Developing High Resolution Climate Data for Alaska
Dynamical Downscaling of Alaska Climate Data (Alaska Host Agreement project)
Dates
Start Date
2017-08-01
End Date
2022-07-31
Release Date
2017
Summary
Alaska has complex topography, with its extensive coastlines, dozens of islands, and mountain ranges that contain the tallest peaks in North America. Topography can have a strong influence on temperature and precipitation, therefore accurate representations of the terrain can improve the quality of simulations of past and future climate conditions. The spatial resolution of globally-available climate data is typically too coarse (~80 to 100 km) to adequately detect local landscape features, meaning these models aren’t useful for predicting future conditions in Alaska. In order for the state to adequately prepare for and adapt to changing conditions, high-resolution climate data is needed. One solution for acquiring this data is to [...]
Summary
Alaska has complex topography, with its extensive coastlines, dozens of islands, and mountain ranges that contain the tallest peaks in North America. Topography can have a strong influence on temperature and precipitation, therefore accurate representations of the terrain can improve the quality of simulations of past and future climate conditions. The spatial resolution of globally-available climate data is typically too coarse (~80 to 100 km) to adequately detect local landscape features, meaning these models aren’t useful for predicting future conditions in Alaska. In order for the state to adequately prepare for and adapt to changing conditions, high-resolution climate data is needed.
One solution for acquiring this data is to use dynamical downscaling, a technique in which higher resolution weather forecasting models are used to provide local context to global-scale data. As part of this project, researchers have applied dynamical downscaling to ERA-Interim global atmospheric data for the years 1979-2015, and to two future climate model projections derived from the CMIP5 RCP8.5 scenario for the years 1970-2100. These data provide hourly information at 20 km resolution for all of Alaska and feature more than 30 meteorological variables, including temperature, precipitation (rain vs. snow), winds, and humidity. Researchers have so far evaluated the accuracy of the downscaled ERA-Interim temperature and precipitation simulations by comparing the downscaled estimates to actual observed conditions.
The downscaled ERA-Interim data can be used by stakeholders to investigate climate- and weather-related phenomena in Alaska. Combining an improved understanding of these phenomena with projections of future climate conditions derived from the downscaled climate scenario data (1970-2100) can help stakeholders effectively plan for future climate conditions in Alaska.
Click on title to download individual files attached to this item.
SukukpakPeak_DaltonHwy_AK_BobWick_BLM.jpg “Sukakpak Peak, Alaska - Credit: Bob Wick, BLM”
389.01 KB
image/jpeg
Purpose
Climate data used to evaluate change in Alaska are often too coarse to resolve key local features especially in complex topography. Dynamical downscaling uses a regional weather forecasting model to help provide the local context. Dynamical downscaling has been applied to the ERA-Interim reanalysis for 1979-2015 and two future climate model projections from the CMIP5 RCP8.5 scenario (NCAR-CCSM4 and GFDL-CM3) from for 1970-2100. These data provide 20km and hourly resolution for the entire Alaska domain and feature over 30 variables including temperature, precipitation (rain vs. snow), winds, humidity and radiative/turbulent fluxes. The downscaled reanalysis has been compared against observed temperature and precipitation to assess biases for Alaska and the data have been used in 5 additional published studies. The downscaled data have been distributed to more than 30 users for uses that range from glacier/hydrological model to blogging. The final Alaska downscaled database is quite large (~100TB) and efforts are currently underway to post-process and organize the data to best serve the needs of users.