This study investigates the variability and predictability of snow depth an
omalies over the Eurasian continent at the end of winter, as represented in
12 ensembles of General Circulation Model simulations performed at the Eur
opean Centre for Medium-Range Weather Forecasts. Each ensemble includes nin
e integrations performed with the same prescribed sea surface temperature,
but started from time-lagged initial conditions. An empirical orthogonal fu
nction (EOF) analysis shows that the leading EOF of Eurasian snow depth in
March has a zonally-oriented dipole structure, with a band of positive anom
alies covering northern Europe and Siberia, and negative anomalies over cen
tral Europe, the Himalayas and north China. A significant relationship is f
ound between the positive/negative phase of this snow-depth anomaly and war
m/cold El Nino Southern Oscillation events. The positive phase of the snow-
depth EOF1 is associated with a wintertime circulation characterized by a s
trengthening of the westerly winds over Europe and Siberia; in the upper tr
oposphere, this westerly anomaly is accompanied by negative zonal wind anom
alies over Eurasia around 30-40 degrees N and positive zonal wind anomalies
between the equator and 25 degrees N over Africa and south-east Asia. A go
od degree of predictability is found in the snow-related circulation anomal
ies: considering 500-hPa height, 850-hPa zonal wind and 200-hPa zonal wind,
the interannual variations of the ensemble-mean fields show a correlation
of 48%, 56% and 65% (respectively) with the corresponding observed anomalie
s over the eastern half (0 degrees to 180 degrees E) of the northern hemisp
here. The tropical component of the zonal wind anomaly associated with snow
-depth EOF1 is strongly predictable; it shows a marked persistence from win
ter to the early summer, and affects the large-scale circulation over south
Asia in the early and central periods of the monsoon season.