Daily fields obtained from a 17-yr atmospheric GCM simulation are used
to study the surface sensible and latent heat flux variability and it
s relationship to the sea level pressure (SLP) field. The fluxes are a
nalyzed over the North Pacific and Atlantic Oceans during winter. The
leading mode of interannual SLP variability consists of a single cente
r associated with the Aleutian low in the Pacific, and a dipole patter
n associated with the Icelandic low and Azores high in the Atlantic. T
he surface flux anomalies are organized by the low-level atmospheric c
irculation associated with these modes in agreement with previous obse
rvational studies. The surface flux variability on all of the timescal
es examined, including intraseasonal, interannual, 3-10 day, and 10-30
day, is maximized along the north and west edges of both oceans and b
etween Japan and the date line at similar to 35 degrees N in the Pacif
ic. The intraseasonal variability is approximately 3-5 times larger th
an the interannual variability, with more than half of the total surfa
ce flux variability occuring on timescales of less than 1 month. Surfa
ce flux variability in the 3-10-day band is clearly associated with mi
dlatitude synoptic storms. Composites indicate upward (downward) flux
anomalies that exceed \30 W m(-2)\ occur to the west (east) of storms,
which move eastward across the oceans at 10 degrees-15 degrees per da
y. The SLP and surface flux anomalies are also strong and coherent in
the 10-30-day band but are located farther north, are broader in scale
, and propagate similar to 3-4 times more slowly eastward than the syn
optic disturbances. The sensible and latent heat flux are proportional
to the wind speed multiplied by the air-sea temperature and humidity
difference, respectively. The anomalous wind speed has the greatest in
fluence on surface flux anomalies in the subtropics and western Pacifi
c, while the air temperature and moisture anomalies have the greatest
impact in the northeast Pacific and north of 40 degrees N in the Atlan
tic. The covariance between the wind speed and the air temperature or
humidity anomalies, while generally small, is nonnegligible on synopti
c timescales.