Activities in the area of COASTAL OCEAN MODELING AND DATA ASSIMILATION (From the GFDL Annual Report for FY94/95) Data Assimilation And Model Evaluation Experiments T. Ezer G.L. Mellor ACTIVITIES FY94 The Data Assimilation and Model Evaluation Experiments (DAMEE) is a Navy sponsored project aimed at the assessment of nowcast/forecast capabilities of several high resolution regional models of the Gulf Stream region. The sigma coordinate Princeton ocean model (POM), one of the models tested, has shown considerable nowcast and forecast skill. An efficient data assimilation scheme, based on a three-dimensional optimum interpolation approach, has been developed and successfully tested for continuous assimilation of satellite altimeter data. The scheme, which is based on the projection of surface data into the deep ocean using preprocessed surface-subsurface correlations, has now been extended to include other data sources such as sea surface temperature and the position of oceanic fronts. Implementation of the scheme to simulations of other regions such as the Gulf of Mexico, the tropical Pacific and the global ocean are being tested now at Princeton, at GFDL and elsewhere. PLANS FY95 As part of a new navy initiative, now in its planning stages, the DAMEE project will be extended from the Gulf Stream region to the entire North Atlantic. Several high resolution models will be evaluated and compared for their capability to predict and simulate meso-scale variability and climatological properties of the Atlantic ocean. The data assimilation methodology, previously developed for regional modeling, will be tested for the North Atlantic Ocean, and important scientific issues concerning boundary current dynamics and multi-scale interactions will be studied. Development of a Coastal Forecast System for the U. S. East Coast P. Chen T. Ezer G.L. Mellor ACTIVITIES FY94 This project, with support from the NOAA Coastal Ocean Program, is a cooperative effort between Princeton University, GFDL, the National Meteorological Center (NOAA/NMC), and the National Ocean Service (NOAA/NOS). The Princeton Ocean Model was coupled (at first only in a one-way interaction) to NMC's ETA, meso-scale, atmospheric model, in order to provide nowcast and forecast information of the coastal ocean. The experimental system has been implemented in an operational environment at NMC with a similar system operated at Princeton/GFDL for scientific and developmental purposes. At this time, the focus is on the prediction of the wind-driven subtidal signal of sea level, including storm surge. The coastal forecast system (CFS) has been in operation since August, 1993, providing a daily 24-hour forecasts of temperatures, salinities, velocities and sea level. Initial evaluation shows considerable skill in the prediction of coastal sea level, when compared with water level tide gages along the coast. Underway now are studies aimed at the improvement of the sea level prediction. For example, the addition of the atmospheric surface pressure loading (otherwise approximated by the so called inverted barometer correction which excludes dynamics) shows improvement in the model prediction. Significant improvement is obtained at Bermuda; unlike most of the east coast stations where the sea level is mostly wind stress-driven, surface pressure plays a more significant role at Bermuda. Another important effect that is being studied now is the thermal expansion effect; this effect is responsible for a seasonal sea level change of 10-15 cm, but is missing from most ocean models due to the Boussinesq approximation. Therefore, a non-Boussinesq version of the model is being tested now. A tidal model, where tidal forcing at open boundaries is derived from a least squares optimization approach, has been developed and will be implemented in the operational system in the future. PLANS FY95 In addition to the heat and momentum surface fluxes, now included in the forecast system, buoyant fluxes (including river runoffs), tidal and surface pressure forcing will be added to the model running at NMC. At Princeton, considerable effort will be also invested in the improvement of the heat flux formulation and different feedback schemes will be tested. SST and Altimetry data will be available soon in real-time, so data assimilation techniques, previously used only in a research environment will be tested in the operational system. North Atlantic And Arctic Ocean Modeling T. Ezer G.L. Mellor ACTIVITIES FY94 This project is done in cooperation with scientists from Goddard/NASA under the support of the Atlantic Climate Change Program (ACCP). The POM model previously has been applied mostly to coastal and estuaries. However, recently it has been used for basin-scale problems. In particular, models of the Arctic ocean, including the sea-ice interaction and a model of the North Atlantic Ocean are being tested. The Atlantic model has been used for diagnostic and prognostic calculations, aimed at the calculation of the ocean circulation and coastal sea level obtained from observed hydrography and surface wind stress, and produced quite realistic meridional circulation and heat flux. Applying the same model to study the interannual variations between the late 1950s and the early 1970s, the changes in circulation, meridional fluxes and coastal sea level due to changes in thermohaline structure and wind pattern were modeled and compared with other models and observations. This study suggests for example that the contribution to changes in the observed sea level from changes in ocean circulation may be as large as the contribution due to thermal expansion. A new high resolution model that includes the Arctic, the Greenland sea and the Atlantic ocean is being tested now and will be used to study the interaction between those oceans. Special emphasis is put on studying deep water formation in high latitudes and the dynamics of flow over sills; these processes are important factors in climate modeling of the North Atlantic. PLANS FY95 With the aid of Arctic and Atlantic models, the influence of the high latitude dynamics and water formation on the climatology of the Atlantic ocean will be studied. The fluctuations of the North Atlantic Ocean, from seasonal to interdecadal time scales will be also studied. A Study of the Mediterranean Sea G. Mellor M. Zavatarelli ACTIVITIES FY94 An earlier work has undergone major revision to focus on climatological processes governing the Mediterranean Sea. A feature of the model is that, with the help of a curvilinear horizontal grid, the larger scales of the entire Mediterranean Sea are modeled and the topography around the narrow and shallow Straits of Gibraltar is also reasonably well represented. The resulting model inflow and outflow seems to mimic the real Mediterranean, often in considerable detail. Levantine Intermediate water is formed in the Levantine Basin and exits through the Strait of Sicily and the Strait of Gibraltar. Deep water formation in the southern Adriatic Sea are processes clearly represented by the model. Deep water is also formed in the northwestern, shallow region of the western Mediterranean but localized, deep convective mixing in the interior northwestern Mediterranean is absent, this being attributed to the lack of intense winter storms in the climatological surface forcing. The model results indicate that, in the western Mediterranean, the wind stress is very important in establishing the summer northward shift of the Atlantic inflow. Lateral boundary runoff and surface heat and salinity fluxes are necessary to the maintenance of the cyclonic circulation in the northern Balearic basin and enhance the seasonal reversal of the circulation in the Tyrrhenian Sea.