Advanced modeling and data assimilation methods for the design of sustained marine monitoring networks

Abstract

An impact assessment of a Fishery Observing System (FOS) network in the Adriatic Sea was carried out with an ocean circulation model fully-coupled with a data assimilation system. The FOS data are single point vertical values of temperature collected in 2007. In this study, we used the Observing System Experiment (OSE) and Observing System Simulation Experiment (OSSE) methodologies to estimate the impact of different FOS design and sensors implementation. OSEs were conducted to evaluate real observations and they show that the FOS network improves the analysis significantly, especially during the stratification season. Root mean square (RMS) of temperature errors are reduced by about 44% and 36% in the upper and lower layers respectively. We also demonstrated that a similar impact can be obtained with a reduced number of vessels if the spatial coverage of the data points does not change significantly. In the OSSE, the impact of the implementation of a CTD (conductivity-temperature-depth) sensor in place of the existing temperature sensor was tested with identical twin approaches between January and April 2007. The results imply that the assimilation of salinity does not improve the analysis significantly during the winter and spring seasons.