System name (version)CORe
Date of implementationMarch 2026


Earth system components included in the analysis system

(e.g., ocean, sea-ice, land, etc.)

Atmosphere (Fang and Tallapragada,  2018)

Horizontal resolution of the model, with indication of grid spacing in km

(for the different Earth system component included in the model)

C128, ~70 km

Number of levels in the different Earth system components

(for the different Earth system component included in the model)

64 levels in the atmosphere
Frequency of the outputs3-hourly
Top of the atmospheric model

0.2 Pa

Number of analysis cycle per day4
Earliest start date1 January 1950
Integration time step720 seconds
Length and frequency of the longest forecast6 hour forecasts initialised four times daily from analyses at 00, 06 12,  and 18 UTC
Dataset latency

1 day behind real time (preliminary product)

Additional comments:



3. Analysis system
Data assimilation method

80-member EnKF

Length of the analysis window6 hours
Number of ensemble members and their resolution80, C128, ~70km

Additional comments:


4. Externally prescribed boundary conditions and their source


Sea surface temperature

1950–1986, HadOIv2

1987 onwards, OSTIA

Sea-ice

1950–1986, HadOIv2

1987 onwards, OSTIA

Snow

1950-1978, bias corrected model snow

1979 onwards, SNODEP combined with IMS
Vegetation

prescribed non-time varying tiles

Land use (and its evolution in time)prescribed non-time varying tiles
Aerosols

1950-1999, monthly across four 45° latitude bands

2000 onwards, no volcanic aerosols
Green House Gases

CO2  derived from actual observation, monthly mean at 15-deg resolution, one-year latency.

Ozone initial conditions come from data assimilation, and is actively predicted during model integration by a parameterized photo-chemistry model.

Solar forcing

modified solar-cycle-dependent solar constant table spanning historical and operational periods

Additional comments:



5. Details of model
Dynamical core (e.g., semi-Lagrangian)

Finite Volume (Harris and Lin, 2013)  — FV3 dynamical core, including global-to-regional nesting

Grid structure

Cubed-Sphere

Hydrostatic or nonhydrostaticNon-hydrostatic
Radiations parameterizationmodified RRTMG longwave and shortwave scheme (Iacono et al, 2008); cloud-radiation interaction: McICA method
Boundary layer parameterization

hybrid Eddy-Diffusivity Mass-Flux (EDMF) PBL and free atmospheric turbulence scheme (Han et al, 2016)

Convection parameterization

Scale-Aware Mass-Flux (SAMF) deep and shallow convection schemes, updated version of  Simplified Arakawa-Schubert (SAS) scheme (Han et al, 2017)                                         

Cloud parameterization

GFDL Cloud Microphysics Scheme (Zhao et al, 2018)

Land surface parameterization

Noah land surface model (4 layers) (Ek et al, 2003)



6. Further information
Operational contact point
URL of the technical note/ reference paper

Yang, F., and V. Tallapragada, 2018: Implementation and evaluation of the NOAA next generation global prediction system with FV3 dynamical core and advanced physics. Eighth Conference on Transition of Research to Operations, Austin, TX, American Meteorological Society, 1.4.

URL for list of products




7. Observational data used
URL with the list of observational data used in the reanalysis
DOI of data product if available


Quality Issues