Exploring the Effects of GCM Uncertainty on the Hydrology and Water Allocation of a Small Mountain Watershed in Northern British Columbia, Canada.

Abstract:

Climate change and allocation of water supplies are causing water shortages and low flow conditions that threaten aquatic ecosystems around the world. Small mountain streams in Western Canada are experiencing increased water use from small diversion hydropower, increasing population, mining, agriculture, and changing energy extraction techniques. In addition, there are very few gauging sites for baseline water data because of the rugged mountain terrain and cold climate. Baseline data is important due to the sensitivity of small mountain streams to shifts in timing of snow pack melt and mid-winter melting, especially near and in coastal regions. Here we use HBV-EC to simulate the range in future flow in a northern mountain watershed under various climate scenarios and explore the uncertainty induced by different GMC models and downscaling for the Goathorn Creek watershed. To explore the effects of GCM model variability we selected four models (CGCM3, ECHAM5, GFDL-CM2.1, and CSIRO-Mk) and used the TreeGen downscaling method to generate multiple ensembles for emissions scenarios (A1B, A2 and B1) for each GCM model.

The calibrated HBV-EC model was sensitive to the climate inputs and produced a 50 percent variation in flows for the 2050’s and 2080’s with the greatest reduction in mean flows by 0.33 m³/s predicted for the 2020’s climate. Although, modeled future discharge is highly variable, some consistent trends are useful for water managers: results suggest spring discharge may occur up to two months earlier (CGCM3, A2 scenario), but was constantly one month earlier for all emission scenarios. This can lead to feasible management strategies such as granting fewer water permits or in areas with high future demand issuing permits with provisions for future storage.