Digital Earths Global Hackathon Data Catalog

This is the data catalog for the Digital Earths Global Hackathon. The datasets are hosted by different providers but sometimes may also be accessed remotely and thus are available at other places.

To work with this catalog, you need intake version 0.7 (not 2.X). Please have a look at the environment.yaml file provided for the global hackathon for a compatible python environment. In order to use the best way of accessing the data from your (or your codes) location, you'll have to specify your current NODE_ID when opening the dataset, the catalog will do the rest. In case you are not working on any participating compute center, you may still select online and access publicly available data.

import intake
cat = intake.open_catalog("https://digital-earths-global-hackathon.github.io/catalog/catalog.yaml")["NODE_ID"]

In order to open a dataset from the catalog below, use the catalog as follows:

ds = cat["DATASET_ID"](zoom=7).to_dask()

Datasets

CERES_EBAF (zoom=6)

zoom:

0, 1, 2, 3, 4, 5, 6

online

CN
EU
JAPAN
NCAR
NERSC
UK
online

ERA5 (zoom=7)

zoom:

0, 1, 2, 3, 4, 5, 6, 7, 8

online

CN
EU
JAPAN
NCAR
NERSC
UK
online

IR_IMERG (zoom=9)

zoom:

NCEP/CPC 4km Global (60N - 60S) IR and GPM IMERG (precipitation) Dataset

NERSC
UK
online

CN
EU
JAPAN
NCAR
NERSC
UK
online

JRA3Q (zoom=7)

zoom:

0, 1, 2, 3, 4, 5, 6, 7, 8

online

CN
EU
JAPAN
NCAR
NERSC
UK
online

MERRA2 (zoom=7)

zoom:

0, 1, 2, 3, 4, 5, 6, 7

online

CN
EU
JAPAN
NCAR
NERSC
UK
online

casesm2_10km_cumulus (time='PT6H', zoom=9)

time:	'PT1H', 'PT3H', 'PT6H'
zoom:	0, 1, 2, 3, 4, 5, 6, 7, 8, 9

CAS-ESM 10 km simulation

CAS-ESM2 atmosphere-only simulation at 10 km resolution with deep convection. Run for 1 year and 1 month and 11 days starting in 20th Jan. 2020.

Simulation:

Horizontal grid: lat-lon, 1801*3600
Vertical grid: 35 full levels
Initial conditions: From IFS for date 2020-01-20T00:00:00Z
Resolution:
Atmosphere data at HEALPix level 9.
Land data at HEALPix level 9.
2d3h contains selected 2D fields every 3 hours.
2d1h contains selected 2D fields every hour.
3d6h contains selected 3D fields every 6 hours.

Processing

Fields were remapped to HEALPix level 9 using nearest-neighbor interpolation.

casesm2_10km_nocumulus (time='PT6H', zoom=9)

time:	'PT1H', 'PT3H', 'PT6H'
zoom:	0, 1, 2, 3, 4, 5, 6, 7, 8, 9

CAS-ESM 10 km simulation

CAS-ESM2 atmosphere-only simulation at 10 km resolution with deep convection off. Run for 1 year and 1 month and 11 days starting in 20th Jan. 2020.

Simulation:

Horizontal grid: lat-lon, 1801*3600
Vertical grid: 35 full levels
Initial conditions: From IFS for date 2020-01-20T00:00:00Z
Resolution:
Atmosphere data at HEALPix level 9.
Land data at HEALPix level 9.
2d3h contains selected 2D fields every 3 hours.
2d1h contains selected 2D fields every hour.
3d6h contains selected 3D fields every 6 hours.

Processing

Fields were remapped to HEALPix level 9 using nearest-neighbor interpolation.

references

https://doi.org/10.1029/2020MS002210

CN
EU
online

CN
EU
JAPAN
NCAR
NERSC
UK
online

icon_d3hp003 (time='P1D', time_method='mean', zoom=0)

time:	'P1D', 'PT1H', 'PT3H', 'PT6H'
time_method:	'inst', 'mean'
zoom:	0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11

ICON 2.5 km simulation

Atmosphere-only simulation at 2.5 km resolution. Run for 1 year and 2 months starting in 2020.

