersheds, while they show acceptable simulation of flow elsewhere. Models showed significant discrepancies in simulating evaporative fluxes over the prairies and snow water equivalent amounts over the Rocky Mountain regions. This study highlights problems with the used models' underlying structure and assumptions, which can be improved in future. Further, the outcomes of this study can help practitioners in simulating the flows of the NCRB accurately by selecting the appropriate/successful candidate models at their area of interest, which will help in better allocation of available water resources in this highly populated basin.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2022
- Bibcode:
- 2022AGUFM.H46B..06S