The dataset in this record is the output of the CCSM4 (Community Climate System Model version 4). Please see the collection reference below for outputs from the other models described here.
These data are the
outputs of three general circulation climate models (GCMs), CCSM4, MRI-CGCM3,
and IPSL-CM5A-LR for the period 1950-2100. Runs of each
GCM were carried out as part of the fifth phase of the Coupled Model
Intercomparison Project. Future runs were forced with the RCP 8.5 emissions
scenario. They were downscaled to
a one km spatial resolution using a quantile matching approach. The three GCMs were
chosen because they were shown to recreate climate well in Alaska during the
last few decades and because they span the range of potential conditions during
the 21st century as projected by all climate models included in the
IPCC AR5. Variables include daily minimum and maximum Temperature (°C), daily
sum of precipitation (mm), daily sum of shortwave radiation (Mj m-2),
and mean VPD (kPa).
This dataset includes the following files. The gridded netCDFs are provided as compressed .tar.gz files. Extensive metadata is embedded within each netCDF.
CCSM4-hist-prcp.tar.gz : Daily precipitation (mm) for
1950-2005 from the CCSM4 GCM.
CCSM4-future-prcp.tar.gz : Daily precipitation (mm) for
2006-2100 from the CCSM4 GCM.
CCSM4-hist-srad.tar.gz : Daily shortwave radiation (mJ m-2)
for 1950-2005 from the CCSM4 GCM.
CCSM4-future-srad.tar.gz : Daily shortwave solar radiation (mJ
m-2) for 2006-2100 from the CCSM4 GCM.
CCSM4-hist-tmax.tar.gz : Daily maximum temperature (deg C)
for 1950-2005 from the CCSM4 GCM.
CCSM4-future-tmax.tar.gz : Daily maximum temperature (deg
C)for 2006-2100 from the CCSM4 GCM.
CCSM4-hist-tmin.tar.gz : Daily minimum temperature (deg C)
for 1950-2005 from the CCSM4 GCM.
CCSM4-future-tmin.tar.gz : Daily minimum temperature (deg
C)for 2006-2100 from the CCSM4 GCM.
CCSM4-hist-vap.tar.gz : Daily vapor pressure (kPa) for
1950-2005 from the CCSM4 GCM.
CCSM4-future-vap.tar.gz : Daily vapor pressure (kPa) for
2006-2100 from the CCSM4 GCM.
Geographic coordinates
Lower right: 63.0998°, -142.5507° ; Lower left: 63.0692°, -158.2863° ;Upper left: 63.0692°, -158.2863°Time period
Begin date: January 1, 2019
End date: September 9, 2019Methodology
The three earth system models considered were the MRI-CGCM3, IPSL-CM5A-LR, and CCSM4, with geographic resolutions (latitude x longitude) of 1.1215° x 1.125, 1.8947° x 3.75°, and 0.9424° x 1.25°, respectively. The model simulations were carried out as part of the 5th phase of the Coupled Model Intercomparison Project (CMIP5; Taylor 2012) and a single simulation was used from each model. The scenarios considered were a Historical (1851–2005) scenario and a future scenario (2006–2100) in which the representative concentration pathway 8.5 (RCP8.5) was assumed, representing a high-emissions scenario in which anthropogenic radiative forcing reaches 8.5 W m-2 by 2100 (van Vuuren et al. 2011). For feasibility, we only included data from the Historical scenario prior to 1950. For each model, the future RCP8.5 simulation was appended to the Historical simulation.
We statistically downscaled the modeled daily fields of maximum temperature (Tmax; °C), minimum temperature (Tmin; °C), mean relative humidity (RH; %), mean solar radiation (W m-2), and precipitation total (mm) to a resolution of 1 km across the boreal forest domain of Alaska using the Daymet daily surface weather dataset, version 3, as the target observational dataset (Thornton et al. 2018). The Daymet dataset covers the time period 1980–2017. There are 1,015,189 1-km grid cells within this domain. For each climate model, the downscaling procedure was as follows:
1. Interpolate the model grids to the 1-km Daymet resolution. First, use bilinear interpolation to interpolate the daily modeled fields to a geographic resolution of 0.1° x 0.2° and then, for each 1-km Daymet grid cell, assign the interpolated model data from the corresponding 0.1° x 0.2° grid cell.
2. For each grid cell, we bias correct monthly means of the full modeled time series (1950–2100). For each of the 12 months, adjust the interpolated model grids such that the modeled monthly climatology for 1980–2017 matches that of Daymet. For all variables except for precipitation, monthly means were bias corrected using addition. For precipitation, multiplication was used.