Visual overview of dataset

Simulation:

Source code: icon-mpim:feature-dy3ha-p-experiments at commit b9c9da5cef4e7faadf13c2fe213de0c990c94bf3
Horizontal grid: R2B10, 83,886,080 cells, nominal resolution dx = ~2.5 km
Vertical grid: 91 half levels, 90 full levels at mid height between upper and lower half level
Initial conditions: From IFS for date 2020-01-01T00:00:00Z
External data:
- SST and SIC as monthly mean fields pre-processed for usage in AMIP simulations
- Well mixed greenhouse gases CO2, CH4, N2O, CFC11, CFC12 as annual means
- O3 simulated or prescribed from monthly and zonal mean fields

Simulation modifications:

2020-05-28T00:00:00Z: timestep atm: 20s -> 15s, ozone: prognostic -> external data, zmaxcloudy: 33km -> 22.5km. zmaxcloudy is the maximum height up to which cloud microphysics is computed.
2020-06-05T00:00:00Z: timestep atm: 15s -> 20s
2020-07-27T00:00:00Z: dt_atm: 20s -> 18s, dynamics substeps: 5 -> 6
2020-08-04T00:00:00Z: dt_atm: 18s -> 20s, dynamics substeps: 6 -> 5

Processing

Fields were remapped from the 2.5 km ICON grid to the Healpix grid with order 12 using the nearest-neighbor interpolation. Lower levels were generated from this Healpix order using conservative aggregation.
Level 12 data were deleted due to the disk space constraints.
2D fields time means are calculated from each simulation timestep values.
3D fields are generated on 30 pressure levels as instantaneous values starting at 6-hourly interval (PT6H). Time means are calculated from the instantaneous values.
Additionally, a 3D field vor on 3 pressure levels is available as instantaneous values starting at PT3H.
Water content of soil layers mrso is available as a 3D field with 5 soil levels starting from PT3H.

references

https://www.icon-model.org/

EU
NERSC
online

CN
EU
JAPAN
NCAR
NERSC
UK
online

icon_d3hp003aug (time='PT15M', time_method='inst', zoom=0)

time:	'PT15M', 'PT3H', 'PT6H'
time_method:	'inst', 'mean'
zoom:	0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11

ICON 2.5 km simulation

Atmosphere-only simulation at 2.5 km resolution. Run for 1 day only. Same initial conditions as in icon_d3hp003.

Processing

PT15M output for a subset (PT1H) of the icon_d3hp003 dataset.
See icon_d3hp003 metadata for more description.

references

https://www.icon-model.org/

EU
online

CN
EU
JAPAN
NCAR
NERSC
UK
online

icon_d3hp003feb (time='PT15M', time_method='inst', zoom=0)

time:	'PT15M', 'PT3H', 'PT6H'
time_method:	'inst', 'mean'
zoom:	0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11

ICON 2.5 km simulation

Atmosphere-only simulation at 2.5 km resolution. Run for 1 day only. Same initial conditions as in icon_d3hp003.

Processing

PT15M output for a subset (PT1H) of the icon_d3hp003 dataset.
See icon_d3hp003 metadata for more description.

references

https://www.icon-model.org/

EU
online

CN
EU
JAPAN
NCAR
NERSC
UK
online

icon_ngc4008 (time='P1D', zoom=0)

time:	'P1D', 'PT15M', 'PT3H'
zoom:	0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10

ICON nextGEMS Cycle 4 simulation

Coupled atmosphere-ocean simulation at 5km resolution in the ocean and 10km in the atmosphere. Run for 30 years starting in 2020.

Resolutions

Ocean data is available at healpix levels 0-10
Atmosphere data at levels 0-9.
PT15M contains selected 2D fields,
PT3H adds atmospheric 3D fields and more 2D fields of atmosphere and ocean
P1D adds ocean 3D fields.

Processing

Fields were remapped to HEALPix level 10 using nearest-neighbor interpolation.
Lower levels were generated from this using conservative aggregation.

references

https://easy.gems.dkrz.de/DYAMOND/NextGEMS/prefinal.html#icon

EU
online

CN
EU
JAPAN
NCAR
NERSC
UK
online

ifs_tco2559_rcbmf (time='PT1H', zoom=7)

time:	'P1D', 'P1M', 'PT1H'
zoom:	7, 10

IFS 4.4km simulation (RCBMF), original IFS variables

Atmosphere-only simulation at 4.4 km resolution, with reduced cloud base mass flux. Run for 1 year and 2 months, starting in 2020.

