Alaska Climate Futures (mid and late 21st century) and Historical References (20th century)

To meet the climate change planning and adaptation needs of Alaska managers and decision makers, I developed a set of statewide summaries of available climate change projections that can be further subset using GIS techniques for requests by management unit, watershed, or other location. This facilitates the development of tailored climate futures for decision makers’ regional or subregional management context. This file describes the source data and summaries for purposes of technical /scientific documentation.

The methods and presentation for these datasets were adapted from products in previous USGS-approved IP products for the AKCASC Building Resilience Today project (e.g, Community of Kotlik et al. 2019).

For each data product included, summaries (averages or totals) are presented for multiple climate models or specific global warming levels and are average dover two time periods: 2040-2069, or the “2050s”, for near-term decision framing; and 2070-2099, or the “2080s”, for longer-term decision framing. In all cases where possible, both moderate emissions (RCP4.5 or +2C global level) and higher emissions (RCP8.5, or +4C global level) are presented. These choices (model averaging, temporal averaging, and scenario presentation) are tailored to the main sources of uncertainty (Hawkins and Sutton 2009) in climate model projections, specifically differences in climate model construction, climatic variability, and emissions scenario uncertainty (e.g., Littell et al. 2011, Snover et al. 2013, Terando et al. 2020). Not all scenario planning or climate impacts modeling needs can be met with these projections – these are intended to characterize a range of futures indicated by the available data products and facilitate further exploration of climate impacts modeling and adaptation development options.

Variable descriptions:
 


1. Statistically downscaled temperature, precipitation, and derived snow variables
 
1.1. Temperature
• Average of 5 statistically downscaled climate models: MRI-CGCM3, GISS-E2-R, GFDL-CM3, IPSL-CM5A-LR and NCAR-CCSM4, bias corrected 
• Annual; winter (DJF); spring (MAM); summer (JJA); autumn (SON)
• 771m resolution (Alaska wide) 
• Original data in degrees C; summary graphics and tables are presented as climatology “deltas” in both degrees C and degrees F, changes (for the period 2040-2069 and 2070-2099) relative to historical (1970-1999) 
• RCP 4.5 and RCP8.5
• Details: Walsh et al. 2018
 
1.2. Precipitation
• Average of 5 statistically downscaled climate models: MRI-CGCM3, GISS-E2-R, GFDL-CM3, IPSL-CM5A-LR and NCAR-CCSM4 , bias corrected 
• Annual; winter (DJF); spring (MAM); summer (JJA); autumn (SON)
• 771m resolution (Alaska wide) 
• Graphics are presented as climatology “deltas” in percent change in total, changes (for the period 2040-2069 and 2070-2099) relative to historical (1970-1999) 
• Details: Walsh et al. 2018
 
1.3. Snowfall water equivalent
• Average of 5 statistically downscaled climate models: MRI-CGCM3, GISS-E2-R, GFDL-CM3, IPSL-CM5A-LR and NCAR-CCSM4, bias corrected 
• Total snowfall water equivalent for October to March
• 771m resolution (Alaska wide) 
• Graphics are presented as climatology “deltas” in percent, changes (for the period 2040-2069 and 2070-2099) relative to historical (1970-1999) 
• RCP 4.5 and RCP8.5
• Details: Littell et al. 2018
 
1.4. Snow index
• Average of 5 statistically downscaled climate models: MRI-CGCM3, GISS-E2-R, GFDL-CM3, IPSL-CM5A-LR and NCAR-CCSM4 , bias corrected 
• Ratio of snowfall water equivalent to total October-March precipitation, equivalent to the fraction of total October to March precipitation entrained in snowfall water equivalent
• 771m resolution (Alaska wide) 
• Graphics are presented as climatology “deltas”, changes (for the period 2040-2069 and 2070-2099) relative to historical (1970-1999)
• RCP 4.5 and RCP8.5
• Details: Littell et al. 2018 
 
1.5. Change in months of reliable snow
• Average of 5 statistically downscaled climate models: MRI-CGCM3, GISS-E2-R, GFDL-CM3, IPSL-CM5A-LR and NCAR-CCSM4, bias corrected 
• Change in count of months with >70% precipitation as snow
• 771m resolution (Alaska wide) 
• Graphics are presented as climatology “deltas”, changes (for the period 2040-2069) relative to historical (1970-1999)
• Only RCP8.5
 
