P.F.J. Lermusiaux, P.J. Haley, C. Mirabito, E. Mule Massachusetts Institute of Technology Center for Ocean Engineering Mechanical Engineering Cambridge, Massachusetts	MSEAS Deterministic Ocean Forecasts MSEAS Probabilistic Ocean Forecasts MSEAS Methods & Systems Atmos. Forecasts Data sources
This research is sponsored by the The National Academies of Sciences, Engineering, and Medicine.

GOFFISH MASTR Sea Exercise Page
GOFFISH Project Main Page

The GRand Adaptive Sampling Experiment (GRASE) is a collaborative sea experiment that occurs in the Gulf of Mexico from April to September 2025. We employ our MIT-MSEAS data-assimilative Primitive-Equation (PE) submesoscale-to-regional-scale ocean-modeling system for real-time deterministic and probability forecasts of ocean fields and derived quantities. Specific objectives include (i) multi-resolution ensemble forecasts with initial conditions downscaled from multiple models and implicit 2-way nesting, (ii) mutual information forecasts for predictability studies, (iii) optimal adaptive sampling guidance for sea sensing platforms, and (iv) reachability forecasts for underwater vehicles. Finally, we provide varied data sets that we process. We thank all of the GRASE team members for their input and collaboration, namely Steve Morey (FAMU); Steve DiMarco, Sakib Mahmud, Anthony Knap, and Xiao Ge (TAMU); Scott Glenn, Travis Miles, Kaycee Coleman, and Michael Smith (Rutgers); Michael Leber, Rafael Ramos, and Jill Storie (Woods Hole Group); Eric Chassignet (FSU); Amy Bower (WHOI); Benjamin Jaimes de la Cruz and Nick Shay (Miami); Miguel Tenreiro, Enric Pallas Sanz, Julio Sheinbaum and Paula Pérez-Brunius (CICESE); Jan van Smirren (Ocean Sierra); Ruoying He (NCSU); and finally Michael Feldman, and Megha Khadka, Arianna Trapp, and Francis Wiese (NAS). We also thank the HYCOM Consortium and Mercator Ocean for their ocean model fields, and NCEP for their atmospheric forcing data. Finally, we thank our MSEAS group members.

Real-time MSEAS Forecasting

• MSEAS Deterministic Ocean Forecasts

  Real-Time Estimates of Present (18 May 0000Z) Environmental Conditions (initialized from downscaled HYCOM)
  Gulf Modeling Domain				Sections Locations Overlaid on Bathymetrry
  Interactive Ocean Physics Forecast
  Reachability	2m Vorticity	0m Sigma-T	0m Sigma-T Unc.	Sec. 1 Temperature	Sec. 1 Temp. Unc.	Sec. 1 Salinity	Sec. 1 Salinity Unc.

   

• Glider Planning:

  Product	April									May
  	7	14	15	17	20	23	25	28	29	1	4	7	11
  0-1000m Avg. Vel.	X	X	X		X	X	X	X	X	X	X	X	X
  Reachability Analysis Reachable sets and reachability front	X	X	X		X	X	X	X	X	X	X	X	X
  Optimal path and headings	X	X	X		X	X	X	X	X	X	X	X	X
  Reachability Prob.				X		X	X			X	X		X


  MSEAS tidally-forced forecasts initialized from downscaled HYCOM

Nowcast and Forecast Products with descriptive dynamics [Product details]			Analyses and Forecasts Issued On
			April						May
			7	14	15	20	23	29	4	7	11
Ocean Physics	Horizontal Maps	Central Forecast	X	X	X	X	X	X	X	X	X
	Interactive Forecast		X	X	X	X	X	X	X	X	X
	Vertical Sections	Central Forecast	X	X	X	X	X	X	X	X	X



• MSEAS Probabilistic Ocean Forecasts: Physics-balanced stochastic models and ESSE dominant subspace decompositions are used to represent the dominant uncertainties in the initial and boundary conditions for T, S, u, v, w, and η fields for the HYCOM and Mercator modeling system fields, in the MSEAS model parameters, in the atmospheric surface forcing flux fields, and in the tidal forcing parameters.

  Nowcast and Forecast Uncertainty Products with dynamics descriptions			Probabilistic Analyses and Forecasts Issued On						1.5 kt Velocity Probability
  			April			May
  			17	23	30	4	7	11
  Probabilistic Ocean Physics	Horizontal Maps	Mean and Std. Dev.	X	X	X	X	X	X
  	Interactive Mean and Std. Dev. Forecast		X	X	X	X	X	X
  	Vertical Sections	Mean and Std. Dev.	X	X	X	X	X	X
  	Hazardous Velocities Statistics		X	X	X	X	X	X



• Optimal Adaptive Sampling: Mutual Information (MI) Forecasts: The objective of our optimal path planning is to maximize information about the Loop Current (LC) dynamics, especially the separation of LC eddies.