General:

The Integrated Forecasting System (IFS), developed at ECMWF, is a spectral-transform atmospheric model with two-time-level semi-implicit, semi-Lagrangian time stepping (Temperton et al., 2001; Hortal, 2002; Diamantakis and Váňa, 2021). It is here running in atmosphere-only mode, that is without an ocean model. It is still coupled to the interactive land and ocean wave models.
This simulation is at 4.4km in atmosphere-only mode, with SST and sea ice concentration being provided by ESA CCI v3.0.1.

Simulation:

This is one of two new atmosphere-only IFS simulations. This one is at 4.4 km resolution, with the other one (ifs_tco3999_rcbmf) is at 2.8 km. The simulations are started on 01.01.2020 and run for 14 months, ending on 01.03.2021. The two simulations differ only in their resolution, and in slight differences in output treatment and availability. Both simulations use "reduced cloud base mass flux" (RCBMF) which has the shallow convection scheme fully active and the deep convection scheme weakly active (with a strongly reduced cloud base mass flux), in agreement with 9 km nextGEMS production simulations.
The simulation has the same physics as the coupled nextGEMS production simulations.
The simulation is closely aligned with the EERIE project and the AMIP simulations done there (1980-2023, 1 member at 9 km and 10 members at 28 km). Those runs use "deep-on" convection, not "RCBMF". A 6-month twin simulation at 4.4 km exists with "deep-on" settings.

Processing:

This dataset uses default IFS naming of variables. Please refer to the parameter database for more information on the available variables.
Output is written to HEALPix H1024 (zoom level 10, about 6 km resolution), and to H128 (level 7, about 50 km resolution). The remapping to healpix was done with linear interpolation (different to ifs_tco3999_rcbmf and the coupled tco3999 runs, which use conservative interpolation for H128 2D variables). Output is available hourly (PT1H), daily means (P1D) and monthly means (P1M). Daily and monthly means also include some min/max fields, these are compute on the model timestep. PT1H includes instantaneous fields (valid at the given timestep) and accumulated fields (fluxes, e.g. precipitation) accumulated over the 1-hour interval preceding the given timestep (precipitation at 12:00 has been accumulated between 11:00 and 12:00). Daily and monthly means have a timestep at the center of the time interval (e.g. daily means are written at 12:00).

references

ifs_tco3999-ng5_deepoff (dim='2D', time='PT1H', zoom=7)

dim:	'2D', '3D'
time:	'P1M', 'PT1H'
zoom:	7, 11

IFS-FESOM 2.8km simulation (DEEPOFF), original IFS variables

Coupled simulation at 2.8 km resolution, with deep convection parameterization switched off. Run for 1 year and 2 months, starting in 2020. (Online access via EERIE cloud).

General:

The Integrated Forecasting System (IFS), developed at ECMWF, is a spectral-transform atmospheric model with two-time-level semi-implicit, semi-Lagrangian time stepping (Temperton et al., 2001; Hortal, 2002; Diamantakis and Váňa, 2021). It is coupled to other Earth system components (land, waves, ocean, sea ice).
The global hackathon simulations at 2.8km are with the IFS-FESOM configuration (Rackow et al. 2025), coupled to the FESOM2 ocean-sea ice model.

Simulation:

This is one of two new IFS simulations at 2.8 km resolution, coupled to FESOM 5 km. The simulations are started on 01.01.2020 and run for 14 months, ending on 01.03.2021. The two simulations differ in the treatment of deep convection. This simulation (DEEPOFF) only has the shallow convection scheme active and the deep convection scheme switched off. The second simulation (RCBMF; catalog entry ifs_tco3999-ng5_rcbmf with standard IFS variables, and ifs_tco3999-ng5_rcbmf_cf the rechunked version following CF naming) has the shallow convection scheme fully active and the deep convection scheme weakly active (with a strongly reduced cloud base mass flux), in agreement with 9 km nextGEMS production simulations.

