P.F.J. Lermusiaux, P.J. Haley, C. Mirabito, M. Doshi, W.H. Ali, A. Gupta, A. Rodriguez 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 Results From Partners Data sources

CALYPSO Project Main Page

The two-part CALYPSO 2022 experiment occurs in the Balearic Sea from February 18 to March 12, 2022 (with modeling work starting February 8) and from March 25 to June 29, 2022 (with modeling work starting on April 13). We employ our MIT-MSEAS data-assimilative Primitive-Equation (PE) submesoscale-to-regional-scale ocean-modeling system for implicit 2-way nested predictions. We issue real-time deterministic and uncertainty forecasts of ocean fields, as well as Lagrangian flow maps, coherent sets, subduction forecasts, and drifter forecasts. Other objectives include (i) issuing nested ensemble forecasts; (ii) forecasts downscaled from multiple models; (iii) forecasts using MSEAS 2D and 3D SeaVizKit, and (iv) performing Lagrangian data assimilation and adaptive sampling. Finally, we provide varied data sets that we process. We thank all of the CALYPSO team members for their inputs and collaboration. We also thank NCEP for their atmospheric forcing data.

This research is sponsored by the Office of Naval Research

MSEAS Forecast Products for Ships

Real-time MSEAS Forecasting

• MSEAS Deterministic Ocean Forecasts

  Real-Time Estimates of Present (17 Mar 0000Z) Environmental Conditions (downscaled from WMOP)
  Full Modeling Domain (900m resolution)				Section 4 (900m resolution)
  Interactive Ocean Physics Forecast		3D Lagrangian Fct. of Subduction	Interactive 96-hr Flow Map Forecast
  0m Sigma-T	2m Vorticity			Temperature	Salinity

   

    MSEAS high-resolution tidally-forced forecasts downscaled from WMOP

  Nowcast and Forecast Products with descriptive dynamics [Product details]

  Analyses and Forecasts Issued On

  February 2022

  8

  9

  10

  11

  12

  13

  14

  15

  16

  17

  18

  19

  20

  21

  22

  23

  24

  25

  26

  27

  28

  Ocean Physics

  Horizontal Maps

  Central Forecast

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  Interactive Forecast

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  Vertical Sections

  Central Forecast

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  Interactive Flow Maps Forward flowmaps display at the initial location of a water parcel the final position of that water parcel as it is advected over a specified time window. For instance, if the value of the 96 hour forward z-flowmap at a given (x, y, z) location is 50m, the final depth of that point will be 50m after 96 hours of being advected in the flow. Hence, the z-flowmap for example does not indicate where the water parcel goes down nor where it ends up. Backward flowmaps display at the final location of a water parcel the initial position of that water parcel as it is advected over a specified time window.

  Forward

  6-hour

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  96-hour

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  Backward

  6-hour

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  96-hour

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  Lagrangian Coherent Structures The figures in this row show the finite-time Lyapunov exponents (FTLEs) in the domain, at different times. These FTLE computations are for a time window of 6 or 96 hours for the forecasts on various dates as specified in the figures captions. The flowmaps were calculated using a PDE-based approach with 5th order spatial and 3rd order temporal schemes, and a novel composition based approach. The FTLE field specifies the amount of stretching of the flowmap in the considered time interval. Specifically, the forward / backward FTLE is a logarithmic scaling of the dominant singular values of the forward / backward flowmap with the time duration of interest. The forward FTLE ridges thus highlight codimension-one manifolds (curves in 2D, surfaces in 3D) that repel particles starting on either side, whereas the backward FTLE ridges highlight codimension-one manifolds that attract particles towards the ridge that originally start far apart. As a reminder, 12:00Z = 13:00 CET, i.e., CET is 1 hour ahead of Zulu time.

