Northeast Shelf Models and Reanalyses¶
Ocean models and reanalyses that cover the US Northeast continental shelf (the Gulf of Maine to the Mid-Atlantic Bight), as candidates for a second opinion beside NECOFS (NECOFS: Northeast Coastal Ocean Forecast System (FVCOM)) or for years NECOFS does not cover. A downstream ocean model that disagrees with one parent can be set against several, built on different grids and vertical coordinates: a feature that survives FVCOM’s sigma layers, ROMS’s s-levels and a z-level model is less likely to be a numerical artifact of any one of them.
Each entry was checked against the provider’s live servers (THREDDS catalogs, OPeNDAP metadata, S3 listings, public STAC metadata) on 2026-11-17. “Not verified” marks what could not be.
Source |
Model and grid |
Vertical |
Coverage and interval |
In forcingkit |
|---|---|---|---|---|
ROMS 4D-Var, curvilinear, about 7 km |
40 s-levels |
2007-01-02 to 2024-12-31, hourly |
|
|
ROMS with assimilation, same grid |
40 s-levels |
2017-11-02 to two days ahead, hourly |
|
|
MOM6-NWA12 |
MOM6, curvilinear C-grid, 1/12 degree |
52 z levels (remapped) |
1993 to 2023, monthly (3D) |
Not yet |
GLORYS12 |
NEMO with assimilation, regular 1/12 degree |
50 z levels |
1993 to about two months ago, daily and monthly |
Not yet (needs an account) |
HYCOM GOFS 3.1 reanalysis |
HYCOM, regular 0.08 degree |
40 z levels |
1994 to 2015, 3-hourly |
HYCOM: global ocean analyses and reanalysis (parent ocean) |
ESPreSSO |
ROMS with assimilation, curvilinear |
36 s-levels |
2009 to 2018 (discontinued) |
Not yet |
NOS OFS (GoMOFS, DBOFS, CBOFS, NYOFS) |
ROMS (POM for NYOFS) |
30, 10, 20 s-levels; 7 sigma |
2015 or 2018 onward, hourly or 3-hourly |
NYOFS: NOAA New York/New Jersey Operational Forecast System, DBOFS (parent ocean) |
ECCOFS |
ROMS 4D-Var, 3 km |
50 s-levels |
Developmental, from 2026 |
Not yet |
Doppio¶
See Doppio (ROMS, Mid-Atlantic Bight and Gulf of Maine) for the reader.
Access: Rutgers THREDDS, OPeNDAP. Reanalysis history (hourly):
https://tds.marine.rutgers.edu/thredds/dodsC/roms/doppio/DopAnV3R3-ini2007_da/his; daily averages at.../DopAnV3R3-ini2007_da/avg, with monthly and yearly averages and atmospheric fluxes beside them. Real time:https://tds.marine.rutgers.edu/thredds/dodsC/roms/doppio/2017_da/his/History_Best(and.../2017_da/avg). The reanalysis is also listed on the AWS Open Data registry (not checked here).Grid: 106 x 242 rho points, about 7 km, C-grid, rotated about 37 degrees from east; 40 s-levels (Vtransform 2, Vstretching 4, theta_s 7, theta_b 2, hc 250 m).
Variables:
u,v(grid-relative, on faces),w,temp,salt,zeta,ubar,vbar, mixing coefficients, surface fluxes; the real-time aggregation addsu_eastward,v_northwardat rho points (empty for 2022).Time: reanalysis 2007-01-02 01:00 to 2024-12-31 00:00, 157,752 hourly records without a gap; real time 2017-11-02 01:00 onward, hourly. Surface elevation relative to NAVD88.
License and credit: “Freely available” in the catalog. Cite López et al. (2020), Wilkin et al. (2022), and for the reanalysis output Wilkin and Levin (2022), doi:10.17882/89673.
MOM6-NWA12 (NOAA CEFI)¶
The Northwest Atlantic 1/12 degree configuration of MOM6 with COBALT biogeochemistry, from NOAA GFDL for NOAA’s Changing Ecosystems and Fisheries Initiative (Ross et al., 2023).