Processing:

This dataset uses default IFS naming of variables. Please refer to the parameter database for more information on the available variables.
Output is written to HEALPix H2048 (zoom level 11, about 3 km resolution), and to H128 (level 7, about 50 km resolution). The remapping to zoom level 7 used conservative interpolation for the 2D variables and linear interpolation for the 3D variables. Remapping to the high-resolution H2048 (level 11) was also done with linear interpolation. Output is available hourly (PT1H). In addition, we also provide monthly means (P1M), but only for those variables that are accumulated in time (all fluxes, e.g. radiation, precipitation, etc.).

references

online

CN
EU
JAPAN
NCAR
NERSC
UK
online

ifs_tco3999-ng5_rcbmf (dim='2D', time='PT1H', zoom=7)

dim:	'2D', '3D'
time:	'P1M', 'PT1H'
zoom:	7, 11

IFS-FESOM 2.8km simulation (RCBMF), original IFS variables

Coupled simulation at 2.8 km resolution, with reduced cloud base mass flux. Run for 1 year and 2 months, starting in 2020. (Online access via EERIE cloud).

General:

The Integrated Forecasting System (IFS), developed at ECMWF, is a spectral-transform atmospheric model with two-time-level semi-implicit, semi-Lagrangian time stepping (Temperton et al., 2001; Hortal, 2002; Diamantakis and Váňa, 2021). It is coupled to other Earth system components (land, waves, ocean, sea ice).
The global hackathon simulations at 2.8km are with the IFS-FESOM configuration (Rackow et al. 2025), coupled to the FESOM2 ocean-sea ice model.

Simulation:

This is one of two new IFS simulations at 2.8 km resolution, coupled to FESOM 5 km. The simulations are started on 01.01.2020 and run for 14 months, ending on 01.03.2021. The two simulations differ in the treatment of deep convection. This simulation (RCBMF) has the shallow convection scheme fully active and the deep convection scheme weakly active (with a strongly reduced cloud base mass flux), in agreement with 9 km nextGEMS production simulations, while the second simulation (DEEPOFF; catalog entry ifs_tco3999-ng5_deepoff) only has the shallow convection scheme active and the deep convection scheme completely switched off.

Processing:

This dataset uses default IFS naming of variables. Please refer to the parameter database for more information on the available variables.
There is a rechunked version of this dataset available online, which may increase performance for some tasks ( ifs_tco3999-ng5_rcbmf_cf ). Note that this dataset largely uses CF naming.
Output is written to HEALPix H2048 (zoom level 11, about 3 km resolution), and to H128 (level 7, about 50 km resolution). The remapping to zoom level 7 used conservative interpolation for the 2D variables and linear interpolation for the 3D variables. Remapping to the high-resolution H2048 (level 11) was also done with linear interpolation. Output is available hourly (PT1H). In addition, we also provide monthly means (P1M), but only for those variables that are accumulated in time (all fluxes, e.g. radiation, precipitation, etc.).

references

online

CN
EU
JAPAN
NCAR
NERSC
UK
online

ifs_tco3999-ng5_rcbmf_cf (time='PT1H', zoom=7)

time:	'PT1H'
zoom:	7, 11

IFS-FESOM 2.8km simulation (RCBMF), CF conventions + rechunked

Main coupled IFS simulation at 2.8 km resolution, with reduced cloud base mass flux. Run for 1 year and 2 months, starting in 2020. Just available via the online catalog (s3 DKRZ cloud).

General:

The Integrated Forecasting System (IFS), developed at ECMWF, is a spectral-transform atmospheric model with two-time-level semi-implicit, semi-Lagrangian time stepping (Temperton et al., 2001; Hortal, 2002; Diamantakis and Váňa, 2021). It is coupled to other Earth system components (land, waves, ocean, sea ice).
The global hackathon simulations at 2.8km are with the IFS-FESOM configuration (Rackow et al. 2025), coupled to the FESOM2 ocean-sea ice model.

Simulation:

This is one of two new IFS simulations at 2.8 km resolution, coupled to FESOM 5 km. The simulations are started on 01.01.2020 and run for 14 months, ending on 01.03.2021. The two simulations differ in the treatment of deep convection. This main simulation (RCBMF) has the shallow convection scheme fully active and the deep convection scheme weakly active (with a strongly reduced cloud base mass flux), in agreement with 9 km nextGEMS production simulations, while the second simulation (DEEPOFF; catalog entry ifs_tco3999-ng5_deepoff) only has the shallow convection scheme active and the deep convection scheme completely switched off.

