Library of Regionally-Refined Model (RRM) Grids

Owned by Peter Caldwell
Last updated: Mar 15, 2024 by Hsi-Yen Ma
6 min read

This page collects well-tested RRM configurations produced for E3SM which others might want to use.

Adding a New Grid to a Branch

General instructions for creating new grids are at Running E3SM on New Atmosphere Grids , which also include specific instructions for RRM.

Atmosphere grids have a grid quality metric which measure the element distortion (see Running E3SM on New Atmosphere Grids). A high quality RRM grid with smooth transition region will have a value <= 4.

New grids require many modified input files, most crucially is the initial condition files. Ideally, these files should be created from a a very long run (several decades) to allow full adjustment of things like soil moisture and ozone. spun-up land conditions can be generated efficiently using the ICRUELM compset.

Creating a Pull Request to Add a New Grid to the Master Branch

The steps below are for making grids available to other users, which is especially important if your RRM grid was used in a publication. Do not create a PR if you have not thoroughly tested your new RRM grid.

  1. Add a row for your new grid in the appropriate table below

  2. Add the grid files to inputdata/share/meshes following the directions at https://acme-climate.atlassian.net/wiki/spaces/ED/pages/707002387.

  3. Create a pull request making your grid available. The changes should be limited to what is needed to support the new grid as outlined below.

  4. Modify cime_config/config_grids.xml to define the new grid

  5. Modify cime_config/config_grids.xml to specify domain file paths

  6. Modify cime_config/config_grids.xml to specify mapping file paths (only for bi- and tri-grid)

  7. Modify components/eam/bld/config_files/horiz_grid.xml to specify the number of columns

  8. Modify components/eam/bld/namelist_files/namelist_defaults_eam.xml to specify namelist defaults specific to the new grid. Typically values that need new defaults are:

    1. ncdata - the atmospheric initial condition

    2. bnd_topo - topography

    3. drydep_srf_file surface dry deposition of aerosols (MAM)

    4. mesh_file - grid definition file in exodus format

    5. se_ne - set to zero for RRM

    6. dtime - dynamics/physics coupling interval

  9. For certain configurations, especially the older preqx dycor option, you may need to modify the following parameters in components/eam/bld/namelist_files/namelist_defaults_eam.xml

    1. se_tstep

    2. nu

    3. nu_top

    4. nu_div

    5. hypervis_scaling

    6. hypervis_subcycle_tom

    7. hypervis_subcycle

  10. Modify components/clm/bld/namelist_files/namelist_defaults_clm4_5.xml to specify land input files needed for the new grid (may not be needed for scenarios that do modify the land grid)

    1. finidat - land model initial condition

    2. fsurdat - input data describing surface types

  11. Optional - modify cime/config/e3sm/allactive/config_pesall.xml to add a new PE layout for your grid

  12. Optional - modify cime_config/tests.py to add a test for the new grid. RRM tests should only be added to specific test suites, such as e3sm_rrm.

Quasi-Uniform cubed-sphere grids (for reference)

Grid

Grid metrics

Grid

Grid metrics

1 degree (NE=30)

Max Dinv-based element distortion:     1.73
area based dx: min,max=95 km, 111 km

1/4 degree (NE=120)

Max Dinv-based element distortion:     1.73
area based dx: min,max=23 km, 28 km

3.25 km (NE=1024)

Max Dinv-based element distortion:     1.73
area based dx: min,max=2.74 km, 3.26 km

 

Conventional High-Resolution Grids

Grid Name

Target Fine Resolution

Target Coarse Resolution

Goals in mesh creation

Link to mesh graphic

Grid metrics

Validation

Contact Person

Grid Name

Target Fine Resolution

Target Coarse Resolution

Goals in mesh creation

Link to mesh graphic

Grid metrics

Validation

Contact Person

v2 North American RRM

northamericax4v1.g

atm/land: 25 km (ne120)
ocean/ice: 14 km

atm/land: 100 km (ne30)
ocean/ice: 60 km

 

See figures in the paper

Max Dinv-based elem distortion: 3.76
area based dx:
min=17 km
max=117 km

Described in Tang et al., (GMDD 2022)

@Qi Tang

v2 Antarctic RRM

antarcticax4v1.g

atm/land: 25 km (ne120)

atm/land: 100 km (ne30)

 

https://acme-climate.atlassian.net/wiki/spaces/ED/pages/934314314

Max Dinv-based elem distortion: 3.25
area based dx:
min=21 km
max=111 km

TBD

@Qi Tang@Wuyin Lin

CONUS Extremes-Optimized RRM (HyperFACETS)

ne0np4natlanticref.ne30x4.g (???)

