Characteristics of the Atlantic Storm-Track Eddy Activity and Its Relation with the North Atlantic Oscillation

Abstract

This study focuses on feedbacks of the high-frequency eddy activity onto the quasi-stationary circulation, particularly with regard to the North Atlantic Oscillation (NAO). The methodology consists of analyzing NCEP?NCAR reanalysis data and sensitivity runs from a high-resolution nonhydrostatic regional model. Consistent with recent studies, results show that the jet displacement characteristic of the NAO phenomenon depends strongly on the dynamics of the synoptic-scale waves and the way they break. Positive and negative phases of the NAO are closely related to anticyclonic and cyclonic wave breaking, respectively. Indeed, the high-frequency momentum flux whose sign is directly related to the type of wave breaking is correlated with the NAO index over the Atlantic. The peak of the momentum flux signal precedes that of the NAO by a few days suggesting that wave breaking is triggering NAO events. Two examples illustrate the significant impact of single storms, in particular those occurring in the east coast of the United States. The wave breaking at the end of their life cycle can suddenly change the NAO index in few days, and as the return to equilibrium takes generally a longer time, it can even affect the sign of the NAO during an entire month. An important issue determining the NAO phase is related to upstream effects. By considering a domain extending from the eastern Pacific to western Europe and by forcing the regional model with real data at the western boundary, sensitivity runs show that the right sign of the NAO index can be recovered. It indicates that waves coming from the eastern Pacific are crucial for determining the NAO phase. According to their spatial scales and frequencies when they reach the Atlantic domain, they can break one way or another and push the Atlantic jet equatorward or poleward. Synoptic waves with periods between 5 and 12 days break anticyclonically whereas those with periods between 2 and 5 days break both anticyclonically and cyclonically with a predominance for cyclonic wave breaking. Another crucial factor concerns surface effects. Cyclonic wave breaking in the upper levels is strongly connected with an explosive cyclonic development at the surface accompanied by strong surface moisture fluxes whereas such an explosive growth is not present in the anticyclonic wave breaking case. Finally, it is proposed that these results are not only useful for explaining the intraseasonal variations of the NAO but would serve also as a basis for understanding its interannual and interdecadal variations.