2. Derived land surface variables consistent with temperature and precipitation above
 
2.1. Changes in fires per century, ALFRESCO landscape fire model
• MRI-CGCM3 (lower fire activity) and NCAR-CCSM4 (higher fire activity)
• Change (difference) between 1900-1999 and 2000-2099 fires per model pixel, derived from raw “relative flammability”, or the count of model replicate fires per pixel per time period 
• 2km resolution (Alaska and NW Canada)
• Graphics are presented as “deltas” changes (for the 21st century relative to historical 20th century) in fires per century
• RCP4.5 and RCP8.5
• Details: McGuire et al. 2018, Euskirchen et al. 2020
 
2.2. Changes in vegetation per century, ALFRESCO landscape fire model
• MRI-CGCM3 (lower fire activity) and NCAR-CCSM4 (higher fire activity)
• Change (difference) between 1900-1999 and 2000-2099 vegetation changes per model pixel
• 2km resolution (Alaska and NW Canada)
• Graphics are presented as “deltas” changes (for the 21st century relative to historical 20th century) in vegetation type per century
• RCP4.5 and RCP8.5
• Details: McGuire et al. 2018, Euskirchen et al. 2020
 
2.3. New Picea glauca (white spruce) establishment, ALFRESCO landscape fire model
• MRI-CGCM3 (lower fire activity) and NCAR-CCSM4 (higher fire activity)
• New spruce colonization in 2050 and 2100 in areas without spruce in historical vegetation map.
• 2km resolution (Alaska and NW Canada)
• Graphics are presented as new establishment at two densities, basal area 0-5m^2/ha for 2050 and 2100 and 5-12m^2/ha for 2050
• Only RCP8.5
• Details: McGuire et al. 2018, Euskirchen et al. 2020
 
3. TerraClimate hydrologic changes
 
3.1. Change in spring runoff
• CMIP5 climate model ensemble of deltas at global warming levels (see below)
• Change in volume of monthly total runoff (Q) per pixel relative to historical (1981-2010) for Feb, Mar, Apr, May
• 4km resolution (global, subset for Alaska region)
• Graphics are presented as change (difference) mm of total Q from historical reference
• No emissions, +2C and +4C global warming levels
• Details: Abatzoglou et al. 2018
 
3.2. Change in summer (JJA) and growing season (AMJJAS) potential evapotranspiration
• CMIP5 climate model ensemble of deltas at global warming levels (see below)
• Change in volume of summer and growing season total potential evapotranspiration (PET) per pixel relative to historical (1981-2010) for Jun-Aug and Apr-Sep
• 4km resolution (global, subset for Alaska region)
• Graphics are presented as change (difference) mm of total PET from historical reference
• No emissions, +2C and +4C global warming levels
• Details: Abatzoglou et al. 2018
 
3.3. Change in summer (JJA) and growing season (AMJJAS) water balance deficit
• CMIP5 climate model ensemble of deltas at global warming levels (see below)
• Change in volume of summer and growing season total potential evapotranspiration (PET) minus actual evapotranspiration (AET) per pixel relative to historical (1981-2010) for Jun-Aug and Apr-Sep
• 4km resolution (global, subset for Alaska region)
• Graphics are presented as change (difference) mm of total PET-AET from historical reference
• No emissions, +2C and +4C global warming levels
• Details: Abatzoglou et al. 2018



Data availability
 
Raw data used to compute the above are available from online archives.
 
Locations:
 

- http://data.snap.uaf.edu/data/Base/AK_771m/projected/AR5_CMIP5_models/P…

 
1.2 -  http://data.snap.uaf.edu/data/Base/AK_771m/projected/AR5_CMIP5_models/P…
 
 
1.3 - http://data.snap.uaf.edu/data/Base/AK_771m/projected/AR5_CMIP5_models/S…
 
1.4 - http://data.snap.uaf.edu/data/Base/AK_771m/SWE_dSWE_SFEtoPR/
 
1.5 – derived
 
2.1 - http://data.snap.uaf.edu/data/IEM/Outputs/ALF/Gen_1a/alfresco_relative_…
 
2.2 - http://data.snap.uaf.edu/data/IEM/Outputs/ALF/Gen_1a/alfresco_relative_…
 
2.3 - http://data.snap.uaf.edu/data/IEM/Outputs/ALF/Gen_1a/best_rep_outputs/A…
 
3.1 –

http://thredds.northwestknowledge.net:8080/thredds/catalog/TERRACLIMATE…
http://thredds.northwestknowledge.net:8080/thredds/catalog/TERRACLIMATE…
http://thredds.northwestknowledge.net:8080/thredds/catalog/TERRACLIMATE…

