Observing System Simulation Experiments (OSSE)

Integrated ocean observation systems (IOOS) have received extensive public and government attention, aiming to predict the response of coastal ecosystems to global climate change, improve the safety and efficiency of maritime operations, mitigate the damages from natural and environmental hazards, and maintain ocean and coastal resources. Building and maintaining an observational network, however, is highly costly. Hence, an optimal design of this system is a critical issue that directly affects whether or not an IOOS will succeed. Observing system simulation experiments (OSSEs), which were used by Charney et al. [1969] for the Global Atmospheric Research Program (GARP), have been adopted for the design of an observing system aimed at improving ocean prediction through the use of data assimilation. Various data assimilation methods were implemented into FVCOM, including a) 4-D nudging, optimal interpolation (OI), and Kalman Filters. Kalman Filters are the most sophisticated statistical approaches and are commonly used for nowcasting/forecasting in the ocean and atmospheric models. The Kalman Filters coded in FVCOM include 1) a Reduced Rank Kalman filter (RRKF), 2) an Ensemble Kalman Filter (EnKF), 3) an Ensemble Transform Kalman Filter (ETKF), 4) an Ensemble Square-Root Kalman Filter (EnSKF), and 5) a singular evaluative interpolated Kalman filter (SEIK). The applications of ensemble Kalman filters to coastal problems were discussed by Chen et al. (2009). This method was also used to conduct  OSSEs in Nantucket Sound (Xie et al., 2011) and Mass Bay  (Xie et al., 2012). The primary objective of the OSSEs in Nantucket Sound is to use data assimilation methods to help design an optimal field measurement plan for the forecast model system in this coastal area. By conducting twin experiments with an ensemble of initial perturbed fields generated using a Monte Carlo approach, we examined the dependence of the success of data assimilation on the memory of the local dynamics system, the impact of ensemble size on the success of OSSE‐based data assimilation experiments, and the optimal design of monitoring sites in this area. The same approach was used to assess the DO monitoring sites in Mass Bay.