Processing:

This main IFS simulation at 2.8 km is provided via the DKRZ cloud (s3), was rechunked, and largely follows CF naming of variables, as in the HK25 data request. You can check the dataset_transformation script here for details how IFS variables were mapped to the CF standard. Metadata may not be fully consistent yet for transformed variables but will be updated step by step (the target was to follow units of the according ICON simulation). There are additional IFS variables provided with their original naming; please refer to the parameter database for more information on those additional variables.
The same simulation without variable transformation/with original IFS variable naming is available under catalog entry ifs_tco3999-ng5_rcbmf. Note that this dataset is not rechunked, so you may experience different performance depending on the task.
Output is written to HEALPix H2048 (zoom level 11, about 3 km resolution), and to H128 (level 7, about 50 km resolution). The remapping to zoom level 7 used conservative interpolation for the 2D variables and linear interpolation for the 3D variables. Remapping to the high-resolution H2048 (level 11) was also done with linear interpolation. Output is available hourly (PT1H). In addition, we also provide monthly means (P1M), but only for those variables that are accumulated in time (all fluxes, e.g. radiation, precipitation, etc.).

references

online

CN
EU
JAPAN
NCAR
NERSC
UK
online

ifs_tco3999_rcbmf (time='PT1H', zoom=7)

time:	'P1M', 'PT1H'
zoom:	7, 11

IFS 2.8km simulation (RCBMF), original IFS variables

Atmosphere-only simulation at 2.8 km resolution, with reduced cloud base mass flux. Run for 1 year and 2 months, starting in 2020.

General:

The Integrated Forecasting System (IFS), developed at ECMWF, is a spectral-transform atmospheric model with two-time-level semi-implicit, semi-Lagrangian time stepping (Temperton et al., 2001; Hortal, 2002; Diamantakis and Váňa, 2021). It is here running in atmosphere-only mode, that is without an ocean model. It is still coupled to the interactive land and ocean wave models.
This simulation is at 2.8km in atmosphere-only mode, with SST and sea ice concentration being provided by ESA CCI v3.0.1.

Simulation:

This is one of two new atmosphere-only IFS simulations. This one is at 2.8 km resolution, with the other one (ifs_tco2559_rcbmf) is at 4.4 km. The simulations are started on 01.01.2020 and run for 14 months, ending on 01.03.2021. The two simulations differ only in their resolution. Both simulations use "reduced cloud base mass flux" (RCBMF) which has the shallow convection scheme fully active and the deep convection scheme weakly active (with a strongly reduced cloud base mass flux), in agreement with 9 km nextGEMS production simulations.
This atmosphere-only simulation is complementary to the 2.8 km coupled IFS-FESOM simulations. It differs from ifs_tco3999-ng5_rcbmf only by the replacement of the ocean model with observed SST forcing.

Processing:

This dataset uses default IFS naming of variables. Please refer to the parameter database for more information on the available variables.
Output is written to HEALPix H2048 (zoom level 11, about 3 km resolution), and to H128 (level 7, about 50 km resolution). The remapping to zoom level 7 used conservative interpolation for the 2D variables and linear interpolation for the 3D variables. Remapping to the high-resolution H2048 (level 11) was also done with linear interpolation. Output is available hourly (PT1H). In addition, we also provide monthly means (P1M), but only for those variables that are accumulated in time (all fluxes, e.g. radiation, precipitation, etc.).

references

mpas_dyamond1 (time='PT3H', zoom=7)

time:	'PT15M', 'PT3H'
zoom:	1, 2, 3, 4, 5, 6, 7, 8, 9, 10

NCAR MPAS 3.75km simulation (partial DYAMOND1)

Atmosphere-land simulation at 3.75km grid spacing. Run from 2016-08-01 to 2016-09-10.

Resolutions

All data is available at HEALPix levels 1-10 for 2D fields

Processing

Fields were remapped to HEALPix level 10 using Delaunay triangulation.
Lower levels were generated from this using coarse-graining.

NCAR

NCAR

mpas_dyamond2 (time='PT3H', zoom=7)

time:	'PT3H'
zoom:	1, 2, 3, 4, 5, 6, 7, 8, 9, 10

NCAR MPAS 3.75km simulation (partial DYAMOND2)

Atmosphere-land simulation at 3.75km grid spacing. Run from 2020-01-20 to 2020-03-01.

Resolutions

All data is available at HEALPix levels 1-10 for 2D fields

Processing

Fields were remapped to HEALPix level 10 using Delaunay triangulation.
Lower levels were generated from this using coarse-graining.