3.2 –


http://thredds.northwestknowledge.net:8080/thredds/catalog/TERRACLIMATE…
http://thredds.northwestknowledge.net:8080/thredds/catalog/TERRACLIMATE…
http://thredds.northwestknowledge.net:8080/thredds/catalog/TERRACLIMATE…

3.3 –


http://thredds.northwestknowledge.net:8080/thredds/catalog/TERRACLIMATE…
http://thredds.northwestknowledge.net:8080/thredds/catalog/TERRACLIMATE…
http://thredds.northwestknowledge.net:8080/thredds/catalog/TERRACLIMATE…

 
 
Methods
 

Temperature

 
Seasonal (Annual – Jan-Dec; Winter – Dec-Feb; Spring – Mar-May; Summer – Jun-Aug; Autumn – Sep-Nov) decadally averaged historical and projected future temperature data (GeoTiff raster files) were obtained from UAF/SNAP/IARC for historical decades 1970-1979, 1980-1989, and 1990-1999 and future decades for 2040-2049, 2050-2059, 2060-2069, 2070-2079, 2080-2089, and 2090-2099. Three climatologies were computed in R (4.1.1 – “Kick Things”, with packages = library(sf); library(raster); library(sp); library(rgdal); library(maptools)) for mean temperature by computing averages of three decades: 1970-1999, 2040-2069, and 2070-2099. The future climatologies were further replicated for all five CMIP5 GCMs available (MRI-CGCM3, GISS-E2-R, GFDL-CM3, IPSL-CM5A-LR and NCAR-CCSM4) and for both RCP4.5 and RCP8.5 pathways, resulting in one historical and ten future scenarios for each annual or seasonal variable. The five-GCM average was computed for both RCP4.5 and RCP4.5 for annual and the four seasons as well. The change (“delta”) in  temperature was computed by subtracting the historical (1970-1999) from each projected future climatology. Data are in SI units (centigrade) and are converted to Farenheit (or scale bars appropriately scaled) for display purposes.
 

Precipitation

 
Same as for temperature above, except that future changes are computed in percent rather than absolute units of mm, i.e., delta Pr = (future climatology – historical climatology)/historical climatology.
 

Snowfall water equivalent

 
Same as for temperature and precipitation, except thirty-year climatologies were pre-calculated as products from Littell et al. 2018. Of note, in Littell et al. 2018, snow day fraction based on McAfee et al. 2013 and downscaled climate in 1.1 and 1.2 are used to calculate 1.3. Raw data are expressed as fractions of 1, so must be scaled to percent for display purposes to be the same as precipitation.
 

Snow Index

 
Same as snowfall water equivalent, except no deltas are calculated. The interpretation of this index is as a percent value out of 100, so the delta is less meaningful
 

Change in months of reliable snow

 
Heuristic exploration of historical data for 1.3 suggest that 70% of precipitation as snowfall results in “snow dominated” watersheds by month. 50% was too low – essentially this results in transitional snowpack, which wouldn’t be considered “reliable”. 70% is likely a generous estimate (probably requires more than 70% to be “reliable” across interannual variability and for most uses dependent on full landscape coverage of snowpack), but also depends on region, mean snowpack, and snow redistribution, which is not simulated here. Individual months of snowfall water equivalent in Littell et al 2018 were computed only for mid-century (2040-2069) and only for RCP 8.5, and the total number of months with snow day fraction > 70 were calculated for the historical and future. Change in months is future minus historical.
 

