Projections of Hydroclimatic Extremes in Southeast Alaska under the RCP8.5 scenario

Abstract

Parts of southeast Alaska experienced record drought in 2019, followed by record daily precipitation in late 2020 with substantial impacts to human health and safety, energy resources and fisheries. To help ascertain whether or not these types of events can be expected more frequently, this study investigated observed trends and projected changes of hydroclimatic extremes indices across southeast Alaska, including measures of precipitation variability, seasonality, magnitude and type. Observations indicated mixed tendencies of inter-annual precipitation variability, but there were consistent trends toward warmer and wetter conditions. Projected changes were assessed using dynamically downscaled climate model simulations at 4-km spatial resolution from 2031-2060 that were compared to a historical period from 1981-2010 using two models – NCAR CCSM4 and GFDL-CM3. Consistent directional changes were found for five of the analyzed indices. The CCSM indicated increased maximum 1-day precipitation (RX1; 12.6%), increased maximum consecutive 5-day precipitation (RX5; 7.4%), longer periods of consecutive dry days (CDD; 11.9%), fewer snow cover days (SNC; -21.4%) and lower snow fraction (SNF; -24.4%); for GFDL these changes were RX1 (19.8%), RX5 (16.0%), CDD (20.1%), SNC (-21.9%) and SNF (-26.5%). While both models indicated substantial snow losses, they also projected annual snowfall increases at high elevations; for CCSM this occurred above 1500 m and above 2500 m for GFDL. Significance testing was assessed at the 95% confidence level using Theil-Sen’s slope estimates for the observed time series and the Wilcoxon-Mann- Whitney U test for projected changes of the hydroclimatic extremes indices relative to their historical distributions.