  LCS FTLE Maps

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  Dilation Maps

  X

  X

  Drifters The figures in this row of the table show the trajectories in the domain for a given time interval. For this project, the drifters are released at every 10 grid points, and the grid has a model resolution of 900m. These trajectory computations correspond to time windows of 96 hours, for the forecasts specified in the headers of the respective columns.

  Trajectories (6 depths)

  96-hour

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  Special Lagrangian Products Special products, commonly still images, of our MSEAS forecasts of the ocean physics and Lagrangian 3D transport of waters (e.g. waters starting at various depths). What is plotted is the depth reached by the water parcel by the end of the forecast, starting on a specific date and time at that (x, y, z) location (i.e., it does not show the horizontal displacement of the parcel, only the depth reach by a parcel released at x, y, z). For the 3D+time Lagrangian subduction of waters, what is shown are the motions in 3D of water volumes that subduct the most from the surface layers to deeper layers (computed based on our forward flowmap forecast).

  3D Lagrangian subduction of waters

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  
  
  Note: In the table below, for columns with two Xs, the left "X" links to the ocean physics forecast from the standalone MSEAS-PE run (900m res); the right "X" links to the forecast from the high-resolution implicit 2-way nested MSEAS-PE run (900m resolution Balearic Sea domain and 300m resolution process domain).

  Nowcast and Forecast Products with descriptive dynamics [Product details]

  Analyses and Forecasts Issued On

  March 2022

  1

  2

  3

  4

  5

  6

  7

  8

  9

  10

  11

  12

  13

  Ocean Physics

  Horizontal Maps

  Central Forecast

  X

  X

  X

  X

  X

  X

  X

  X

  X X

  X

  X X

  X

  X X

  Interactive Forecast

  X

  X

  X

  X

  X

  X

  X

  X

  X X

  X

  X X

  X

  X X

  Vertical Sections

  Central Forecast

  X

  X

  X

  X

  X

  X

  X

  X

  X X

  X

  X X

  X

  X X

  Interactive Flow Maps Forward flowmaps display at the initial location of a water parcel the final position of that water parcel as it is advected over a specified time window. For instance, if the value of the 96 hour forward z-flowmap at a given (x, y, z) location is 50m, the final depth of that point will be 50m after 96 hours of being advected in the flow. Hence, the z-flowmap for example does not indicate where the water parcel goes down nor where it ends up. Backward flowmaps display at the final location of a water parcel the initial position of that water parcel as it is advected over a specified time window.

  Forward

  6-hour

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  96-hour

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  Backward

  6-hour

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  96-hour

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  Lagrangian Coherent Structures The figures in this row show the finite-time Lyapunov exponents (FTLEs) in the domain, at different times. These FTLE computations are for a time window of 6 or 96 hours for the forecasts on various dates as specified in the figures captions. The flowmaps were calculated using a PDE-based approach with 5th order spatial and 3rd order temporal schemes, and a novel composition based approach. The FTLE field specifies the amount of stretching of the flowmap in the considered time interval. Specifically, the forward / backward FTLE is a logarithmic scaling of the dominant singular values of the forward / backward flowmap with the time duration of interest. The forward FTLE ridges thus highlight codimension-one manifolds (curves in 2D, surfaces in 3D) that repel particles starting on either side, whereas the backward FTLE ridges highlight codimension-one manifolds that attract particles towards the ridge that originally start far apart. As a reminder, 12:00Z = 13:00 CET, i.e., CET is 1 hour ahead of Zulu time.

  LCS FTLE Maps

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  Dilation Maps

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  Drifters The figures in this row of the table show the trajectories in the domain for a given time interval. For this project, the drifters are released at every 10 grid points, and the grid has a model resolution of 900m. These trajectory computations correspond to time windows of 96 hours, for the forecasts specified in the headers of the respective columns.