Changes in fires per century

 
Raw data (GeoTiff rasters) for ALFRESCO “relative flammability” were downloaded for two bracketing GCMs and for two different time periods: 1900-2099 and 2000-2099. However, “relative flammability” doesn’t’ give much insight into probability of permanent vegetation change. But if fires per century exceed approximately one, it is unlikely that boreal forest will remain dominated by spruce and will instead transition to deciduous forest. Also a general increase or decrease in number of fires per century is more intuitively scalable to most users than negative or positive relative flammability. The relative flammability data indicate the number of times each cell burned during the specified period given the transient summer climate from the GCM across 200 ALFRESCO model replicates. To convert this value to fires per century, the total number of fires per cell for 2000-2099 (future) must be separated (via subtraction) from the full 1900-2099 (historical + future) period, and then the results must be divided by the number of replicates to get mean fires per century per replicate. The change is calculated as (2000-2099 – 1900-1999)/200.
 

Changes in vegetation per century

 
Same as 2.1, except for vegetation change data.
 

New Picea glauca establishment

 
Raw data (GeoTiff rasters) for ALFRESCO basal area of white spruce were downloaded for two bracketing GCMs. For year 2050 and year 2100, cells with 0-5 m^2/ha were extracted; for 2100, cells with 5-12 m^2/ha were extracted. These are consistent with “new spruce establishing” and/or “new spruce established” densities and indicate places on the landscape where tundra vegetation is being rapidly invaded by spruce.
 
3.1 Changes in spring runoff
 
Original monthly runoff (Q) data was downloaded from the TerraClimate thredds site as NetCDF files and converted in R to GeoTiff rasters to include with variables in 1 and 2 above, but at coarser resolution (4km native). Historical (1980-2010) monthly and future (analogous +2C global and +4C global warming levels) monthly Q were extracted for Feb, Mar, Apr, and May and the monthly deltas calculated as differences (1980-2010 minus +2C future and, separately, minus +4C future).
 
3.2 Change in summer (JJA) and growing season (AMJJAS) potential evapotranspiration
 
Same as 3.1, but for potential evapotranspiration and for Apr, May, Jun, Jul, Aug and Sep. Growing season (April-September total) and summer (June-August total) potential evapotranspiration were calculated for historical (1980-2010) and future (analogous +2C global and +4C global warming levels). Seasonal deltas calculated as differences (1980-2010 minus +2C future and, separately, minus +4C future).
 
3.3 Change in summer (JJA) and growing season (AMJJAS) water balance deficit
 
Same as 3.2, except deficit is equal to PET minus actual evapotranspiration (PET-AET). AET summaries were calculated as in 3.2, then subtracted from PET before the differencing for historical relative to future.
 
 
Data files
 
Three archives nominally described according to source/resolution are included.
 
1. Deltas_771m_sources includes summary temperature (dTa), precipitation (dPr), and snowpack (dSFWE) outputs 

1.1 dTa includes changes (d) in air temperature (Tas) in C degrees

dtas_ANN_5mm.2050s.r45.tif
dtas_ANN_5mm.2050s.r85.tif
dtas_ANN_5mm.2080s.r45.tif
dtas_ANN_5mm.2080s.r85.tif
dtas_DJF_5mm.2050s.r45.tif
dtas_DJF_5mm.2050s.r85.tif
dtas_DJF_5mm.2080s.r45.tif
dtas_DJF_5mm.2080s.r85.tif
dtas_JJA_5mm.2050s.r45.tif
dtas_JJA_5mm.2050s.r85.tif
dtas_JJA_5mm.2080s.r45.tif
dtas_JJA_5mm.2080s.r85.tif
dtas_MAM_5mm.2050s.r45.tif
dtas_MAM_5mm.2050s.r85.tif
dtas_MAM_5mm.2080s.r45.tif
dtas_MAM_5mm.2080s.r85.tif
dtas_SON_5mm.2050s.r45.tif
dtas_SON_5mm.2050s.r85.tif
dtas_SON_5mm.2080s.r45.tif
dtas_SON_5mm.2080s.r85.tif
 