NCAR

NCAR

mpas_dyamond3 (time='PT1H', zoom=7)

time:	'PT1H', 'PT30M', 'PT3H', 'PT6H'
zoom:	1, 2, 3, 4, 5, 6, 7, 8, 9, 10

NCAR MPAS 3.75km simulation (partial DYAMOND3)

Atmosphere-land simulation at 3.75km grid spacing. Run from 2020-01-20 to 2020-03-05.

Resolutions

All data is available at HEALPix levels 1-10 for 2D fields

Processing

Fields were remapped to HEALPix level 10 using Delaunay triangulation.
Lower levels were generated from this using coarse-graining.

NCAR

NCAR

nicam_220m_test (zoom=0)

zoom:

0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15

NICAM 220 m simulation

Atmosphere-only simulation at 220 m resolution. Simulation

Horizontal grid: Glevel-15, 10737418240 control volumes (hexagon/pentagon), average interaval (dx) ~ 220 km
Vertical grid: 78 full levels with the model top of about 50 km

JAPAN
online

CN
EU
JAPAN
NCAR
NERSC
UK
online

nicam_gl11 (time='PT3H', zoom=0)

time:	'PT1H', 'PT3H', 'PT6H'
zoom:	0, 1, 2, 3, 4, 5, 6, 7, 8, 9

NICAM 3.5 km simulation

Atmosphere-only simulation at 3.5 km resolution. One-year data from 10-year simulations starting in 2011.

Simulation

Horizontal grid: Glevel-11, 41,943,042 control volumes (hexagon/pentagon), average interaval (dx) ~ 3.5 km
Vertical grid: 78 full levels with the model top of about 50 km
Initial conditions: From ERA-Interim for date 2011-01-01T00:00:00Z
External data:
- SST and SIC predicted by a slab-ocean model (mixed layer depth 15m), nudged to daily mean fields from OSTIA with a 7-day relaxation time
- Greenhouse gases as daily values, following HighResMIP protocol
- O3 prescribed from monthly fields, following HighResMIP protocol
- Greenhouse gases and O3 With SSP585 scenario after 2015, following HighResMIP protocol

Processing

Fields were remapped from the 3.5 km NICAM to the Healpix grid with conservative remapping.
2d1h fields have instantaneous values except for precipitation (pr).
2d3h fields have mean values calculated from each simulation timestep values.
3d6h fields are generated on 25 pressure levels as instantaneous values starting at 6-hourly interval.

references

https://doi.org/10.1029/2023MS003701

JAPAN
NERSC
online

CN
EU
JAPAN
NCAR
NERSC
UK
online

nicam_gl11_latlon (time='PT3H')

time:

'PT3H', 'PT6H'

NICAM 3.5 km simulation

This is the data from nicam_gl11, but on a lat/lon grid.

JAPAN

JAPAN

scream-dkrz (time='PT1H', zoom=2)

time:	'P1D', 'PT1H', 'PT3H'
zoom:	0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10

SCREAM CESS experiment subset for Hamburg node

Subset of what is available via Globus. Each entry merges existing datasets as follows:

scream_PT1H (Resolution 0-10):

scream2D_hrly_pr_hp*_v7
scream2D_hrly_rlut_hp*_v7 scream_PT3H (Resolution 0-8):
scream2D_ne120_all_hp*_v7
scream3D_ne120_hus_hp*_v7 scream_P1D (Resolution 0-8):
scream_lnd_mrso_hp*_v4
scream_lnd_swe_hp*_v4

references

https://github.com/digital-earths-global-hackathon/hk25/blob/main/content/models/scream.md

online

CN
EU
JAPAN
NCAR
NERSC
UK
online

scream2D_hrly (zoom=2)

zoom:

0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10

SCREAM CESS experiment 2D Hourly instantaneous Fields

2-D Hourly instantaneous Fields from the SCREAM CESS experiment.

Resolutions

Atmosphere data at levels 0-10 for 2D fields.

Processing

Fields were remapped to HEALPix level 10 using Delaunay triangulation.
Lower levels were generated from this using coarse-graining.

references

https://github.com/digital-earths-global-hackathon/hk25/blob/main/content/models/scream.md

NCAR
NERSC

NCAR
NERSC

scream_lnd (zoom=2)

zoom:

0, 1, 2, 3, 4, 5, 6, 7, 8

SCREAM CESS experiment 3hrly Fields from 25km data

2D daily average Fields from the SCREAM CESS experiment (at 3.5km resolution).)

