A modelling and observational study of the relationship between sea surface temperature in the North‐West atlantic and the atmospheric general circulation

Abstract

AbstractResults from four pairs of 50‐day wintertime integrations of the Meteorological Office's 5‐level general circulation model, with warm and cold sea surface temperature (s.s.t.) anomalies of about 3 K in the northwest Atlantic, are described. Difference fields between the warm and cold integrations are statistically significant at the 1% level with positive geopotential height over the central north Atlantic, and weaker negative height over Europe. the storm track over the Atlantic is displaced from its normal position. Results from four further pairs of integrations with halved s.s.t. anomalies are also described. the response is approximately linear, with systematic differences in 500 mb geopotential height over the Atlantic, parts of which are just significant at the 10% level with half the full s.s.t. anomaly. Overall, however, the model's response is weaker than could be obtained with tropical s.s.t. anomalies of the same magnitude.Results from the model integrations are compared with results from an observational study of the relationship between wintertime s.s.t. in the north‐west Atlantic, and mean sea level pressure and 500mb height. Two independent 30‐year periods were chosen for study, thus minimizing the influence of long‐term trends in s.s.t. Over the Atlantic and Europe the model results compared well with the observations. With s.s.t. data lagging the atmosphere by one month, the observational study appears to show that the s.s.t. anomalies are initially forced by perturbations in the atmospheric circulation. With s.s.t. data leading the atmosphere by one month results show that atmospheric and s.s.t. anomalies are most persistent in the period October to December. Throughout the winter these lagged relationships are much weaker and not statistically significant.Diagnostics of E‐vector divergence from the GCM experiments are used to suggest that anomalous baroclinic wave activity over the Atlantic is important as a momentum forcing for the anomalous time‐mean flow pattern. On the other hand, the role of thermal forcing, provided by anomalous diabatic heating and transient eddy heat flux convergence, may be important. to substantiate this statement, a simple linear steadystate two‐layer model of the response to extratropical thermal forcing is described. With a suitable basic state flow, and a mid tropospheric heat source (given mainly by the transient eddy heat flux convergence), the response is shown to be equivalent barotropic with a downstream ridge and ascent over the thermal source.Conversely, results from an ocean mixed layer model are discussed which suggest that warm s.s.t. anomalies could be maintained by a positive surface pressure perturbation positioned downstream of the anomaly, through anomalous northward advection of warm ocean water by Ekman drift currents. This northward advection would balance the sensible and latent heat loss into the atmosphere over the s.s.t. anomaly. Hence it is possible that some positive ocean‐atmosphere feedback may account for the persistence of such atmospheric and oceanic anomalies.