A Working Paper

Rapid Environmental Assessment (REA) (Pouliquen et al., 1997) is defined in the military environment as "the acquisition, compilation and release of tactically relevant environmental information in a tactically relevant time frame" (Sellschopp, 1999a). Tactically relevant time scales range from several months in an early operational planning phase down to days during operations. REA is a necessity for military operations when critical environmental information is unavailable to the military planner. REA is ideal for the distribution of information that is valid for a limited time, such as transient ocean circulation patterns or other time-varying phenomena. REA adds synoptic real-time environmental knowledge to climatological data bases, or, in a broader sense, it adds 'weather' to 'climate'.

Ocean forecasting is essential for effective and efficient operations on and within the sea, and such forecasting has been initiated, e.g. for military operations, coastal zone management and scientific research. Observations are used to initialize dynamical forecast models, and further observations are continually assimilated into the models as the forecasts advance in time (Robinson et al., 1996a). Such observations are generally difficult, costly and sparse. If a region of the ocean were to be sampled uniformly over a predetermined space-time grid, adequate to resolve scales of interest, only a small subset of those observations would have significant impact on the accuracy of the forecasts. The impact subset is related to intermittent energetic synoptic dynamical events. For most of the energetic variability in the ocean, the location and timing of such events is irregular and not a priori known. However, a usefully accurate forecast targets such events and forms the basis for the design of a sampling scheme tailored to the ocean state to be observed. Such adaptive sampling of the observations of greatest impact is efficient and can drastically reduce the observational requirements, i.e. by one or two orders of magnitude. The development of a regional forecast system and capability depends both upon the scales and processes of interest and the scales and processes that are dominant in the region (Robinson and Glenn, 1999).

A Coastal Predictive Skill Experiment (CPSE) (Curtin et al., 1993) is an exercise involving both a model and observations in a specific coastal domain. Observations are used to initialize the model, which is then used to forecast future states of the system. Subsequent observations are used to test the accuracy of the forecast and may also be assimilated to keep the model on track. These tests typically take the form of differences and of correlation or coherence in one, two or three spatial dimensions (Miller et al., 1995; Robinson et al., 1996b; Onken, 1998). The accuracy of the forecast and the trend of that accuracy over time determine the skill of the model. Presumably skill with data assimilation will be greater than without. The relationship between forecast error and forecast time (error growth over the domain (Lermusiaux, 1999a; Lermusiaux and Robinson, 1999)) is of particular interest. A REA predictive skill experiment must be designed to determine forecast skill on the basis of minimal and covertly attainable observations and thus may be most efficiently carried out in the context of the definitive over-sampling provided by a CPSE.

For about a decade, there has been a fruitful collaboration between the Harvard Ocean Modeling Group and the Oceanography Department of the SACLANT Undersea Research Centre. Most recently, the partners have combined their skills in ocean data sampling and modeling during Rapid Response. The Harvard Ocean Prediction System, its initialization and the adaptive sampling program have found exceptional attention by navy staff (Hammond, 1997). Rapid Response has been a three year long series of demonstrations of rapid environmental assessment capabilities under the umbrella of NATO's military oceanography (MILOC) organization. It took place in the Sicilian Channel, the Ionian Sea and the Gulf of Cadiz, which are challenging ocean areas because of their complex bathymetric, thermohaline and dynamical conditions. The successful Harvard/SACLANTCEN collaboration has, thus far, been carried out in the context of real-time operations (Sellschopp and Robinson, 1997; Robinson, 1997). The experiments for the project proposed here will be designed as definitive predictive skill experiments.

Advanced ocean observing and prediction systems (OOPS) now exist for field estimation. An OOPS consists of an observational network, data analysis and assimilation schemes and a suite of interdisciplinary dynamical models. The Littoral Ocean Observing and Prediction System (LOOPS) is an advanced OOPS, which has been addressing some of the fundamental issues associated with the development of the multi-disciplinary observation, assimilation and modeling forecasting concept for coastal processes. The LOOPS system is modular, based on a distributed information concept, utilizing a network-based infrastructure for heterogeneous data and computation elements, providing sharable, scalable, flexible and efficient workflow and management for interdisciplinary data collection, assimilation and forecasting (Patrikalakis et al., 1999). The Harvard Ocean Prediction System (HOPS) (Robinson, 1997) is at the heart of LOOPS. The LOOPS partnership has also developed a unique, coupled physical/acoustical assimilation formulation with consistent handling of model and measurement errors. These joint efforts have established LOOPS as a mature multi-disciplinary partnership uniquely qualified to develop and implement a fully integrated generic, multi-scale, multi-disciplinary ocean forecasting capability.

Two coastal predictive skill/rapid environmental assessment experiments are proposed to be carried out as a joint research project (JRP) between Harvard University and SACLANT Undersea Research Centre and possibly additional partners. The experiments should take place in the years 2001 and 2002 in ocean areas that are fairly well known. An important objective of the experiments will be to investigate the performance of adaptive sampling concepts and discipline-coupled data assimilation concepts for littoral ocean forecasting under realistic environmental conditions. Suggested locations are: the areas of Massachusetts Bay/Gulf of Maine (where a LOOPS/AFMIS/AOSN interdisciplinary, multi-scale sea trial was carried out in 1998 (Robinson and Glenn, 1999)); and the transition area between Western and Eastern Mediterranean See from the Strait of Sardinia to the western Ionian Sea (where Rapid Response experience has been gained) (Sellschopp, 1999b; Lermusiaux, 1999b). The experiments should cover appropriate areas for model initialization and updating and nested high resolution sub-regions.