3. For each grid cell we bias corrected daily departures from monthly means using quantile matching. For each month of the Daymet and modeled time series, we calculated daily departures from that month’s mean daily value (for precipitation this was calculated as fraction of the mean). For each of the 12 months, the observed distribution of daily departures was represented by the empirical cumulative distribution function (CDF) of the Daymet data during 1980–2017. Modeled daily departures for each of the 12 months were converted to quantiles based on the 1980–2017 baseline period and these quantile values were then applied to the observed CDF using quantile matching to calculate bias corrected modeled daily departures. We prevented unrealistically large modeled daily departures by, for each of the 12 months, not allowing bias corrected daily departures to stray beyond the observed range of daily departures.
4. Bias corrected daily departures from monthly means were added to the bias corrected monthly means (this was done using multiplication for precipitation, as the daily departures values were in units of fraction of the monthly mean).
5. Daily mean vapor pressure (Ea) was calculated from the bias corrected grids of daily Tmax, Tmin, and RH. Daily maximum and minimum saturation vapor pressure (Es_max and Es_min) were calculated from Tmax and Tmin, respectively, following:
Es = (a0+Ta1 + T2a2 + T3a3 + T4a4 + T5a5 + T6a6) / 10,
where Es is saturation vapor pressure in kPa, T is temperature in degrees Celsius (replaced with Tmax and Tmin for calculation of Es_max and Es_min, respectively), a0 = 6.107799961, a1 = 0.4436518521, a2 = 1.428945805x10-2, a3 = 2.650648471x10-4, a4 = 3.031240396x10-6, a5 = 2.034080948x10-8, and a6 = 6.136820929x10-11 (Lowe 1977). Daily mean Es was estimated as the mean of Es_max and Es_min and daily mean Ea was then calculated as Ea = Es x RH/100.
6. Any negative daily Ea, solar radiation, and precipitation values were set to zero.
7. Additionally, within the bias correction of daily departures for precipitation, we also performed a bias correction procedure on daily precipitation frequency. For each month and grid cell in the bias corrected modeled dataset, fraction of mean precipitation frequency (Fmodel) was calculated as the number of days with precipitation and divided by that month’s baseline daily precipitation frequency during 1980–2017. For each month, the bias-corrected precipitation frequency (Fmodel_bc) was then calculated as Fmodel multiplied by that month’s observed 1980–2017 precipitation frequency according to Daymet. Values of Fmodel_bc were rounded and not allowed to exceed the number of days in the month. When Fmodel_bc was less than Fmodel by X days, then the X days with the lowest non-zero precipitation totals were re-assigned to zero. When Fmodel_bc was more than Fmodel by X days, then the X days with zero precipitation with the highest likelihood of precipitation based on the Daymet record were assigned a precipitation value corresponding to the observed 10th percentile of Daymet-based precipitation totals from that month. For each month, daily precipitation totals were rescaled to match the pre-determined bias-corrected monthly precipitation total bias-correction of daily precipitation frequency.Software
NetCDF documentation located at,
https://www.unidata.ucar.edu/software/netcdf/docs/
CCSM4 documentation located at,
https://www.cesm.ucar.edu/models/ccsm4.0/ccsm/Investigator 1
Park WilliamsInvestigator 2
Winslow HansenInvestigator 3
Richard Seager (PI)Additional references
Lowe, P.R., 1977. An approximating polynomial for the computation of saturation vapor pressure. Journal of the American Meteorological Society 59:100-103. https://doi.org/10.1175/1520-0450(1977)016<0100:AAPFTC>2.0.CO;2
Taylor, K.E. 2012. An overview of CMIP5 and the experimental design. Bulletin of the American Meteorological Society:485-498. http://dx.doi.org/10.1175/BAMS-D-11-00094.1
Thornton, P.E., M.M. Thornton, B.W. Mayer, Y. Wei, R. Devarakonda, R.S. Vose, and R.B. Cook. 2018. Daymet: Daily Surface Weather Data on a 1-km Grid for North America, Version 3. ORNL DAAC, Oak Ridge, Tennessee, USA. https://doi.org/10.3334/ORNLDAAC/1328.
van Vuuren, D. P., J. Edmonds, M. Kainuma, K. Riahi, A. Thomson, K. Hibbard, G. C. Hurtt, T. Kram, V. Krey, J. F. Lamarque, T. Masui, M. Meinshausen, N. Nakicenovic, S. J. Smith, and S. K. Rose. 2011. The representative concentration pathways: An overview. Climatic Change 109:5–31. https://doi.org/10.1007/s10584-011-0148-z