Access: NOAA PSL THREDDS,
https://psl.noaa.gov/thredds/catalog/Projects/CEFI/regional_mom6/cefi_portal/northwest_atlantic/full_domain/catalog.xml, with productshindcast,seasonal_forecast,seasonal_reforecast,decadal_forecast,multi_decadal_outlookandlong_term_projection, each split into daily, monthly and yearly,raw(model grid) andregrid(regular grid), and by release (r20230520,r20250715;latestisr20250715). One file per variable, for examplehindcast/monthly/raw/r20250715/uo.nwa.full.hcast.monthly.raw.r20250715.199301-202312.nc. The same tree is in the anonymous S3 bucketnoaa-oar-cefi-regional-mom6-pds, with kerchunk sidecars and Zarr stores underarraylake/. PSL returned HTTP 429 after a few quick requests, so a reader should pace itself or use the bucket.Grid: native C-grid, 845 x 775 tracer points (
yh,xh), 98 W to 36 W and 5 N to 52 N; curvilinear (2Dgeolon,geolatinocean_static.nc). The regridded files are on a regular 844 x 774 grid.Vertical: remapped to 52 z levels (
z_l, 2.5 m to 6250 m; 5 m spacing to 50 m), withz_iinterfaces. The native coordinate is not published.Variables:
uo,vo(grid-relative, on faces),uo_rotate,vo_rotate(east and north at tracer points),thetao,so,zos,ssh, surface and bottom fields, mixed layer depth.Time: 3D physics only as monthly means, 1993-01 to 2023-12 (
r20250715). The daily output holds surface and bottom fields:r20230520(1993 to 2019) has daily SSH and surface currents;r20250715daily has biogeochemistry only.License and credit: no license attribute in the files; NOAA’s open data terms for its buckets presumably apply (not verified). Cite Ross et al. (2023) and the data DOI 10.5281/zenodo.7893386, both named in the files.
Use here: monthly means cannot resolve tides or an eddy that lives a few days, so NWA12 is a climatological check, not an hourly parent.
GLORYS12 (Copernicus Marine)¶
Mercator Ocean’s global reanalysis for the Copernicus Marine Service, product GLOBAL_MULTIYEAR_PHY_001_030 (Lellouche et al., 2021).
Access: Copernicus Marine (the
copernicusmarinetoolbox, ARCO Zarr), which needs a free account. forcingkit does not create one; a user who wants GLORYS registers and supplies their own credentials. Public metadata (no login):https://stac.marine.copernicus.eu/metadata/GLOBAL_MULTIYEAR_PHY_001_030/product.stac.json. Dataset IDscmems_mod_glo_phy_my_0.083deg_P1D-m_202311(daily) andcmems_mod_glo_phy_my_0.083deg_P1M-m_202311(monthly), with static coordinate, bathymetry and mean dynamic topography sets.Grid: regular 1/12 degree, 80 S to 90 N; 50 z levels, 0.494 m to 5727.9 m.
Variables:
uo,vo(east, north),thetao(potential temperature),so,zos,bottomT,mlotst, sea ice.Time: daily means 1993-01-01 to 2026-10-06 and monthly means to 2026-09 in the STAC metadata; the recent months are the interim extension, folded into the same dataset IDs (the date where the reanalysis becomes the interim run is not verified). Daily means, so tides come from elsewhere.
License and credit: the Copernicus Marine Service licence (STAC
"license": "proprietary", linking to the service commitments and licence page). Products must be credited as “Generated using E.U. Copernicus Marine Service Information”, with the product DOI 10.48670/moi-00021; publications state “This study has been conducted using E.U. Copernicus Marine Service Information”. Cite Lellouche et al. (2021). Whether the ARCO stores can be read anonymously was not tested.
HYCOM GOFS 3.1 reanalysis¶
The HYCOM consortium’s global reanalysis, GLBv0.08 expt_53.X, which HYCOM: global ocean analyses and reanalysis already uses
as the first segment of its parent-ocean record.
Access:
https://tds.hycom.org/thredds/dodsC/GLBv0.08/expt_53.X/data/<year>per year (1994 to 2015) and the aggregationGLBv0.08/expt_53.X; the raw 3-hourly files are also in the S3 buckethycom-gofs-3pt1-reanalysis(us-west-2, about 4.8 GB each).Grid: regular, 4500 x 3251, 0.08 degree in longitude (0.04 degree in latitude poleward of 40 degrees); 40 standard z levels, 0 m to 5000 m (the model’s 41 hybrid layers interpolated).
Variables:
water_u,water_v(east, north),water_temp,salinity,surf_el, and*_bottomvariants, stored as scaled Int16. HYCOM states that the temperature is in situ, not potential.Time: 1994-01-01 12:00 to 2015-12-31 09:00, 3-hourly, with gaps (2010 holds 2888 of 2920 steps, 2015 holds 2861). The analysis experiments that follow (
expt_56.3toexpt_93.0, then ESPC-D-V02 from 2024-08-10) are listed in HYCOM: global ocean analyses and reanalysis.License and credit: “Approved for public release. Distribution unlimited.” (file metadata); the AWS registry entry states no restrictions on use. No acknowledgment sentence was found; cite Bleck (2002) and Chassignet et al. (2007), listed in Sources and Citations.
Use here: 9 km, so it resolves the shelf break and Gulf Stream rings but not canyon-scale flow.
ESPreSSO (Rutgers, discontinued)¶
Doppio’s predecessor, also ROMS with data assimilation for MARACOOS.
Access:
https://tds.marine.rutgers.edu/thredds/catalog/roms/espresso/catalog.xml, marked discontinued in favor of Doppio. Version 2 (2013 to 2018):.../dodsC/roms/espresso/2013_da/avg/ESPRESSO_Real-Time_v2_Averages_Bestand.../2013_da/his/ESPRESSO_Real-Time_v2_History_Best; version 1 (2009 to 2013) underroms/espresso/2009_da; hindcast experiments underroms/esp2006.Grid: 82 x 130 rho points, C-grid, 36 s-levels; the catalog says about 6 km (not verified against the grid metrics).
