This dataset consists of 6000 GeoTIFFs produced by the Geophysical Institute Permafrost Lab (GIPL) Permafrost Model. Six distinct CMIP5 model-scenario combinations were used to force the GIPL model output. Each model-scenario combination includes annual (2021-2120) summaries of the following ten variables: - Mean Annual Ground Temperature (MAGT) at 0.5 m below the surface (°C) - MAGT at 1 m below the surface (°C) - MAGT at 2 m below the surface (°C) - MAGT at 3 m below the surface (°C) - MAGT at 4 m below the surface (°C) - MAGT at 5 m below the surface (°C) - Mean Annual Surface (i.e., 0.01 m depth) Temperature (°C) - Permafrost top (upper boundary of the permafrost, depth below the surface in m) - Permafrost base (lower boundary of the permafrost, depth below the surface in m) - Talik thickness (perennially unfrozen ground occurring in permafrost terrain, m) There are 1000 GeoTIFF files per model-scenario combination. The model-scenario combinations are: - GFDL-CM3, RCP 4.5 - GFDL-CM3, RCP 8.5 - NCAR-CCSM4, RCP 4.5 - NCAR-CCSM4, RCP 8.5 - A 5-Model (GFDL-CM3, NCAR-CCSM4, GISS-E2-R, IPSL-CM5A-LR, MRI-CGCM3) Average, RCP 8.5 - A 5-Model (GFDL-CM3, NCAR-CCSM4, GISS-E2-R, IPSL-CM5A-LR, MRI-CGCM3) Average, RCP 4.5 The file naming convention is `gipl_model_scenario_variable_year.tif` for example: `gipl_GFDL-CM3_rcp45_talikthickness_m_2090.tif` Each GeoTIFF uses the Alaska Albers (EPSG:3338) projection and has a spatial resolution of 1 km x 1 km. All rasters in this dataset have indentical extents, spatial references, and metadata objects. Once extracted, the entire dataset (all 6000 GeoTIFFs) requires 39 GB of disk space. Data are compressed into ten .zip files, one per variable. Each archive will contain all model-scenario combinations and all years for that variable. Each .zip file contains 600 GeoTIFFs. This research was funded by the Broad Agency Announcement Program and the U.S. Army Engineer Research and Development Center and Cold Regions Research and Engineering Laboratory (ERDC-CRREL) under Contract No. W913E521C0010. The GIPL2-MPI/GCM simulations were supported in part by the high-performance computing and data storage resources operated by the Research Computing Systems Group at the University of Alaska Fairbanks Geophysical Institute.

These GeoTIFFs include annual spatial representations of the following variables produced through summarization of ALFRESCO model outputs across 200 replicates: Flammability: likelihood of a pixel to burn across 200 replicates Modal vegetation type: statistical mode of vegetation type across 200 replicates Percent vegetation type: percent of each possible vegetation type across 200 replicates These outputs were derived from AR5/CMIP5 climate inputs, historical fire inputs from the Alaska Interagency Coordination Center (AICC), and several fire management options (FMO) inputs.

This dataset consists of spatial representations of relative vegetation change produced through summarization of ALFRESCO model outputs. These specific outputs are from the Integrated Ecosystem Model (IEM) project, and are from the linear coupled version using AR5/CMIP5 climate inputs (IEM Generation 2).

This set of files includes downscaled historical estimates of monthly temperature (in degrees Celsius, no unit conversion necessary) from 1901 - 2013 (CRU TS 3.22) at 10 min x 10 min spatial resolution. The downscaling process utilizes CRU CL v. 2.1 climatological datasets from 1961-1990.

This data set includes weekly (January 1954 to December 2013) and monthly (January 1850 to December 2022) midpoint historical sea ice concentration (0 - 100%) estimates at 1/4 x 1/4 degree spatial resolution for the ocean region around the state of Alaska, USA. This value-added dataset was developed by compiling the below historical data sources into spatially and temporally standardized datasets. Gaps in temporal or spatial resolutions were filled in with spatial and temporal analog month approaches. This dataset is no longer being updated. The NSIDC provides a new version in netCDF format receiving ongoing updates: https://nsidc.org/data/nsidc-0051/versions/2.

This dataset includes PRISM derived 1961-1990 climatologies of monthly average, maximum, and minimum temperature and total precipitation across Alaska and Western Canada including the Yukon, British Columbia, Alberta, Saskatchewan, and Manitoba. These were obtained from the PRISM Climate Group and mosaicked into a single continuous transboundary extent. Please cite the PRISM Climate Group when using this data.