1.2 dPr includes changes (d) in precipitation (Pr), in %

dpr_ANN_5mm.2050s.r45.tif
dpr_ANN_5mm.2050s.r85.tif
dpr_ANN_5mm.2080s.r45.tif
dpr_ANN_5mm.2080s.r85.tif
dpr_DJF_5mm.2050s.r45.tif
dpr_DJF_5mm.2050s.r85.tif
dpr_DJF_5mm.2080s.r45.tif
dpr_DJF_5mm.2080s.r85.tif
dpr_JJA_5mm.2050s.r45.tif
dpr_JJA_5mm.2050s.r85.tif
dpr_JJA_5mm.2080s.r45.tif
dpr_JJA_5mm.2080s.r85.tif
dpr_MAM_5mm.2050s.r45.tif
dpr_MAM_5mm.2050s.r85.tif
dpr_MAM_5mm.2080s.r45.tif
dpr_MAM_5mm.2080s.r85.tif
dpr_SON_5mm.2050s.r45.tif
dpr_SON_5mm.2050s.r85.tif
dpr_SON_5mm.2080s.r45.tif
dpr_SON_5mm.2080s.r85.tif
 
1.3. dSFWE includes changes in “reliable” snow (fs_monr70, in months), changes in SWE (dSWE, in ratio, or %/100) and SFE:Pr ratios (SFEtoP, in ratio):

dfs_monr70_5mm.2050.rcp85.tif
dswe_ONDJFM_5mm.2050s.r45.tif
dswe_ONDJFM_5mm.2050s.r85.tif
dswe_ONDJFM_5mm.2080s.r45.tif
dswe_ONDJFM_5mm.2080s.r85.tif
SFEtoP_ONDJFM_5mm.2050s.r45.tif
SFEtoP_ONDJFM_5mm.2050s.r85.tif
SFEtoP_ONDJFM_5mm.2080s.r45.tif
SFEtoP_ONDJFM_5mm.2080s.r85.tif
swe_ONDJFM_hist.1980s.tif

2. Deltas_2km_sources includes ALFRESCO outputs (fires per century, vegetation per century, new spruce in “dFpCen, dVegpCen, NuPi” and​

2.1 dFpCen_dVegpCen_NuPi includes new spruce (NuPi) for basal area 0-5 (BA0t5) and 5-12 (BA5t12), and changes (d) in fires per century (fpcen) and vegetation per century (vgcen):

dfpcen.ccsm4.2100.r45.tif
dfpcen.ccsm4.2100.r85.tif
dfpcen.cgcm3.2100.r45.tif
dfpcen.cgcm3.2100.r85.tif
dvgcen.ccsm4.2100.r45.tif
dvgcen.ccsm4.2100.r85.tif
dvgcen.cgcm3.2100.r45.tif
dvgcen.cgcm3.2100.r85.tif
NuPiBA0t5.ccsm4.2050.tif
NuPiBA0t5.ccsm4.2100.r85.tif
NuPiBA0t5.cgcm3.2050.r85.tif
NuPiBA0t5.cgcm3.2100.r85.tif
NuPiBA5t12.ccsm4.2050.r85.tif
NuPiBA5t12.cgcm3.2050.r85.tif
 
3. Deltas_TerraClimate_sources includes changes (d) in potential evapotranspiration (PET), deficit (DEF), and runoff (Q) outputs:

dDEF_AMJJAS.2C.tif
dDEF_AMJJAS.4C.tif
dDEF_JJA.2C.tif
dDEF_JJA.4C.tif
dPET_AMJJAS.2C.tif
dPET_AMJJAS.4C.tif
dPET_JJA.2C.tif
dPET_JJA.4C.tif
dq_APR.2C.tif
dq_APR.4C.tif
dq_FEB.2C.tif
dq_FEB.4C.tif
dq_MAR.2C.tif
dq_MAR.4C.tif
dq_MAY.2C.tif
dq_MAY.4C.tif 

 
Citations
 
Abatzoglou, J.T., S.Z. Dobrowski, S.A. Parks, K.C. Hegewisch, 2018, Terraclimate, a high-resolution global dataset of monthly climate and climatic water balance from 1958-2015, Scientific Data.
 
Community of Kotlik, Littell, J.S., Fresco, N., Toohey, R.C., and Chase, M., editors. 2020. Looking Forward, Looking Back: Building Resilience Today Community Report. Aleutian Pribilof Islands Association. Kotlik and Fairbanks, AK. 48 pp.
 
