A total of 16 drifters are proposed to sample the circulation in the area of study for about a year. Deployments will be conducted in groups of 2 drifters each season in the coastal waters of Lebanon. Drifters will also be deployed along a meridional transect south of Cyprus. All drifters will be of the Surface Velocity Program (SVP) type with a holey-sock drogue at a nominal depth of 15 m. They will provide hourly observations of near-surface currents (calculated from successive positions) and SST in near-real time via the Iridium satellite system. The drifters will not be recovered.
Three Argo floats will be deployed contemporarily with the drifter operations. Floats will be programmed to measure temperature and salinity profiles down to either 700 or 2000 m (MedArgo standard). The vertical resolution will vary from 2 m near the surface to 25 m below 700 m. The MedArgo standard cycle length of 5 days will be reduced to 1 or 2 days during the period of drifter and glider operations in order to obtain more data in the water column with higher horizontal resolution.
Periodic glider missions will be planned in order to sample the thermohaline and dynamical structures of the area. Three 1000 m-rated Seagliders and two 750 m-rated SeaExplorers will be available. They will carry a sensor payload of conductivity, temperature, depth, dissolved oxygen, chlorophyll-alpha fluorescence, and optical backscatter at 470 nm and 700 nm. Typically, the biogeochemical parameters are only measured in the upper 200-500 m in order to conserve energy. Sampling rates are typically every 10 sec (< 1 m vertical spacing) for temperature and salinity.
Initially, a broad survey using one glider over one month will characterize the initial conditions. Since the glider will cover about 25 km/day, a total of 750 km horizontal track will scan the area between Cyprus and Lebanon. The mission will be planned to be as synoptic as possible: segments will be broken into straight lines of about 3-4 days in length, connected in a lawnmower-type pattern to cover the entire area.
The positions of interesting features such as coastal meanders and eddies will be estimated from the initial characterization and will be used to plan the second phase of the field experiment involving a more intense effort of surface drifters, profiling floats and a pair of simultaneously-operating gliders.
The glider data will be assimilated in near-real time into a high resolution (1 km) version of the Princeton Ocean Model (POM) merged with the OceanVar data assimilation scheme (for temperature and salinity). The high resolution POM model will be run in the domain bounded by 31ºE to the west and 33ºN to the south. The north and east boundaries are defined by the northeast Mediterranean coastline.
The results of the above month-long model-glider experiment will produce a best estimate for the sea dynamics. Even after glider data stop coming in, the model will continue to run in forecast mode, since re-initialization would likely destroy many of the features introduced by the assimilation. A control run will also be carried out, with the exact same initial and boundary conditions, but with no data assimilation. This will allow us to estimate the impact of the assimilation.