Integration of Observational Data and Dynamical Downscaling for Projections of Future Winds in Alaska

Evaluations of wind climatologies and high-wind events present challenges in regions such as Alaska where observational data are sparse and winds are affected by complex topography and coastal configuration. In order to provide a more comprehensive and homogeneous database for studies of winds in Alaska, we utilize 35 years of hourly wind data from 70 stations, adjusted for discontinuities identified by a change-detection algorithm, to quantile-map the wind distributions in a reanalysis-driven simulation by the Weather Research and Forecasting Model (WRF). The quantile-mapped wind distributions provide a basis for adjusting historical and future simulations (1980-2100) forced by two global climate models (GFDL CM3 and CCSM4) under the RCP 8.5 scenario. The quantile mapping is especially useful for depicting the extremes of the distributions of the hourly winds. The results show that wind climatologies and high-wind events have strong seasonal and regional variations within Alaska. When the percentage change from the present is used as a metric, future changes in high-wind events and wind-energy generation potential are greater than the changes in the mean wind speeds at many locations in Alaska. The high-wind events represent potential hazards through coastal flooding, wildfire spread and infrastructure damage.