Chris Thomas, University of Reading

Using lagged covariances to improve our understanding of circulation in the Atlantic Ocean

Keith Haines, Irene Polo, Jon Robson


The RAPID mooring array, located at 26N in the Atlantic Ocean, has been collecting data for approximately a decade. Data from this array play an important role in improving our understanding of the Atlantic Meridional Overturning Circulation (AMOC). Previous model runs have indicated robust covariances between the AMOC at 26N and density anomalies in high-latitude regions such as the Labrador Sea. These covariances have a significant time lag - often up to several years. The objective of the RAMOC project is to assimilate RAPID data into a high-resolution NEMO-CICE model, taking advantage of the lagged covariances which have been determined from simulation. 4DVar is too expensive and hard to implement for this model therefore ideas from 4DEnVar will be employed. Many additional data sources will also be used in the assimilation and the impact of the RAPID observations on top of these needs to be determined. The RAMOC project ultimately aims to improve the knowledge of the ocean state enabling greater predictability on yearly to decadal timescales. We show results from simple simulation studies in an idealised system and also preliminary results from the NEMO-CICE model.

File available here