The predictive skill experiment should be designed as a real time verification experiment with quantitative skill measures. Adaptive sampling planning should be automated, i.e. the requirements for updated hydrographic measurements or current vectors obtained from ships, AUVs or aircraft should be the result of sensitivity calculations rather than being established intuitively. Because of adaptive sampling, there are minimum requirements for in situ data acquisition. Oversampling is however required for rigorous verification of the modeling results. By comparison with the over-sampled data set, it will be possible to design optimal sampling for given accuracy requirements.

The rapid assessment experiment should provide the best possible predictions as stated above. Operational REA suffers from limited data and restrictions in platform deployments. In a realistic REA experimental scenario, a REA modeling team will be given the opportunity to define the measurement program of a ship, but only for a fraction of its time at sea. The rest of the ship time will be used to acquire a complete data set, adequate for unconstrained forecast validation. A second modeling team will make use of the full data set and produce more accurate real-time forecasts. For the value of forecasts under data sampling restrictions, answers given by this experiment are considered to be more realistic than forecasts from an a posteriori reduction of a rich data set in part, and importantly, because of the demanding conditions under which the forecast team must perform.

Curtin, T., J.G. Bellingham, J. Catipovic and D. Webb, 1993. Autonomous Ocean Sampling Networks, Oceanography, 6 (3), 86-94.

Hammond, N., 1997. RAPID RESPONSE 96 - A Demonstration of NATO's Rapid Environmental Assessment Capability, in Rapid Environmental Assessment (E. Pouliquen, A.D. Kirwan and R.T. Pearson, eds.), pp. 21-23.

Lermusiaux, P.F.J., 1999a. Data assimilation via Error Subspace Statistical Estimation, Part II: Mid-Atlantic Bight shelfbreak front simulations and ESSE validation, Monthly Weather Review, 127 (7), 1408-1432.

Lermusiaux, P.F.J., 1999b. Estimation and study of mesoscale variability in the Strait of Sicily, Dynamics of Atmospheres and Oceans, to appear.

Lermusiaux, P.F.J. and A.R. Robinson, 1999. Data assimilation via Error Subspace Statistical Estimation, Part I: Theory and schemes, Monthly Weather Review, 127 (7), 1385-1407.

Miller, A.J., H.G. Arango, A.R. Robinson, W.G. Leslie, P.-M. Poulain and A. Warn-Varnas, 1995. Quantitative skill of quasigeostrophic forecasts of a baroclinically unstable Iceland-Faeroe Front, Journal of Geophysical Research, 100 (C6), 10833-10849.

Onken, R., 1998. Nowcasting and forecasting of sound velocity structure by ocean modeling,

SACLANTCEN Report SR-292, SACLANT Undersea Research Centre, La Spezia, Italy, 1998.

Patrikalakis, N.M., P.J. Fortier, Y.E. Ioannidis, C.N. Nikolaou, A.R. Robinson, J.R. Rossignac and A. Vinacua, 1999. Distribution Information and Computation in Scientific and Engineering Environments, D-Lib Magazine, 5(4), http://www.dlib.org/dlib/april99/04abrams.html

Pouliquen, E., A.D. Kirwan, R.T. Pearson, editors, 1997. Rapid Environmental Assessment, Proceedings of a conference held in Lerici, Italy, 10-14 March 1997, SACLANTCEN Conference Proceedings Series CP-44.

Robinson, A.R., 1997. Forecasting and Simulating Coastal Ocean Processes and Variabilities with the Harvard Ocean Prediction System, in Rapid Environmental Assessment (E. Pouliquen, A.D. Kirwan and R.T. Pearson, eds.), pp. 187-198; also in Coastal Ocean Prediction, (C.N.K. Mooers, ed.), AGU Coastal and Estuarine Studies Series, American Geophysical Union, pp. 77-100.

Robinson, A.R. and S.M. Glenn, 1999. Adaptive sampling for ocean forecasting, Naval Research Reviews, to appear. Available via the WWW at /PAPERS/Naval_Review.html

Robinson, A.R., H.G. Arango, A. Warn-Varnas, W.G. Leslie, A.J. Miller, P.J. Haley and C.J. Lozano, 1996a. Real-time regional forecasting, Modern Approaches to Data Assimilation in Ocean Modeling (P. Malanotte-Rizzoli, ed.), Elsevier Oceanography Series, Elsevier Science, The Netherlands, 377-412.

Robinson, A.R., H.G. Arango, A.J. Miller, A. Warn-Varnas, P.-M. Poulain and W.G. Leslie, 1996b. Real-time operational forecasting on shipboard of the Iceland-Faeroe frontal variability, Bulletin of the American Meteorological Society, 72 (2), 243-259.

Sellschopp, J., 1999a. Rapid Environmental Assessment, Naval Forces, International Forum for Maritime Power, June Æ99, to appear.

Sellschopp, J., 1999b. Experience gained and lessons learnt in three years of Rapid Response,

UDT 1998, 23-25 June 1998, London England, Conference Proceedings, 69-73.

Sellschopp, J. and A.R. Robinson, 1997. Describing and Forecasting Ocean Conditions during Operation Rapid Response, in Rapid Environmental Assessment (E. Pouliquen, A.D. Kirwan and R.T. Pearson, eds.), pp. 35-42.