  Product	Verification Field	April	May
  		26	1	4	7	11
  Mutual Information	T + S	X
  	U + V	X	X	X	X
  	σT	X	X	X	X



• Methods and Systems: The probabilistic MIT-MSEAS Primitive Equation (PE) ocean modeling system is utilized in real time to provide ocean forecasts in the region. The ocean forecasts are initialized from HYCOM or Mercator, downscaled to higher resolution and updated with ocean data from varied open sources of opportunity (ARGO floats, gliders, SST, etc.). Ensemble forecasts are initialized using ESSE procedures. These ocean simulations are forced by atmospheric flux fields forecast by the Global Forecast System (GFS) 0.25° or the North American 12km Model (NAM) from the National Centers for Environmental Prediction (NCEP) and by tidal forcing from TPXO-10, but adapted to the bathymetry and coastlines. For ESSE probabilistic forecasts, initial conditions, model parameters, boundary conditions, atmospheric forcing and tidal forcing amplitudes and phases, are all perturbed according to their respective uncertainties.


• MSEAS-Processed Data
  • Summary of all available data of opportunity:
    • Summary by type: March 20 to present
    • Locations with time: March 20 to present
  • Data of opportunity:
    • Argo profiles:
      • March 20 to present
      • Argo float summary (profiles by float ID, dive depth, and time)
    • Gliders (Source: IOOS Glider DAC): March 20 to present | Depth-avg velocity
    • NDBC Buoys: March 20 to present
    • SST: MetOffice/CMEMS OSTIA Analysis (daily 0.05°): March 20 to present | Caribbean region | Interior GoM region
    • SSH (March 20 to present):
      • AVISO (DUACS-processed NRT multi-satellite; Source: CMEMS):
        • Along-Track ADT (7km; daily; filtered)
        • Gridded ADT via OI (0.125°; daily)
      • SWOT Swath ADT (Source: PO.DAAC): March 20 to present (2km; daily)
    • SSC: Multi-sensor DINEOF global gap-filled Chl-a (March 20 to present; 9km; daily)
    • Multi-sensor CMEMS global 4km Chla (March 20 to present; 4km; daily) | Caribbean region | Interior GoM region


• MSEAS-processed atmospheric forcing flux forecasts:
  

  Merged NCEP NAM 12km and GFS 0.25°: April-May 2025
  Daily average wind stress, E-P, heat flux, and SW Radiation (from 0Z forecast each day)
  Flux snapshot plots (1 fct issued daily)	April	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21	22	23	24	25	26	27	28	29	30
  	May	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21	22	23	24	25	26	27	28	29	30	31



• Woods Hole Group Mapper

Top of page

Data sources

• Initialization and BCs
  • HYbrid Coordinate Ocean Model (HYCOM) Description (HYCOM Consortium)
  • Operational Mercator Ocean 1/12° Model Output (Copernicus Marine Data Store)
• Ocean synoptic
  • ARGO float data access
  • Gliders:
    • IOOS ERDDAP/TableDAP page
    • UGOS-3 gliders ERDDAP/TableDAP page
  • National Data Buoy Center station selection
  • SST Imagery/Data: Global Ocean OSTIA Sea Surface Temperature and Sea Ice Analysis (CMEMS)
  • SSH Imagery/Data:
    • AVISO+ (distributed by CMEMS) L3 NRT along-track SSH
    • AVISO+ (distributed by CMEMS) L4 NRT 0.125° gridded SSH
    • SWOT Level 2 KaRIn Low Rate Sea Surface Height Data Product, Version 2.0
  • SSC Imagery/Data: NOAA CoastWatch DinEOF3 9km gridded product
  • HF Radar:
    • HFRnet THREDDS server
    • UGOS-3 - HFR: Yucatan - 6km - Sea Surface Currents
• Ocean historical/climatological
  • Data sets and products NOAA National Centers for Environmental Information
    • World Ocean Database (WOD)
• Atmospheric forcing
  • Weather Research and Forecasting (WRF) Model Real-Time Forecasts
  • National Centers for Environmental Prediction (NCEP) products:
    • Global Forecast System Model: version 4 (GFS), at 0.5 degree resolution (GFSp5) and also at 0.25 degree resolution (GFSp25)
      • Direct link to GFSp5/GFSp25 model output
      • NOMADS - NOAA Operational Model Archive and Distribution System (main page)
      • General information about the GFS model
      • Index of GFSp5/GFSp25 GRIB records

Top of page