modeling

Available Real-Time Forecasts




MARCOOS/HOPS Real-Time Forecasting System

The MARCOOS/HOPS Real-Time Forecast System has been operational since March 9, 2009. The operational forecasting system was built on the feature-oriented initialization scheme developed by Gangopadhyay et al. (1997) for the Gulf Stream Meander and Ring (GSMR) region and that by Gangopadhyay et al (2003) for the Gulf of Maine and Georges Bank (GOMGB) region. The deep water feature model set for GSMR has been melded with shallow water feature model set in the GOMGB region and further supplemented with the Levitus climatology as the background in a multi-scale objective analysis framework. The ring and front analysis from Jenifer Clark (http://users.erols.com/gulfstrm/) are used to feed into the feature modeling scheme for generating a three-dimensional initial field. The initialization field is dynamically adjusted with wind-forcing and used in a SST-assimilative forecast model using the methodology described by Brown et al (2007 a, b).

The Harvard Ocean Prediction System (HOPS) is an integrated system of data analysis and assimilation schemes, and a suite of coupled interdisciplinary (physical, acoustical, optimal, biochemical-ecosystem) dynamical models (Robinson, 1996; Robinson et al., 1999; Robinson et al., 2002). It is composed of various program packages, and its core is a primitive equation (PE) physical dynamical circulation model with modules composed of packages necessary for setting up the model domain including grid, conditioning of bathymetry, management of observational data, objective analysis, preparation of assimilation fields, etc. Details of the model equations and their implementation can be found in Spall and Robinson (1989). The latest version of the HOPS model code and its associated pre- and post-processing software and methods are being developed and maintained by the Multidisciplinary Simulation, Estimation, and Assimilation Systems (MSEAS) group at MIT (see http://mseas.mit.edu/ ) with whom we are collaborating.

The model is forced by atmospheric fields (surface momentun flux, surface heat flux, surface water flux and shortwave radiation) from the Global Forecast System (GFS) at 1/2 degree resolution, which provides 7-day forecast fields.

The forecast system is presently assimilating 3-day composite SST from the Johns Hopkins University/Applied Physics Laboratory and similar AVHRR passes processed by the MARCOOS group at University of Delaware/College of Marine and Earth Studies. The week-long forecasts are issued generally by Wednesday morning; Monday zero-hour is a typical model initialization state, with SST assimilation (Lozano et al. 1996) carried out on Monday afternoon and Tuesday noontime. Real-time Glider data is being assimilated for several hindcast runs, which will become operational in the coming weeks. The forecast fields (Temperature, Salinity, Currents) are available at www.smast.umassd.edu/modeling/RTF/ for different levels at 6-hourly intervals for the full domain, for a zoom domain for Mid-Atlantic Shelf, and for another zoom region for the Gulf of Maine. The model forecast data are available from the thredds server http://aqua.smast.umassd.edu:8080/catalog.hml. The CF-compliant version will be available during summer of 2009.

This real-time forecast system has been possible due to many years of research at the Harvard University, UMass Dartmouth, MIT and other Institutions by many researchers. Funding was made available from the Office of Naval Research (ONR), the National Aeronautics and Space Administration (NASA), the National Oceanic and Atmospheric Administration (NOAA) and the National Science Foundation (NSF) at different times to different PIs at different Institutions. We acknowledge all their efforts and supports.

References



Avijit Gangopadhyay
agangopadhya@umassd.edu • 508-910-6330
The School for Marine Science and Technology
200 Mill Road, Suite 325, Fairhaven, MA 02719