Rachel White

Global climate models (GCMs) often lack the spatial resolution to capture regional features of high topography and coastlines such as the Pacific Northwest. Orographic precipitation and convection greatly influence the region’s climate, yet these processes are typically heavily parameterized in climate models and contribute to well-known modelling uncertainties. This study uses the Weather Research and Forecast (WRF) model to dynamically downscale the Canadian Earth System Model (CanESM2), a coarse GCM, to generate high-resolution (3 km) convection-permitting climate projections over southwestern British Columbia, Vancouver Island, and Washington. Dynamical downscaling is a modeling technique that extrapolates the effects of large-scale climate processes from the coarse models, using them as initial and boundary conditions to drive high-resolution models, allowing for small-scale processes, especially those affected by local geography (e.g., coasts, topography) such as convection, to be resolved within the model rather than parameterized. Three 20-year simulations were generated: a historical period (1986-2005) for validation and to assess the changes in the Pacific Northwest, and two mid-century simulations (2046-2065) under two climate mitigation scenarios (moderate and no mitigation). Evaluated against its coarser, non-convection-permitting nest, as well as CanESM2 and CanRCM4, the 3-km CanESM2-WRF model reveals significant reductions in biases for temperature, precipitation, and wind speed across over 100 stations. Following the strong response to emissions of CanESM2, 3-km CanESM2-WRF projects seasonal temperature increases between 2 and 3.5ºC. Summers are projected to become drier, while winters are expected to be wetter with increased frequency and intensity of precipitation events. Winter wind speeds are projected to increase along the coast but decrease in other seasons. Despite the increase in mean temperatures, extreme spring and winter temperatures show limited increases and even decreases compared to historical values. This study examines temperature and precipitation compound events using both historical and future extreme thresholds; providing insights into the frequency of such events under shifting climatic conditions. Projections based on future extreme thresholds show seasonally-dependent changes in extreme and compound events, with an increase in cold spells in most regions, a reduction in interior summer warm spells, and an increase in spring dry-hot days but decrease in spring dry-cold days.

View record