Are stronger North‐Atlantic southwesterlies the forcing to the late‐winter warming in Europe?

We examine a possible mechanism leading to late‐winter warming, and thus to an early spring in Europe. From the National Centers for Environmental Prediction reanalysis, we extract for the years 1948–99 ocean‐surface winds over the eastern North Atlantic, and air temperatures at the surface T s , and at the 500 hPa level T 500 in late‐winter and spring. T s is extracted at six European locations, all at 50.5°N, ranging in longitude from 1.9°E (northeastern France) to 26.2°E (Ukraine). To quantify the advection of maritime air into Europe, we evaluate for three‐pentad groups the index I na of the southwesterlies at 45°N, 20°W; I na is the average wind speed at this point if the direction is from the quadrant 180–270° (when the direction is different, the contribution counts as zero). In late winter, correlations C it between I na and T s are substantial, up to the 0.6 level in western Europe (but weaker correlations for Poland and Ukraine). C it drops sharply by mid‐March, occasionally taking negative values subsequently. This drop in C it indicates that maritime air advection is no longer associated closely with the surface‐air warming; the role of insolation becomes important, and thus the drop in C it marks the arrival of spring. Correlations C iΔ between I na and our lapse‐rate parameter Δ, the difference between T s and T 500 , indicate that the flow of warm maritime‐air from the North Atlantic into this ‘corridor’ at 50.5°N is predominantly at lower tropospheric level. By computing the best linear fit to I na and T s , the trends for the period 1948–99 are evaluated. The trends are appreciable in the second half of February and the first half of March: for I na , the trends are 0.41 m s −1 and 0.15 m s −1 per decade in pentad groups 10–12 and 13–15 respectively ( I na increased from 1948 to 1999 by 2.10 m s −1 and 0.77 m s −1 ); for T s , the trends for western Germany are 0.36°C and 0.43°C per decade in these two respective pentad groups ( T s in this location increased from 1948 to 1999 by 1.86°C and 2.19°C). Such higher near‐surface temperatures would markedly influence snow‐melt, and thus absorption of insolation by the surface. Our three‐pentad analysis points to the interval from mid‐February to mid‐March as the end of‐winter period in which the southwesterlies over the eastern North Atlantic become stronger and the surface‐air temperatures in Europe rise markedly, the lapse rate becomes steeper, and concurrently the longitudinal temperature gradient between the Somme (France) and the Oder (Germany–Poland border) (about −4°C in 1948 for the 10° longitude distance) is reduced by 0.8°C, i.e. by 20% of its 1948 value. Our thesis, that the observed late‐winter warming and the concomitant advancement of spring in Europe results, at least in part, from stronger southwesterlies over the North Atlantic, merits further investigations. Copyright © 2002 Royal Meteorological Society.