Explorer of the Seas
About the Model
The ocean surface forcings used to force the model are based on scatterometer observations and NCEP product reanalyses and operational forecasts.
The workhorse scatterometer wind product used is the blended QSCAT and NCEP reanalysis surface winds archived at UCAR's ds744.4 Home Page. These winds are provided four-times daily for the globe on a regular 0.5° resolution. The data record currently runs through Mar. 2004 and is updated about every 3 months as the QSCAT mission continues. Milliff et al. (1998, 2004) document realistic kinetic energy content in the blended winds at the grid resolution, in contrast to surface winds from weather-center analyses and forecasts. Scatterometer-derived winds are used in the assimilation step during real-time forcasting, and ar the near real time (NRT) winds provided by Dr. Paul Chang at manati.orbit.nesdis.noaa.gov/quikscat/. Complete coverage of the IAS is typically achieved every 24-hours. Because of NRT constraints, the processing for these winds has to be less precise than the wind retrievals for the QSCAT standard products. Nonetheless, the error statistics are well known and will be used in the 4DVAR procedures.
NCEP and COAMPS operational forecast winds and surface fluxes will be used for the real-time forecasts. We will quantify the effect of switching to the NCEP or COAMPS forecast model winds. SST fields (e.g. Reynolds and/or TRMM/TMI) will be used to constrain model surface temperatures during assimilation hindcast and forecast experiments (see below). Climatological data from the USGS will be used for the major rivers that impact the IAS.
Numerous ocean observation data sets are available in the IAS as summarized below. These data are comprised of (i) various satellite missions that measure surface winds, SST, sea surface height (SSH) and ocean color, (ii) research cruises and field missions collecting standard hydrographic and ocean current measurements, and (iii) buoys, ocean moorings, drifters and floats, and submarine cables. All of the ocean data for the period 1995-present below will be used.
For the hindcast and forecast experiments, in addition to satellite observations we will use data from semi-permanent in situ moorings, buoys and cables, repeat cruise tracks, and available floats and drifters. These data are illustrated schematically below.
The primary source of real-time upper ocean observations are the data sets collected by the Royal Caribbean Cruise Line (RCCL) ship, the Explorer of the Seas, a 138, 000 ton, 1020 foot commercial cruiseliner. Explorer collects upper ocean (and meteorological) observations on a continuous basis along two separate cruise tracks in the northern Caribbean indicated above (see also www.rsmas.miami.edu/rccl). The two cruise tracks are each plied once every 2 weeks meaning that Explorer circumnavigates the region once every 14 days. The primary ocean sensors are two ADCPs, and a bow-thruster seawater intake that measures bulk temperature and salinity. Data from the Explorer has been collected and archived routinely since Jan. 2001, and will continue indeffinitely. The ADCPs operate in different frequency ranges; one at 38 kHz that measures currents down to 1000 m at a vertical resolution of ~ 8 m, and one at 150 kHz that measures down to 200 m at a vertical resolution of ~ 2-4 m. Dr. Lisa Beal at the University of Miami Rosenstiel School of Marine and Atmospheric Science (RSMAS) is the scientist responsible for the ADCP data processing (Beal et al. 2007). The Explorer of the Seas ADCP data set is produced with support from SEACOOS (SouthEast U.S. Atlantic Coastal Ocean Observing System) and NOAA. Other onboard instruments (e.g. a skin-temperature radiometer, acoustic anemometers, rain gauges) will be used to estimate the forcing error statistics required to model Cf.
Satellite-derived SST products (Reynolds and Smith, 1994) and SSH from TOPEX/POSEIDON, ERS2 and JASON will be utilized heavily. Weekly SST products (e.g. TRMM/TMI) will be used both to constrain the surface heat flux forcing SST during the data assimilation phase. Following the usual practice (Pinardi et al., 2003), SSH will be corrected to remove the signal component due to the seasonal thermal expansion of the water column.
In addition to the Explorer and satellite observations, other in situ observations are available. Specifically, a number of repeat cruise tracks exist across the Yucatan Channel, as well as 8 ocean moorings. These data were collected by the Canek group at CICESE (Sheinbaum et al., 2002) and we will work with Dr Julio Sheinbaum, a member of this group, in assimilating these data. In addition, monthly timeseries of hydrographic data are available in the Cariaco basin of the southern Caribbean (F. Muller-Karger, pers. comm.). Long Caribbean Timeseries (CaTs) also exist in the northern Caribbean around Puerto Rico and US Virgin Islands (J. Corredor, pers. comm.). A submarine cable is now in place along the entire Antilles arc to measure the depth averaged transport. At the present time only the portions of cable across the Florida Straits and Grenada Passage are operational. A permanent array of moorings is also being installed just east of the Lesser Antilles (the "Antilles Array") by the University of Kiel.
Other data that are available on a regular basis are drifter data (Centurioni and Niiler, 2003) and ARGO, PALACE and S-PALACE float data which sample temperature and salinity profiles in the upper 2000 m of the water column, returning to the surface every 10 days to broadcast data via the GTS. Initially, we will treat the float data as point measurements of temperature and salinity, but later we may experiment with assimilating the Lagrangian velocity component of the data. When assimilating only temperature observations, we will also experiment with adjusting the salinity profile as described by Ricci et al. (2004).
Concurrent with the Explorer data set, data from Volunteer Observing Ships (VOS) of opportunity and yachts ("Seakeepers") are available. Also data are available from an array of NDBC (National Data Buoy Center) and TABS (Texas Automated Buoy System) buoys that monitor wind velocity and ocean surface temperature. The Canek group at CICESE maintains an array of buoys and moorings along the Yucatan coast near Cozumel and within the coastal and central Gulf of Mexico. All of the required ocean and satellite data will be obtained in near-realtime (12-24 hour delay) from NOAA Coastwatch (http://cwcaribbean.aoml.noaa.gov).