  Trajectories (6 depths)

  96-hour

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  Special Lagrangian Products Special products, commonly still images, of our MSEAS forecasts of the ocean physics and Lagrangian 3D transport of waters (e.g. waters starting at various depths). What is plotted is the depth reached by the water parcel by the end of the forecast, starting on a specific date and time at that (x, y, z) location (i.e., it does not show the horizontal displacement of the parcel, only the depth reach by a parcel released at x, y, z). For the 3D+time Lagrangian subduction of waters, what is shown are the motions in 3D of water volumes that subduct the most from the surface layers to deeper layers (computed based on our forward flowmap forecast).

  3D Lagrangian subduction of waters

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  X

  
  

  Nowcast and Forecast Products with descriptive dynamics [Product details]			Analyses and Forecasts Issued On
  			May 2022	June 2022
  			30	6	14	19
  Ocean Physics	Horizontal Maps	Central Forecast	X X	X X	X X	X
  	Vertical Sections	Central Forecast	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 eta fields for the WMOP modeling system fields, 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
  			February/March 2022
  			25	2	4	6	8	10
  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

 

• Methods and Systems: The probabilistic MIT-MSEAS Primitive-Equation (PE) ocean modeling system is utilized in real-time to provide ocean forecasts for Lagrangian transport and analyses in the region. The modeling system was set-up in an implicit 2-way nesting configuration (900m resolution Balearic Sea domain and 300m resolution process domains). The ocean forecasts are initialized from WMOP, downscaled to higher resolution and updated with ocean data from varied open sources of opportunity (CTDs, 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° model from the National Centers for Environmental Prediction (NCEP) and by tidal forcing from TPXO8, but adapted to the high-resolution bathymetry and coastlines.

 

• MSEAS-processed atmospheric forcing flux forecasts:
  

  NCEP GFS 0.25°	January 2022						February 2022
  Daily average wind stress, E-P, heat flux, and SW Radiation (from 0Z forecast each day)
  Flux snapshot plots (a single 6-day fct issued daily)	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X
  Forecast snapshot plots (Four 6-hour forecasts/day)	26	27	28	29	30	31	1	2	3	4	5	6	7	8	9	10	11	12	13	14

  
  

  NCEP GFS 0.25°	February 2022														March 2022
  Daily average wind stress, E-P, heat flux, and SW Radiation (from 0Z forecast each day)
  Flux snapshot plots (a single 6-day fct issued daily)	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X
  Forecast snapshot plots (Four 6-hour forecasts/day)	15	16	17	18	19	20	21	22	23	24	25	26	27	28	1	2	3	4	5	6	7	8	9	10	11	12

  
  

  NCEP GFS 0.25°	April 2022
  Daily average wind stress, E-P, heat flux, and SW Radiation (from 0Z forecast each day)
  Flux snapshot plots (a single 6-day fct issued daily)	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X
  Forecast snapshot plots (Four 6-hour forecasts/day)	14	15	16	17	18	21	22	23	24	25	26	27	28	29	30

  
  

  NCEP GFS 0.25°	May 2022
  Daily average wind stress, E-P, heat flux, and SW Radiation (from 0Z forecast each day)
  Flux snapshot plots (a single 6-day fct issued daily)	X	X	X	X	X	X	X			X	X		X	X						X	X	X	X	X	X	X	X	X	X	X	X
  Forecast snapshot plots (Four 6-hour forecasts/day)	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

  
  

  NCEP GFS 0.25°	June 2022
  Daily average wind stress, E-P, heat flux, and SW Radiation (from 0Z forecast each day)
  Flux snapshot plots (a single 6-day fct issued daily)	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X
  Forecast snapshot plots (Four 6-hour forecasts/day)	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

 