Variables:
zeta,u,v,temp,salt,h.Time: version 2 daily averages 2013-05-19 to 2018-09-19 (the last days may be forecast).
License and credit: no license attribute found. The paper to cite was not verified.
NOAA NOS Operational Forecast Systems¶
NOAA CO-OPS’s regional forecast systems. Besides NECOFS (served by SMAST; see NECOFS: Northeast Coastal Ocean Forecast System (FVCOM)), four cover parts of the region.
Access: the anonymous S3 bucket
noaa-nos-ofs-pds(per model, per day, for examplegomofs/netcdf/2026/11/15/gomofs.t00z.20261115.fields.f003.nc; older months inYYYYMM/folders, from 2022-01 for CBOFS and 2024 for the others); the 7-day CO-OPS THREDDS aggregationshttps://opendap.co-ops.nos.noaa.gov/thredds/dodsC/<MODEL>/fmrc/Aggregated_7_day_<MODEL>_Fields_Forecast_best.ncd; and NCEI before that,https://www.ncei.noaa.gov/thredds/catalog/model-<model>-files/.Systems:
System
Model and grid
Levels
Output and archive
GoMOFS
ROMS, 777 x 1173, Gulf of Maine (about 700 m, not verified)
30 s
3-hourly fields (hourly 2D surface files); NCEI from 2018-01-10
DBOFS
ROMS, 732 x 119, Delaware Bay
10 s
Hourly fields; NCEI from 2015-08-12
CBOFS
ROMS, 291 x 332, Chesapeake Bay
20 s
Hourly fields; NCEI from 2015-08-12
NYOFS
POM, 134 x 73, NY/NJ Harbor
7 sigma
Nowcast and forecast files four times a day; no temperature or salinity; NCEI from 2015-08-19
Variables:
zeta,u,v,temp,salt(ROMS, grid-relativeu,von faces, so the same destaggering and rotation as Doppio); NYOFSzeta,u,v,w. The ROMS systems’ Vtransform and Vstretching were not checked.License and credit: NOAA data on NODD “are open to the public and can be used as desired”; NOAA requests attribution (AWS registry). Cite the model papers in Sources and Citations.
ECCOFS (developmental)¶
The East Coast Coastal Ocean Forecast System, a ROMS 4D-Var system by Rutgers, UC Santa Cruz and
NOS at 3 km with 50 s-levels, from the Grand Banks to Venezuela, planned for operations at NCEP
in 2028. Output is in the S3 bucket noaa-nos-eccofs-pds (avg/, his/, and others; for
example avg/eccofs_avg_20260608_0001.nc, 7.6 GB a day) and on Rutgers THREDDS under
roms/eccofs. Too young for a reanalysis window, but a future successor to Doppio.
Next steps¶
MOM6-NWA12 monthly 3D fields through
uo_rotate,vo_rotate,thetao,sofrom the S3 mirror, as a climatological comparison.ESPreSSO and the NOS ROMS systems through the Doppio reader, once their vertical transform parameters are checked (the reader supports Vtransform 2 only).
GLORYS12, for a user with a Copernicus Marine account.
References¶
Lellouche, J.-M., E. Greiner, R. Bourdallé-Badie, et al. (2021). The Copernicus global 1/12° oceanic and sea ice GLORYS12 reanalysis. Frontiers in Earth Science, 9, 698876. doi:10.3389/feart.2021.698876
López, A. G., J. L. Wilkin and J. C. Levin (2020). Doppio: a ROMS (v3.6)-based circulation model for the Mid-Atlantic Bight and Gulf of Maine: configuration and comparison to integrated coastal observing network observations. Geoscientific Model Development, 13(8), 3709-3729. doi:10.5194/gmd-13-3709-2020
Ross, A. C., C. A. Stock, A. Adcroft, et al. (2023). A high-resolution ocean model for the Northwest Atlantic: model description and evaluation (MOM6-COBALT-NWA12 v1.0). Geoscientific Model Development, 16, 6943-6985. doi:10.5194/gmd-16-6943-2023
Wilkin, J. and J. Levin (2022). Outputs from a Regional Ocean Modeling System (ROMS) data assimilative reanalysis (version DopAnV3R3-ini2007) of ocean circulation in the Mid-Atlantic Bight and Gulf of Maine for 2007-2021. SEANOE. doi:10.17882/89673
Wilkin, J., J. Levin, A. Moore, H. Arango, A. López and E. Hunter (2022). A data-assimilative model reanalysis of the U.S. Mid Atlantic Bight and Gulf of Maine: Configuration and comparison to observations and global ocean models. Progress in Oceanography, 209, 102919. doi:10.1016/j.pocean.2022.102919