A statistical-dynamical analysis of the relation between the Mediterranean storm track and the North Atlantic Oscillation based on wave activity diagnostics

Authors

Abstract

 The North Atlantic Oscillation (NAO) is one of the most prominent modes of low-frequency variability over the Atlantic basin in the Northern Hemisphere. In the past decades, the impact of NAO has attracted increasing scientific interest because the NAO exerts an important impact on the regional climate and weather in the North Atlantic region and adjacent continents. Of particular interest is the impact of NAO on the Mediterranean storm track through which NAO can extend its influence to the climate far downstream including the Middle East and southwest Asia. The problem has been previously studied using the energetics by comparing ensemble averages of the terms involved in the eddy kinetic and available potential energy, where ensemble averages are taken separately over the critical positive and negative NAO months. Such analysis has resulted in certain specific results regarding the behavior of the transient eddies in Mediterranean storm track during the two phases of NAO. For example, the energy flux vectors indicate a stronger source in the central Mediterranean with a stronger sink in the Red Sea and Northeast Africa in the positive NAO. There is, however, a fundamental issue with any energy-based analysis, that is the non-uniqueness way of writing the conversion and flux terms.
As a more powerful diagnostic tool, wave activity conservation law resolves the non-uniqueness issues encountered in dealing with the conversion terms. In this way, wave activity diagnostics proves useful for investigating propagation characteristics of stationary and migratory wave disturbances and their interaction with mean flows, as well as inferring preferred position of emission and absorption of Rossby waves. First introduced for waves defined by perturbation with respect to zonal mean leading to the Eliassen–Palm (EP) diagnostics, the wave activity conservation law has now been extended to other averages as well as to more general definition of waves and mean flows with no resort to averaging. In this study a form of the wave activity and its flux introduced by Esler and Haynes in 1999 is used.
The data used are the NCEP/NCAR reanalysis data covering years 1950–2011 for the winter months from December to February. The critical months are defined on the basis of the monthly index of NAO and grouped in two ensembles of 31 positive and 37 negative NAO months. A critical positive (negative) is considered a month whose monthly NAO index is greater (smaller) than the mean NAO index by more than one standard deviation. The wave activity and the three components of its flux are computed for all days of each winter season, then the averages are taken and the composite maps are prepared for the two ensembles. To investigate the net flux of wave activity to the Mediterranean region, a three-dimensional domain extended vertically from 600 to 200 hPa and horizontally from 15W to 45E in longitude and from 30N to 50N in latitude is selected. For further analysis, the domain thus defined is divided to the three equal subdomains in the west, center and east of the Mediterranean domain.
The main results can be summarized as follows. The connection of the Mediterranean storm track to the north east of Atlantic and north of Europe is stronger in the positive phase of NAO. However, there is a stronger connection of the Mediterranean storm track to the cyclogenesis in the west of the North Atlantic in the negative phase of NAO. In other words, the Mediterranean storm track receives stronger activity from the north and the west in, respectively, the positive and negative phases of NAO. In the upper troposphere, wave activity flux vectors indicate the dominance of anticyclonic (cyclonic) Rossby wave breaking and northward (southward) transfer of momentum in the positive (negative) phase of NAO over the Mediterranean region. In both phases, while the west and east subdomains act as sinks (receivers) of wave activity, the central subdomain acts as a source (emitter). In accordance with the results from energetics, the central Mediterranean acts as a considerably stronger source of wave activity in the positive phase. Overall, results of wave activity analysis confirm those of the energetics. In particular, the southwest Asia is expected to receive a stronger influence from the North Atlantic storm track via the Mediterranean in the positive phase of NAO.
The above results are solely based on the simultaneous analysis of wave activity over the whole North Atlantic and Mediterranean storm tracks as well as the southwest Asia in critical months. It remains to see how such results carry over to the actual episodes of positive and negative NAO with proper time lags. Such analysis is expected to have the potential to lead to some seasonal forecasting capability. 

Keywords