|dc.description.abstract||The Boreal Plains (BPs) ecoregion spans the northern potions of Alberta, Saskatchewan, and Manitoba and is an area of high ecological sensitivity. With large industrial developments in the region, including the Athabasca Oil Sands extraction projects, informed decision making and reclamation is critical. Hydrologic models are tools which are often used to inform such tasks. The BPs are characterized by their deep soils, their mosaic of forests and wetlands and their corresponding complicated hydrology. This complicated hydrology, including variable hydrologic connectivity, fill and spill mechanisms, and variable annual moisture deficit make selecting or developing appropriate hydrologic models a challenge. Current fixed model approaches have thus far been unable to demonstrate good representation of the hydrology of the BPs.
To address this gap in the literature for BPs hydrologic representation, three model structures were developed which attempt to capture the complicated physical nature of the BPs. This was achieved by utilizing an iterative, step-wise, and flexible model development approach within the Raven Hydrologic Modelling Framework (Raven). Additionally, physical realism was checked throughout the development process using multiple model diagnostic criteria and hydrologic signatures. Three study basins in the Athabasca River Basin were used to calibrate and validate the model structures. The results were compared to a baseline model which employed standard fixed modelling approaches.
The model development process faced numerous challenges including limited data availability, limited understanding of the physical environment at large scales, accurately representing wetland functional groups, representing the variable contributing area, equifinality of calibrated data sets, and limiting available winter algorithms in Raven. For the three data-limited basins examined here, it was found that lack of sufficient data made it difficult to properly constrain model structure and parameterization. Due to the complexity of BPs hydrology, it was found that while inclusion of additional BPs-specific hydrologic structures generally improved model performance in calibration, similar performance was unattainable in validation. This indicates that basins in the BPs will likely required additional data sources, beyond hydrographs, to properly inform and constrain local models.||en