Euskirchen, E. S., Timm, K., Breen, A. L., Gray, S., Rupp, T. S., Martin, P., Reynolds, J. H., Sesser, A., Murphy, K., Littell, J. S., Bennett, A., Bolton, W. R., Carman, T., Genet, H., Griffith, B., Kurkowski, T., Lara, M. J., Marchenko, S., Nicolsky, D., … McGuire, A. D. (2020). Co-producing knowledge: the Integrated Ecosystem Model for resource management in Arctic Alaska. Frontiers in Ecology and the Environment, 18(8), 447–455. https://doi.org/https://doi.org/10.1002/fee.2176
 
McAfee, S. A., Walsh, J., & Rupp, T. S. (2014). Statistically downscaled projections of snow/rain partitioning for Alaska. Hydrological Processes, 28(12), 3930–3946. https://doi.org/10.1002/hyp.9934
 
McGuire, A. D., Genet, H., Lyu, Z., Pastick, N., Stackpoole, S., Birdsey, R., D’Amore, D., He, Y., Rupp, T. S., Striegl, R., Wylie, B. K., Zhou, X., Zhuang, Q., & Zhu, Z. (2018). Assessing historical and projected carbon balance of Alaska: A synthesis of results and policy/management implications. Ecological Applications, 28(6), 1396–1412. https://doi.org/10.1002/EAP.1768
 
 
Littell, J. S., McAfee, S. A., & Hayward, G. D. (2018). Alaska Snowpack Response to Climate Change: Statewide Snowfall Equivalent and Snowpack Water Scenarios. Water 2018, Vol. 10, Page 668, 10(5), 668. https://doi.org/10.3390/W10050668
 
Littell, J. S., McKenzie, D., Kerns, B. K., Cushman, S., & Shaw, C. G. (2011). Managing uncertainty in climate-driven ecological models to inform adaptation to climate change. Ecosphere, 2(9), art102. https://doi.org/10.1890/ES11-00114.1
 
Snover, A. K., Mantua, N. J., Littell, J. S., Alexander, M. A., Mcclure, M. M., & Nye, J. (2013). Choosing and Using Climate-Change Scenarios for Ecological-Impact Assessments and Conservation Decisions. Conservation Biology, 27(6), 1147–1157. https://doi.org/10.1111/cobi.12163
 
Walsh, J. E., Bhatt, U. S., Littell, J. S., Leonawicz, M., Lindgren, M., Kurkowski, T. A., Bieniek, P. A., Thoman, R., Gray, S., & Rupp, T. S. (2018). Downscaling of climate model output for Alaskan stakeholders. Environmental Modelling & Software. https://doi.org/10.1016/j.envsoft.2018.03.021
 

Custom geospatial summaries for management / planning / adaptation requests:

 
The data layers above have been used for a series of custom summaries for specific management units in Alaska. As of Feb 2023, there are 18 products for specific areas in “draft” format that use these data layers.
 

Middle Kuskokwim community adaptation (ANTHC): Feb 2020
Copper River Native Association (CRNA): Apr 2020
Tetlin National Wildlife Refuge (USFWS): May 2020
Chugach National Forest (USFS): Oct 2020
Yukon Delta National Wildlife Refuge (USFWS): Mar 2021
Yukon – Koyukuk community adaptation (ANTHC): Mar 2021
Yukon Flats National Wildlife Refuge (USFWS): Apr 2021
Arctic National Wildlife Refuge (USFWS): Apr 2021
Izembek National Wildlife Refuge (USFWS): Jun 2021
Koyukuk, Innoko, Nowitna Wildlife Refuges (USFWS): Aug 2021
Alaska National Park Networks (USNPS): Feb – Aug 2021

SW Network: Feb 2021
SE Network: Feb 2021
Central AK Network: Aug 2021
Arctic Network: Aug 2021


Togiak National Wildlife Refuge (USFWS): Apr 2022
Tongass National forest (USFS): May 2021, Feb 2023
Aleutian Bering Sea Island region (APIA): Oct 2021, Jan 2023
Klukwan/Chilkat upper Lynn Canal (Chilkat Indian Village): Aug 2019, Feb 2023

 
These are not currently hosted in one place – access is variable depending on the governmental home of the requestor.