Resolutions

Levels 0-8 for 2D fields.

Processing

Fields were saved as coarse (25km, ne120) data
Then remapped to HEALPix level 8 using Delaunay triangulation.
Lower levels were generated from this using coarse-graining.

references

https://github.com/digital-earths-global-hackathon/hk25/blob/main/content/models/scream.md

NCAR
NERSC

NCAR
NERSC

scream_ne120 (zoom=2)

zoom:

0, 1, 2, 3, 4, 5, 6, 7, 8

SCREAM CESS experiment 3hrly Fields from 25km data

2D and 3D 3hourly aerage Fields from the SCREAM CESS experiment (at 3.5km resolution).)

Resolutions

Levels 0-8 for 2D fields.
Levels 0-8 for 3D fields.

Processing

Fields were saved as coarse (25km, ne120) data
Then remapped to HEALPix level 8 using Delaunay triangulation.
Lower levels were generated from this using coarse-graining.

references

https://github.com/digital-earths-global-hackathon/hk25/blob/main/content/models/scream.md

NCAR
NERSC

NCAR
NERSC

scream_ne120_inst (zoom=2)

zoom:

0, 1, 2, 3, 4, 5, 6, 7, 8

SCREAM CESS experiment 2D 3hrly instantaneous Fields

2-D 3-Hourly instantaneous Fields from the SCREAM CESS experiment.

Resolutions

Atmosphere data at levels 0-8 for 2D fields.

Processing

Fields were saved as coarse (25km, ne120) data
Fields were remapped to HEALPix level 8 using nearest-neighbor interpolation.
Lower levels were generated from this using coarse-graining.

references

https://github.com/digital-earths-global-hackathon/hk25/blob/main/content/models/scream.md

NERSC

NERSC

tracking (zoom=9)

zoom:

6, 7, 8, 9, 10

Remapped Pixel-level feature tracking (MCS Mask) data to HEALPix grid

2D hourly feature tracking data from the SCREAM CESS experiment.

Resolutions

Levels 10-6 for 2D fields.

NERSC

NERSC

tracking-d3hp003

EU
online

CN
EU
JAPAN
NCAR
NERSC
UK
online

um_Africa_km4p4_RAL3P3_n1280_GAL9_nest (time='PT1H', zoom=8)

time:	'PT1H', 'PT3H'
zoom:	0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10

4.4km resolution RAL3 Africa LAM nested in 10km GAL9 global driving model

Atmosphere-only RAL3 simulation of the Africa LAM at 4.4 km resolution. Run for 1 year and 2 months starting in 2020.

Simulation:

Source code: Regional Nesting Suite (RNS)
Horizontal grid: 4.4km resolution global lat-lon grid
Vertical grid: L90_40km - 90 levels, model top at 40km
Initial conditions: From operational model for date 2020-01-20T00:00:00Z

UK
online

CN
EU
JAPAN
NCAR
NERSC
UK
online

um_CTC_km4p4_RAL3P3_n1280_GAL9_nest (time='PT1H', zoom=8)

time:	'PT1H', 'PT3H'
zoom:	0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10

4.4km resolution RAL3 CTC nested in 10km GAL9 global driving model

Atmosphere-only RAL3 simulation of the Cyclic Tropical Channel (CTC) at 4.4 km resolution. Run for 1 year and 2 months starting in 2020.

Simulation:

Source code: Regional Nesting Suite (RNS)
Horizontal grid: 4.4km resolution global lat-lon grid
Vertical grid: L90_40km - 90 levels, model top at 40km
Initial conditions: From operational model for date 2020-01-20T00:00:00Z

UK
online

CN
EU
JAPAN
NCAR
NERSC
UK
online

um_SAmer_km4p4_RAL3P3_n1280_GAL9_nest (time='PT1H', zoom=8)

time:	'PT1H', 'PT3H'
zoom:	0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10

4.4km resolution RAL3 SAmer LAM nested in 10km GAL9 global driving model

Atmosphere-only RAL3 simulation of the South America (SAmer) LAM at 4.4 km resolution. Run for 1 year and 2 months starting in 2020.

