High-resolution weather research forecasting (WRF) modeling and projection over western Canada, including mackenzie watershed

Abstract

Weather Research Forecasting (WRF) model was run at a Convection-Permitting (CP) 4-km resolution to dynamically downscale the 19-member CMIP5 ensemble mean projection to assess the hydroclimatic risks in Western Canada under high-end emission scenario RCP8.5 by the end of twenty-first century. A retrospective simulation (CTL, 2000-2015) forced by ERA-Interim and a Pseudo-Global Warming (PGW) forced with the reanalysis plus the climate change forcing (2071-2100-1976-2005) were derived using CMIP5 ensemble. The surface air temperature of WRF-CTL, evaluated against gridded analysis ANUSPLIN, shows good agreements in the geographical distribution. There are cold biases east of the Canadian Rockies, especially in spring. WRF-CTL's precipitation resembles the geographical distribution of CaPA and ANUSPLIN. The wet bias mainly resides near the British Columbia coast in winter and over on the eastern side of the Canadian Rockies in summer. WRF-PGW shows much larger warming over the polar region in the northeast during the cold season relative to WRF-CTL. Precipitation increases in most areas in spring and autumn, whereas unchanged or decreased precipitation in summer occurs in the Saskatchewan River Basin and southern Canadian Prairies. The flat precipitation changes cannot compensate the enhanced evapotranspiration over the region causing the water stress for the rain-fed agriculture during the growing season in the future. WRF-PGW projects lower warming than that by the CMIP5 ensemble throughout the year. The CMIP5 ensemble projects a much drier future over the Canadian Prairies with a 10-20% decrease of summer precipitation. The CMIP5 ensemble mean generally agrees with WRF-PGW except for regions with significant terrain, which may be due to WRF's higher resolution can represent small-scale summer convection and orographic lifting better. A larger increase of high-intensity precipitation events compared to lower intensity events, which indicates a higher risk for extreme events and lower effective rainfall for agriculture. New bias correction methods need to be developed to capture the shift in the precipitation intensity distribution in the future. The study also reveals the urgent need for high-quality meteorological observation to provide forcing data and evaluation benchmarks in Western Canada. The high-resolution dynamical downscaling over Western Canada provides opportunities for studying local-scale atmospheric dynamics and providing hydroclimatic data for cold region ecosystems, agriculture, and hydrology.