This set of files includes downscaled future projections of vapor pressure (units=hPa) at a 1km spatial scale. This data has been prepared as model input for the Integrated Ecosystem Model (IEM). There can be errors or serious limitations to the application of this data to other analyses. The data constitute the result of a downscaling procedure using 2 General Circulation Models (GCM) from the Coupled Model Intercomparison Project 5 (CMIP5) for RCP 8.5 scenario (2006-2100) monthly time series and Climatic Research Unit (CRU) TS2.0 (1961-1990,10 min spatial resolution) global climatology data. Please note that this data is used to fill in a gap in available data for the Integrated Ecosystem Model (IEM) and does not constitute a complete or precise measurement of this variable in all locations. RCPs: 8.5 Centers, Model Names, Versions, and Acronyms: National Center for Atmospheric Research,Community Earth System Model 4,NCAR-CCSM4 Meteorological Research Institute,Coupled General Circulation Model v3.0,MRI-CGCM3 Methods of creating downscaled relative humidity data: 1. The GCM input data are distributed as relative humidity along with the CRU CL 2.0, therefore no conversion procedure was necessary before beginning the downscaling procedure. 2. Proportional Anomalies generated using the 20c3m Historical relative humidity data 1961-1990 climatology and the projected relative humidity data (2006-2100). 3. These proportional anomalies are interpolated using a spline interpolation to a 10min resolution grid for downscaling with the CRU CL 2.0 Relative Humidity Data. 4. The GCM proportional anomalies are multiplied by month to the baseline CRU CL 2.0 10min relative humidity climatology for the period 1961-1990. Creating a downscaled relative humidity projected time series 2006-2100. 5. Due to the conversion procedure and the low quality of the input data to begin with, there were values that fell well outside of the range of acceptable relative humidity (meaning that there were values >100 percent), these values were re-set to a relative humidity of 95 at the suggestion of the researchers involved in the project. It is well known that the CRU data is spotty for Alaska and the Circumpolar North, due to a lack of weather stations and poor temporal coverage for those stations that exist. 6. The desired output resolution for the AIEM modeling project is 1km, so the newly created downscaled time series is resampled to this resolution using a standard bilinear interpolation resampling procedure. 7. The final step was to convert the downscaled relative humidity data to vapor pressure using the calculation below, which uses a downscaled temperature data set created utilizing the same downscaling procedure. EQUATION: saturated vapor pressure = 6.112 x exp(17.62 x temperature/(243.12+temperature)) vapor pressure = (relative humidity x saturated vapor pressure)/100

This set of files includes downscaled projections of monthly means, and derived annual, seasonal, and decadal means of monthly mean temperatures (in degrees Celsius, no unit conversion necessary) from Jan 2006 - Dec 2100 at 771x771 meter spatial resolution. For seasonal means, the four seasons are referred to by the first letter of 3 months making up that season: * `JJA`: summer (June, July, August) * `SON`: fall (September, October, November) * `DJF`: winter (December, January, February) * `MAM`: spring (March, April, May) The downscaling process utilizes PRISM climatological datasets from 1971-2000. Each set of files originates from one of five top-ranked global circulation models from the CMIP5/AR5 models and RCPs or is calculated as a 5 Model Average.

These files include downscaled projections of decadal average monthly snow-day fraction ("fs", units = percent probability from 1 – 100) for each month of the decades from 2010-2019 to 2090-2099 at 771 x 771 m spatial resolution. Each file represents a decadal average monthly mean. Output is available for the CCSM4, GFDL-CM3, GISS-E2-R, IPSL-CM5A-LR, and MRI-CGCM3 models and three emissions scenarios (RCP 4.5, RCP 6.0 and RCP 8.5). These snow-day fraction estimates were produced by applying equations relating decadal average monthly temperature to snow-day fraction to downscaled decadal average monthly temperature. Separate equations were used to model the relationship between decadal monthly average temperature and the fraction of wet days with snow for seven geographic regions in the state: Arctic, Western Alaska, Interior, Cook Inlet, SW Islands, SW Interior, and the Gulf of Alaska coast, using regionally specific logistic models of the probability that precipitation falls as snow given temperature based on station data fits as in McAfee et al. 2014. These projections differ from McAfee et al. 2014 in that updated CMIP5 projected temperatures rather than CMIP3 temperatures were used for the future projections. Although the equations developed here provide a reasonable fit to the data, model evaluation demonstrated that some stations are consistently less well described by regional models than others. It is unclear why this occurs, but it is likely related to localized climate conditions. Very few weather stations with long records are located above 500m elevation in Alaska, so the equations used here were developed primarily from low-elevation weather stations. It is not clear whether the equations will be completely appropriate in the mountains. Finally, these equations summarize a long-term monthly relationship between temperature and precipitation type that is the result of short-term weather variability. In using these equations to make projections of future snow, as assume that these relationships remain stable over time, and we do not know how accurate that assumption is. These snow-day fraction estimates were produced by applying equations relating decadal average monthly temperature to snow-day fraction to downscaled projected decadal average monthly temperature. The equations were developed from daily observed climate data in the Global Historical Climatology Network. These data were acquired from the National Climatic Data Center in early 2012. Equations were developed for the seven climate regions described in Perica et al. (2012). Geospatial data describing those regions was provided by Sveta Stuefer. Perica, S., D. Kane, S. Dietz, K. Maitaria, D. Martin, S. Pavlovic, I. Roy, S. Stuefer, A. Tidwell, C. Trypaluk, D. Unruh, M. Yekta, E. Betts, G. Bonnin, S. Heim, L. Hiner, E. Lilly, J. Narayanan, F.Yan, T. Zhao. 2012. NOAA Atlas 14. Precipitation-Frequency Atlas of the United States.

This set of files includes downscaled modeled historical estimates of monthly temperature (in degrees Celsius, no unit conversion necessary) from 1901 - 2005 at 15km x 15km spatial resolution. Each set of files originates from one of five top-ranked global circulation models from the CMIP5/AR5 models and RCPs or is calculated as a 5 Model Average. These outputs are from the Historical runs of the GCMs. The downscaling process utilizes CRU CL v. 2.1 climatological datasets from 1961-1990 as the baseline for the Delta Downscaling method.