Simulation:

Source code: Regional Nesting Suite (RNS)
Horizontal grid: 4.4km resolution global lat-lon grid
Vertical grid: L90_40km - 90 levels, model top at 40km
Initial conditions: From operational model for date 2020-01-20T00:00:00Z

UK
online

CN
EU
JAPAN
NCAR
NERSC
UK
online

um_SEA_km4p4_RAL3P3_n1280_GAL9_nest (time='PT1H', zoom=8)

time:	'PT1H', 'PT3H'
zoom:	0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10

4.4km resolution RAL3 SEA LAM nested in 10km GAL9 global driving model

Atmosphere-only RAL3 simulation of South East Asia (SEA) LAM at 4.4 km resolution. Run for 1 year and 2 months starting in 2020.

Simulation:

Source code: Regional Nesting Suite (RNS)
Horizontal grid: 4.4km resolution global lat-lon grid
Vertical grid: L90_40km - 90 levels, model top at 40km
Initial conditions: From operational model for date 2020-01-20T00:00:00Z

UK
online

CN
EU
JAPAN
NCAR
NERSC
UK
online

um_glm_n1280_CoMA9_TBv1p2 (time='PT1H', zoom=8)

time:	'PT1H', 'PT3H'
zoom:	0, 1, 2, 3, 4, 5, 6, 7, 8, 9

CoMorph9 (CoMA9) 10km simulation - global

Atmosphere-only CoMorph9 simulation at 10 km resolution. Run for 1 year and 2 months starting in 2020.

Simulation:

Source code: Regional Nesting Suite (RNS)
Horizontal grid: 10km resolution global lat-lon grid
Vertical grid: L85_80km - 85 levels, model top at 80km
Initial conditions: From operational model for date 2020-01-20T00:00:00Z

UK
online

CN
EU
JAPAN
NCAR
NERSC
UK
online

um_glm_n1280_GAL9 (time='PT1H', zoom=8)

time:	'PT1H', 'PT3H'
zoom:	0, 1, 2, 3, 4, 5, 6, 7, 8, 9

GAL9 10 km simulation - global driving model

Atmosphere-only simulation at 10 km resolution. Run for 1 year and 2 months starting in 2020.

Simulation:

Source code: Regional Nesting Suite (RNS)
Horizontal grid: 10km resolution global lat-lon grid
Vertical grid: L85_80km - 85 levels, model top at 80km
Initial conditions: From operational model for date 2020-01-20T00:00:00Z

UK
online

CN
EU
JAPAN
NCAR
NERSC
UK
online

um_glm_n2560_RAL3p3 (time='PT1H', zoom=8)

time:	'PT1H', 'PT3H'
zoom:	0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10

RAL3.3 5 km simulation

Atmosphere-only simulation at 5 km resolution. Run for 1 year and 2 months starting in 2020.

Simulation:

Source code: Regional Nesting Suite (RNS)
Horizontal grid: 5km resolution global lat-lon grid
Vertical grid: L85_80km - 85 levels, model top at 80km
Initial conditions: From operational model for date 2020-01-20T00:00:00Z

references

https://doi.org/10.5194/gmd-2024-201

UK
online

CN
EU
JAPAN
NCAR
NERSC
UK
online

wrf_conus (zoom=8)

zoom:

7, 8, 9, 10

WRF Coterminus US Experiment (CONUS404)

2D hourly average Fields from the WRF (Weather and Research Forecast) CONUS404 experiment, i.e. regional down-scaled WRF model simulation at 4km grid spacing. Simulations are run from 1979-10-01 to 2022-09-30, but the HEALPix subdataset is available from 2000-10-01 to 2020-09-30.

Resolutions

Seven 2D variables are available at HEALPix levels 7-10.

Processing

Fields were extracted from raw data
Then remapped to HEALPix level 10 using Delaunay triangulation.
Lower levels were generated from this using coarse-graining.

references

https://rda.ucar.edu/datasets/d559000/

NCAR
NERSC

NCAR
NERSC

wrf_samerica (zoom=8)

zoom:

7, 8, 9, 10

WRF South America Experiment

2D hourly average Fields from the WRF South America experiment. Simulations are 20-40 years long, but only 2020 stored here.

Resolutions

Levels 0-10 for 2D fields.

Processing

Fields were sextracted from raw data
Then remapped to HEALPix level 10 using Delaunay triangulation.
Lower levels were generated from this using coarse-graining.

references

https://rda.ucar.edu/datasets/d616000

NCAR
NERSC

NCAR
NERSC