Not on E3SM inputdata server

atm/land: 25 km (ne120)

atm/land: 100 km (ne30)

28km over CONUS and Atlantic TC region; 55km over eastern Pacific to better capture ARs

 

elem distortion: 3.35
area based dx:
min=23 km
max=113 km

TBD

@Paul Ullrich

arcticx4v1pg2

atm/land: 25 km (ne120)

atm/land: 100 km (ne30)

 

 

elem distortion: 4.26
area based dx:
min=23 km
max=125 km

TBD

@Ben Hillman

Storm-Resolving Grids

Grid Name

Target Fine Resolution

Target Coarse Resolution

Goals in mesh creation

Link to mesh graphic

grid quality metric

Validation

Contact Person

Grid Name

Target Fine Resolution

Target Coarse Resolution

Goals in mesh creation

Link to mesh graphic

grid quality metric

Validation

Contact Person

CONUS ref(100W,40N)

CONUS_ne32x4_lon100W_lat40N.g
CONUS_ne32x8_lon100W_lat40N.g
CONUS_ne32x16_lon100W_lat40N.g
CONUS_ne32x32_lon100W_lat40N.g

(no on E3SM inputdata server)

28 km (ne128), 14 km (ne256), 7km (ne512), 3.5 km (ne1024)

111km (ne32), 28 km (ne128)

Refinement region with complete coverage of CONUS using several coarse and fine resolutions

 

Max Dinv elem distortion: 3.00, 3.00, 3.01, 3.02

area based dx:
min=24, 12, 6, 3 km
max=105, 106,106,106 km

TBD

@Paul Ullrich

2012 North American Derecho

derecho_ne512.g
derecho_ne1024.g
derecho_ne2048.g

(not on E3SM inputdata server)

7 km (ne512), 3.5 km (ne1024), 1.75 km (ne2048)

100 km (ne32)

Include the region important for the 2012 North American Derecho

 

Max Dinv elem distortion: 2.89, 2.89, 2.97

area based dx:
min=6.4, 3.2, 1.6 km
max=28, 28, 28 km

Liu et al. (2023; under review)

@Paul Ullrich

California

CA_ne32_x32_v1.g
CA_ne32_x32_v2.g
CA_ne32_x32_v3.g
CA_ne32_x64_v1.g
CA_ne32_x128_v1.g

(not on E3SM inputdata server)

3.25 km (ne1024)

100 km (ne32)

High resolution over CA to capture topographic effects of precipitation, but small enough HR domain so that climate length simulations can be performed. Note that other experimental variants of this grid exist (contact @Peter Bogenschutz for details).

 

Max Dinv-based element distortion:     3.02, 3.02, 5.45, 3.04, 3.05

area based dx:
min=3.3, 3.2, 2.4, 1.6, 0.8 km
max=113, 113, 113, 113, 113 km

Zhang et al. (2023; in prep); Bogenschutz et al. (2023; in prep)
Tang et al. (2023; in prep)

@Peter Bogenschutz@Qi Tang

Jishi Zhang (zhang73@llnl.gov)

Amazon ref(60W,5S)

(not on E3SM inputdata server)

14 km (ne256), 7 km (ne512), 3.5 km (ne1024)

28 km (ne128)

Large high resolution region over Amazon

 

 

TBD

@Paul Ullrich

ARM GoAmazon

GoAmazon_ne32_x32_v1.g

(not on E3SM inputdata server)

3.25 km (ne1024)

100 km (ne32)

For ARM GoAmazon field campaign

 

Max Dinv-based element distortion:     2.88
area based dx:
min=3.2 km
max=109 km

TBD

@Hsi-Yen Ma @Yunyan Zhang

ARM TRACER

3.25 km (ne1024)

100 km (ne32)

For ARM TRACER field campaign

 

Max Dinv-based element distortion:     3.06
area based dx:
min=3.2 km
max=109 km

TBD

@Hsi-Yen Ma @Yunyan Zhang

ARM COMBLE

3.25 km (ne1024)

100 km (ne32)

For ARM COMBLE field campaign

 

Max Dinv-based element distortion:     2.98
area based dx:
min=3.2 km
max=109 km

TBD

@Hsi-Yen Ma @Yunyan Zhang

ARM CACTI

3.25 km (ne1024)

100 km (ne32)

For ARM CACTI field campaign

 

Max Dinv-based element distortion:     2.98
area based dx:
min=3.2 km
max=109 km

TBD

@Hsi-Yen Ma @Yunyan Zhang

ARM MAGIC

3.25 km (ne1024)

100 km (ne32)

For ARM MAGIC field campaign

 

Max Dinv-based element distortion:     3.05
area based dx:
min=3.2 km
max=109 km

TBD

@Hsi-Yen Ma @Yunyan Zhang

ARM EPCAPE

3.25 km (ne1024)

100 km (ne32)

For ARM EPCAPE field campaign

 

Max Dinv-based element distortion:     2.88
area based dx:
min=3.2 km
max=109 km

TBD

@Hsi-Yen Ma @Yunyan Zhang

ARM CONUS

3.25km (ne1024)

100 km (ne32)

For SGP, SEUS, TRACER

 

Max Dinv-based element distortion:     2.98
area based dx:
min=3.2 km
max=109 km

TBD

@Hsi-Yen Ma @Yunyan Zhang

Indian Ocean, Maritime Continent

3.25km (ne1024)

100 km (ne32)

For MJO study

 

Max Dinv-based element distortion:     3.40
area based dx:
min=3.2 km
max=109 km

TBD

@Hsi-Yen Ma