CMIP5 GCMs

For the downscaling, GCM data is taken from the historical (1950-2005) and future RCP4.5/RCP8.5 (2006-2100) daily outputs of several available CMIP5 models. Though over 40 models participated in CMIP5, only 20 of these models had daily outputs for all of the variables we are interested in for this downscaling project (i.e. minimum/maximum temperature,precipitation, wind, humidity, solar radiation).

The 20 models used for MACA downscaling are:

Model Name Model Country Model Agency Atmosphere Resolution(Lon x Lat) Ensemble Used
bcc-csm1-1 China Beijing Climate Center, China Meteorological Administration 2.8 deg x 2.8 deg r1i1p1
bcc-csm1-1-m China Beijing Climate Center, China Meteorological Administration 1.12 deg x 1.12 deg r1i1p1
BNU-ESM China College of Global Change and Earth System Science, Beijing Normal University, China 2.8 deg x 2.8 deg r1i1p1
CanESM2 Canada Canadian Centre for Climate Modeling and Analysis 2.8 deg x 2.8 deg r1i1p1
CCSM4 USA National Center of Atmospheric Research, USA 1.25 deg x 0.94 deg r6i1p1
CNRM-CM5 France National Centre of Meteorological Research, France 1.4 deg x 1.4 deg r1i1p1
CSIRO-Mk3-6-0 Australia Commonwealth Scientific and Industrial Research Organization/Queensland Climate Change Centre of Excellence, Australia 1.8 deg x 1.8 deg r1i1p1
GFDL-ESM2M USA NOAA Geophysical Fluid Dynamics Laboratory, USA 2.5 deg x 2.0 deg r1i1p1
GFDL-ESM2G USA NOAA Geophysical Fluid Dynamics Laboratory, USA 2.5 deg x 2.0 deg r1i1p1
HadGEM2-ES United Kingdom Met Office Hadley Center, UK 1.88 deg x 1.25 deg r1i1p1
HadGEM2-CC United Kingdom Met Office Hadley Center, UK 1.88 deg x 1.25 deg r1i1p1
inmcm4 Russia Institute for Numerical Mathematics, Russia 2.0 deg x 1.5 deg r1i1p1
IPSL-CM5A-LR France Institut Pierre Simon Laplace, France 3.75 deg x 1.8 deg r1i1p1
IPSL-CM5A-MR France Institut Pierre Simon Laplace, France 2.5 deg x 1.25 deg r1i1p1
IPSL-CM5B-LR France Institut Pierre Simon Laplace, France 2.75 deg x 1.8 deg r1i1p1
MIROC5 Japan Atmosphere and Ocean Research Institute (The University of Tokyo), National Institute for Environmental Studies,and Japan Agency for Marine-Earth Science and Technology 1.4 deg x 1.4 deg r1i1p1
MIROC-ESM Japan Japan Agency for Marine-Earth Science and Technology, Atmosphere and Ocean Research Institute (The University of Tokyo), and National Institute for Environmental Studies 2.8 deg x 2.8 deg r1i1p1
MIROC-ESM-CHEM Japan Japan Agency for Marine-Earth Science and Technology, Atmosphere and Ocean Research Institute (The University of Tokyo), and National Institute for Environmental Studies 2.8 deg x 2.8 deg r1i1p1
MRI-CGCM3 Japan Meteorological Research Institute, Japan 1.1 deg x 1.1 deg r1i1p1
NorESM1-M Norway Norwegian Climate Center, Norway 2.5 deg x 1.9 deg r1i1p1

Quick Differences between the Models

  • ESM in the name indicates that the global climate model is an 'Earth System Model', so that it has the added capability to explicitly represent biogeochemical processes that interact with the physical. climatebiogeochemical processes that interact with the physical climate. These 7 ESMs are: BNU-ESM, CanESM2, GFDL-ESM2G,GFDL-ESM2M,MIROC-ESM,MIROC-ESM-CHEM,NorESM1-M. HadGEM2-ES/CC also seem to be earth system models.
  • GFDL-ESM2M vs GFDL-ESM2G: These models differ mainly in the physical ocean component. In GFDL-ESM2M, pressure-based vertical coordinates are used along the developmental path of GFDL’s Modular Ocean Model version 4.1. In GFDL-ESM2G, an independently developed isopycnal model using the Generalized Ocean Layer Dynamics (GOLD) code base was used.
  • IPSL-CM5A-LR vs IPSL-CM5A-MR: The difference between these models is their atmospheric resolution. LR refers to 'low resolution' which is 3.75° x 1.9° and MR refers to 'mid-resolution' which is 2.5° x 1.25°.
  • IPSL-CM5A vs IPSL-CM5B:
  • HadGEM2-ES vs HadGEM2-CC: The difference between these models is in the included earth system components. ES refers to 'Earth System', while CC refers to 'Carbon Cycle'. HadGEM2 has earth system components that include a terrestrial and ocean carbon cycle and tropospheric chemistry. The major differences of HadGEM2-CC from HadGEM2-ES are the inclusion of a non-orographic gravity wave drag scheme, production of stratospheric water vapour from methane oxidation, and the removal of the UKCA interactive tropospheric chemistry component.
  • MIROC-ESM vs MIROC-ESM-CHEM: MIROC-ESM-CHEM has an atmospheric chemistry (CHEM) component coupled to the earth system model of MIROC-ESM.
  • bcc-csm1-1 vs bcc-csm1-1-m: The major difference between these models is there atmospheric model resolution. bcc-csm1-1-m has moderate resolution of 1.12 deg x 1.12 deg, whereas bcc-csm1-1 has low resolution of 2.8 deg x 2.8 deg.
  • NorESM1-M vs CCSM4: NorESM is based on the Community Climate System Model (CCSM) and Community Earth System Model (CESM) projects. NorESM1-M differs from CCSM in the following aspects: NorESM1-M has its own code for chemistry-aerosol-cloud-radiation interactions in the atmosphere, an isopycnic coordinate ocean model adapted to the isopycnic ocean model framework.
  • MACA Downscaling of CMIP5 GCMs

    Only one ensemble run was downscaled for each model, even if some of the models had multiple ensemble runs. We used ensemble r1i1p1 for all the models except CCSM4 for which we used r6i1p1.

    As HadGEM2-ES/CC are both 360-day models, but 365-day outputs are needed for most impact analyses, we interpolated these to 365-day before downscaling them to create the models HadGEM2-ES365 and HadGEM2-CC365.

    For CCSM4, the wind components uas/vas were taken from the variables ua/va at the lowest pressure level in which there was data. Data from ua/va are avaiable at pressure levels: 100,85,70,50,25,10,5,1 kPa.

    CCSM4 and NorESM1-M did not have relative humidity available at daily timescales. Therefore the MACA data does not have relative humidity (i.e. rhsmin/rhsmax) for either of these models.

    bcc-csm1-1 did not have daily data available for 12/31/2099 for the RCP8.5 scenario only. Therefore the MACA does not have data(i.e all missing values) for this day for any of the variables for RCP8.5.