| Summary: | Thesis (M.S.) University of Alaska Fairbanks, 2014. Alaska is experiencing effects of global climate change due, in large part, to the positive feedback mechanisms associated with polar amplification. The major risk factors include loss of sea ice, glaciers, thawing permafrost, increased wildfires, and ocean acidification. Reanalyses, which are weather forecast models that assimilate observations, are integral to understanding mechanisms of Alaska's past climate and to help calibrate future modeling efforts. This study evaluates five reanalyses using monthly gridded datasets of temperature, precipitation, and snowwater equivalent, as well as daily station data of maximum and minimum temperature, precipitation, and snow depth across six climate regions in Alaska, and at eight stations from 1979-2009. The reanalyses evaluated in this study include the: NCEP-NCAR Reanalysis (NCEP-R1), North American Regional Reanalysis (NARR), Climate Forecast System Reanalysis (CFSR), ERA-Interim, and Modern-Era Retrospective Analysis for Research and Applications (MERRA). MERRA was the top-performing reanalysis for the station-based assessment, has the lowest statewide precipitation bias, and is the most reliable model for snow-water equivalent. NARR and ERA-Interim have the lowest near-surface air temperature biases across Alaska. The quality of reanalysis data varies by region, season, and variable. This thesis provides guidance for reanalysis users to make informed decisions. Chapter 1. Alaska’s climate and modeling needs -- 1.1. Alaska’s changing climate -- 1.2. Previous usage of reanalysis for Alaska -- 1.3. Project goals -- Chapter 2. Meteorological surface observations and reanalysis data -- 2.1. Meteorological surface observations -- 2.1.1. Surface data -- 2.1.2. Gridded temperature and precipitation verification datasets -- 2.1.3. Gridded snow verification dataset -- 2.2. Reanalysis models and topography -- 2.2.1. NCEP-NCAR Reanalysis -- 2.2.2. North American Regional Reanalysis -- 2.2.3. Climate Forecast System ...
|
|---|