MERRA-2 climate reanalysis over the Greenland Ice Sheet subsetted for forcing the Community Firn Model

This dataset is a subset of MERRA-2 climate reanalysis data over the Greenland Ice Sheet (GrIS) created to provide forcing for the Community Firn Model (CFM; https://doi.org/10.5281/zenodo.3585884). It comprises five (zip) zarr stores, each containing reanalysis data for a decade collated from several MERRA-2 data collections. The spatial grid is the native MERRA-2 grid (0.5° lat x 0.625° lon) but only includes data between 55° and 90° latitude and -80° and -10° longitude. The data is temporally resampled to daily resolution.

The files contain variables from four different collections, listed below. The variables are all in the same zarr store, regardless of collection they are taken from (as opposed to the native MERRA-2 formatting, which has different variables in different files.)

tavg1_2d_rad_Nx: (doi:10.5067/Q9QMY5PBNV1T)

- 'ALBEDO','LWGAB','LWGEM','SWGDN','SWGNT','TS','EMIS'

tavg1_2d_int_Nx: (doi:10.5067/Q5GVUVUIVGO7)

- 'PRECCU','PRECLS','PRECSN'

tavg1_2d_flx_Nx: (doi:10.5067/7MCPBJ41Y0K6)

- 'EVAP','HFLUX','EFLUX','GHTSKIN','HLML'

tavg1_2d_slv_Nx: (doi:10.5067/VJAFPLI1CSIV)

- 'T2M'

tavg3_2d_glc_Nx: (doi:10.5067/9ETB4TT5J6US)

- 'RUNOFF','ASNOW_GL','SNOMAS_GL'

As MERRA-2 does not provide surface melt as an output variable, this dataset contains an estimate of the surface meltwater generated (variable SMELT). This is calucated using the radiation fluxes from MERRA-2 assuming a top layer that is 8 cm thick with a density of 400 kg/m3. With this assumption and using the turbulant fluxes from MERRA-2, the SEB calculation is a 4th-order polynomial that can be solved directly. This caculation provided both the melt flux and a surface temperature estimation (TScalc), which is slightly different than TS from the tavg1_2d_rad_Nx collection.

There is also an variable called "EMIS_eff" (effective emissivity), which calculated by:

Although this is a crude estimate, it is helpful in a SEB calculation (as it allows for an estimate of how much longwave radiation is reaching the surface given a 2-m temperature).

There are five zarr stores: one for each decade, 1980-01-01 to 2025-06-25. They are subsetted in time this way to keep file sizes reasonable. Note that these are zip zarr stores, which are effectively uncompressed zips of directories. These can be read directly by xarray (i.e., you do not need to unzip them to use them).

The CFM repository includes an example script (run_CFM_example.py) and notebook (run_CFM_example_notebook.ipynb) that are configured to use these zarr stores for model runs. Note that you will need to edit file paths in those scripts to ensure that they point to the zarr stores on your system.

The script used to create these zarr stores is presently hosted here: https://github.com/maximusjstevens/ATL_masschange/blob/main/MERRA2_subsetter_array.py

CITATION INFORMATION:

Any use of these data for publication should include reference to the original MERRA-2 data and follow NASA GMAO citation recommendations, which can be found here: https://gs6101-gmao.gsfc.nasa.gov/gmao-products/merra-2/citing-merra-2-data_merra-2/. The DOIs for each collection are listed above for convenience. The MERRA-2 file specifications, which include details about the variables can be found here: https://gmao.gsfc.nasa.gov/pubs/docs/Bosilovich785.pdf.