source_id registration of E3SM-1-0

[mccoy20]

label = E3SM 1.0 label_extended = E3SM 1.0 (Energy Exascale Earth System Model) source_id = E3SM-1-0 institution_id = E3SM-Project release_year = 2018 activity_participation = [CMIP]

aerosol: description = MAM4 with resuspension, marine organics, and secondary organics (same grid as atmos ) nominal_resolution = 100 km atmos: description = EAM (v1.0, ne30L72 spectral-element native atmosphere grid; 256 x 129 longitude/latitude regridded; 72 levels; top level 0.1 hPa) nominal_resolution = 100 km atmosChem: description = Troposphere specified oxidants for aerosols. Stratosphere linearized interactive ozone (LINOZ v2) (same grid as atmos) nominal_resolution = 100 km land: description = ELM (v1.0, ne30 satellite phenology mode), MOSART (v1.0, 0.5 degree latitude/longitude grid) nominal_resolution = 100 km landIce: description = none nominal_resolution = none ocean: description = MPAS-Ocean (v6.0, oEC60to30 mesh, variable resolution 60 km to 30km , 60 levels; top grid cell 0-510 m) nominal_resolution = 50 km ocnBgchem: description = none nominal_resolution = none seaIce: description = MPAS-Seaice (v6.0, same grid as ocean)

nominal_resolution = 50 km

[durack1]

@mccoy20 I have a couple of clarifications for the details above:

atmos:
description = EAM (v1.0, ne30L72 spectral-element native atmosphere grid; 256 x 129 longitude/latitude regridded; 72 levels; top level 0.1 hPa)

Is the regridded indicating that this is not the model origin/native grid? If yes, we're registering the native configurations, so please amend this information above

ocean:
description = MPAS-Ocean (v6.0, oEC60to30 mesh, variable resolution 60 km to 30km , 60 levels; top grid cell 0-510 m)

I'm guessing the top grid cell is 0-5 m rather than 510, and I'd also query the oEC60to30 mesh entry, can a more descriptive entry be provided?

[mccoy20]

Here are the modifications

atmos: description = EAM (v1.0, ne30L72 spectral-element grid; 72 levels; top level 0.1 hPa)

ocean: description = MPAS-Ocean (v6.0, oEC60to30 mesh, variable resolution 60 km to 30km , 60 levels; top grid cell 0-10 m)

For the atmosphere - I have deleted the regridding information, the native grid is a "cubed sphere" with 30 spectral elements along one side. The ocean top layer should have been 0-10m.

@durack1 - is it ok now? Thanks!

[durack1]

Thanks @mccoy20, there is one other cube sphere model that is registered, that is th VRESM1-0 (see here)

It has the atmos entry:

VCAM-1.0 (C192; 192 x 192 x 6 longitude/latitude/cubeface; 35 levels; top level 35km)

Would something similar work?

[mccoy20]

@taylor13 @golaz -- how should we describe E3SM atmosphere spectral element ne30 grid? - thanks!

[golaz]

I believe the technical name for our low-res spectral element cubed sphere grid is ne30np4. I don't know whether this is a generally accepted naming convention or whether we should have a notation similar to the C192 above.

@mt5555 might have an opinion as well.

[mccoy20]

Thanks Chris -- I tagged the wrong Taylor -- sorry Karl, I meant to tag Mark @mt5555

[mt5555]

The notation "ne30np4" is a CIME thing and meaningless to most people, so I'm not sure if we should use it or not. the FV3 convention of "C192" is also not standard, so we probably shouldn't adopt that either.

As for the grid, it should be "90 x 90 x 6".

It's confusing because is a 30 x 30 x 6 grid of elements, and each element contains a 4x4 grid, but the points on the element boundaries are shared, so in terms of points / degrees of freedom, the grid is 90 x 90 x 6.

The points do align with latitudes, but not longitudes, so I dont think we should use longitude/latitude/cubeface phrase either.

How about we go the MPAS route and just say:

Cubed sphere finite element mesh with 5400 elements. Each element contains approximately 9 degrees of freedom. Quasi-uniform 1km resolution.

[mccoy20]

Here is the modified text

atmos: description = EAM (v1.0, ne30np4L72 spectral-element grid; 90 x 90 x 6 longitude/latitude/cubeface; 72 levels; top level 0.1 hPa)

[durack1]

@mccoy20 great, this looks good, I'll get it registered later today

[mt5555]

How about:

description = EAM (v1.0, cubed sphere spectral-element grid; 5400 elements with p=3; 1 km average grid spacing; 72 levels; top level 0.1 hPa)

The "p=3" notation is standard in the finite element community, representing polynomials up to degree 3.

[mt5555]

I recommend we avoid the "ne30np4" notation. I dont think it should be used in papers either. It's a CIME naming convention.

[durack1]

@mccoy20 @mt5555 there's no limit to the detail we include, so how about:

description = EAM (v1.0, cubed sphere spectral-element grid; 5400 elements with p=3; 1 km average grid spacing; 90 x 90 x 6 longitude/latitude/cubeface; 72 levels; top level 0.1 hPa)

[mccoy20]

Looks good to me, thanks!

[mt5555]

And the ELM grid should be updated to match.

[golaz]

@mt5555: I think you meant "1 deg average grid spacing" instead of "1 km".

[mccoy20]

OK then lets try agin atmos: description = EAM (v1.0, cubed sphere spectral-element grid; 5400 elements with p=3; 1 deg average grid spacing; 90 x 90 x 6 longitude/latitude/cubeface; 72 levels; top level 0.1 hPa)

land: description = ELM (v1.0, cubed sphere spectral-element grid; 5400 elements with p=3; 1 deg average grid spacing; 90 x 90 x 6 longitude/latitude/cubeface; satellite phenology mode), MOSART (v1.0, 0.5 degree latitude/longitude grid)

[mt5555]

oops! yes, 1 deg!

[durack1]

@mccoy20 as I'm starting to pull this across, it seems we're still missing the XX x XX longitude/latitude info for the ocean, if it's a standard grid, can we add that in? For ref, there are other models that have refined tropical resolutions, so higher res tropical, and lower res high latitudes that still have XX x XX denoted in CMIP6_source_id.html

FYI, I've already merged the above, but would like to correct the ocean description if we can

[mccoy20]

The ocean uses unstructured mesh Spherical Centroidal Voronoi Tessellations (SVTs) and it is variable resolution, so can’t give you xx x xx in lon/lat.

[durack1]

@mccoy20 maybe we should add the SVT information into the description so this is more clear. The total element count could also be useful, this has been done for AWI-CM-1-1-HR see CMIP6_source_id.html which details:

FESOM 1.4 (unstructured grid in the horizontal with 1306775 wet nodes; 46 levels; top grid cell 0-5 m)

[mccoy20]

OK, we can add this info, how about

ocean: description = MPAS-Ocean (v6.0, oEC60to30 unstructured SVTs mesh with 235160 cells and 714274 edges, variable resolution 60 km to 30km , 60 levels; top grid cell 0-510 m)

[durack1]

@mccoy20 that change is now done, and I think we're good - final info is visible at CMIP6_source_id.html