• MSEAS-Processed Data
  • Summary of all available data of opportunity:
    • Summary by type: January 2022 | February 2022 | March 2022 | April 2022 | May 2022 | June 2022
    • Locations with time: January 2022 | February 2022 | March 2022 | April 2022 | May 2022 | June 2022 | Last 3 days
  • ARGO: January 2022 | February 2022 | March 2022 | April 2022 | May 2022 | June 2022
  • SST: Sentinel-3 A/B Near-Real-Time Imagery: January 2022 | February 2022 | March 2022 | April 2022 | May 2022 | June 2022
  • SSH: AVISO Near-Real-Time Multi-Satellite Imagery
    • Along-Track ADT: January 2022 | February 2022 | March 2022 | April 2022 | May 2022 | June 2022
    • Gridded ADT (0.25° resolution): January 2022 | February 2022 | March 2022 | April 2022 | May 2022 | June 2022
  • SSC (Chl-a) data from merged VIIRS (4 km resolution): January 2022 | February 2022 | March 2022 | April 2022 | May 2022 | June 2022
  • Balearic Sea Buoys (T/S/u/v)
  • P. Testor Glider: Location and Profile Summary
  • SOCIB Slocum Gliders: Location and Profile Summary
  • R/V Pourquoi Pas? EcoCTD profiles: Locations and Profile Summary
  • R/V Pourquoi Pas? Shipboard ADCP: Location and Current Velocity
  • R/V Pelagia Underway: Locations and T/S
  • R/V Pelagia UCTD: Locations and Profile Summary
  • R/V Pelagia Shipboard ADCP: Location and Current Velocity
  • SOLO-II Floats: Locations and Profile Summary
  • Spray Gliders: Location and Profile Summary

 

• Results from our CALYPSO partners:
  • Latest SOCIB WMOP Real-time Forecasts, Satellite Images, and Assimilated Data

 

• Acknowledgements: We thank our CALYPSO colleagues for their data, inputs, and collaboration. We especially thank Baptiste Mourre and the SOCIB team for the WMOP model fields and data. We also thank Prof. Michael Allshouse and graduate student Aravind Meenambika from Northeastern University for their collaboration with the dilation map computations. We also thank Tamay Özgökmen and Pierre Poulain for their drifter data. We also wish to thank Matthew Pyle, Eric Rogers, Geoff DiMego, and Arun Chawla of NCEP for help and support for atmospheric forcing data.

Top of page

Data sources

• Copernicus Mediterranean Sea Physics Analysis and Forecast System - Evaluation and Data Assimilation (CMCC)
• Ocean synoptic
  • The Global Temperature and Salinity Profile Programme (GTSPP)
    • Data Access
      • Real-Time Data
  • SST Images/Data
    • AVHRR on METOP data overview
    • AVHRR on METOP data from OSI SAF, hosted by JPL
  • HF Radar:
    • Estrecho de Gibraltar HF Radar system information
    • Estrecho de Gibraltar HF Radar data
    • Estrecho de Gibraltar HF Radar Current Maps
    • Other Mediterranean HF Radar Sites (from Mongoos)
  • PERSEUS project cast database
  • Copernicus data page
  • EMODnet
  • CMEMS Products for the Mediterranean Sea:
    • CMEMS Near-Real-Time In Situ Observations
    • Mediterranean Sea Observations and Analysis Product List
  • Coriolis Operational Oceanography data selection page
• Ocean historical/climatological
  • The Global Temperature and Salinity Profile Programme (GTSPP)
    • Data Access
      • Best-Copy Data
  • Data sets and products NOAA National Centers for Environmental Information
    • World Ocean Database (WOD)
• Data/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 data
      • NOMADS - NOAA Operational Model Archive and Distribution System (main page)
      • General information about the GFS model
      • Index of GFSp5/GFSp25 GRIB records
  • The European Centre for Medium-Range Weather Forecasts (ECMWF) ERA Intrim (global Re-Analysis)
    • General information
    • Data access
  • Coupled Ocean / Data/Atmosphere Mesoscale Prediction System (COAMPS)
  • Navy Global Environmental Model (NAVGEM) products:
    • Real-time data
    • NAVGEM 0.5: Real-time and historical data
  • Agencia Estatal de Meteorología (AEMET) products:
    • HARMONIE-AROME (formerly HIRLAM) model forecast images

Top of page