Climatological Observations for Maritime Prediction and Analysis Support Service (COMPASS)

The US Navy faces several limitations when planning operations in regard to forecasting environmental conditions. Currently, mission analysis and planning tools rely heavily on short-term (less than a week) forecasts or long-term statistical climate products. However, newly available data in the form of weather forecast ensembles provides dynamical and statistical extended-range predictions that will produce more accurate predictions if ensemble members can be combined correctly. Charles River Analytics proposes a Climatological Observations for Maritime Prediction and Analysis Support Service (COMPASS), which performs data fusion over extended-range multi-model ensembles to produce a unified forecast for several weeks to several seasons in the future. We will evaluate and select the data using probabilistic programming (PP) and data fusion and association techniques. Further, we will perform state estimation and prediction in an effort to produce an all-encompassing and superior forecast that can be used to inform the Navys decision planning process.