By Sabrina Matteucci
Surface drift prediction in the Adriatic Sea using hyper-ensemble statistics on atmospheric, ocean and wave models: uncertainties and probability distribution areas. NURC-FR-2006-012. May 2006.
An increasing number of current models routinely provide weather forecasts and climate predictions, offering multiple options on resolutions, range, domains and derived fields. NATO requirements include reliable tactical knowledge and forecasts of the sea surface components, where potential mine threats have to be mitigated or avoided and where search and rescue efforts have to be optimized. These issues become more challenging and relevant when considering support for Expeditionary Warfare (in remote areas with limited access) and countering naval asymmetric warfare (need for high accuracy and reliability). The surface drift is the resultant of many different direct and indirect contributions of the atmosphere, the ocean and the sea surface itself. However, the prediction of the surface drift resultant still remains a challenge when the different components have competing contributions, like in coastal or near-shore areas. One of the possible solutions to address these issues is to migrate from the traditional deterministic approaches towards probabilistic-stochastic methodologies. When multiple models and data become available, the envisaged probabilisticstochastic alternative is the multi-model super-ensemble technique which uses an optimized combination of an ensemble of models. This technique has previously been demonstrated to improve forecast skills in the atmospheric and is applied here to the prediction of surface drift in the Adriatic. The technique combines optimally an atmospheric, an ocean and a wave model and is shown to outperform traditional forecast methods.