A method has been developed to estimate daily surface fluxes of moment
um and sensible and latent heat over the global oceans using a stabili
ty-dependent bulk scheme. Daily fluxes are computed from daily values
of special sensor microwave imager (SSM/I) surface winds, SSM/I surfac
e humidity, National Centers for Environmental Prediction sea surface
temperatures (SSTs), and European Centre for Medium-Range Weather Fore
casts (SSTs minus 2-m temperatures). Dairy surface specific humidity i
s estimated from the SSM/I water vapor for an atmospheric column and t
he lower 500 m of the planetary boundary layer, using the method of Ch
ou et al. [1995] with two modifications for the extratropical oceans.
The modified method is described using two simple equations. Gustiness
parameterization for the weak winds and convective situations is foun
d to have an insignificant impact on the air-sea fluxes derived from t
he SSM/I data and hence is not included. The SSM/I-radiosonde comparis
on (over the global oceans for the entire annual cycle of 1993) shows
that for a 25-km resolution the instantaneous SSM/I surface humidity h
as a root-mean-square (rms) difference of 1.83 g kg(-1). Daily SSM/I l
atent heat fluxes (and wind stresses) agree well with the flux measure
ments over the western Pacific warm pool, with a bias of 6.2 W m(-2) (
0.0061 N m(-2)), an rms difference of 29.0 W m(-2) (0.0187 N m(-2)), a
nd a correlation of 0.83 (0.86). Monthly results of February and Augus
t 1993 show that the patterns and seasonal variabilities of the SSM/I
surface humidity, latent, and sensible heat fluxes are generally in go
od agreement with those of the Comprehensive Ocean-Atmosphere Data Set
(GOADS) and climatologies derived from ship measurements. The SSM/I s
ensible heat flux is generally within +/-10 W m(-2) of GOADS. However,
the SSM/I latent heat flux is generally larger, especially over the w
intertime trade wind belts. The result is consistent with previous cli
matological studies in that the latent heat fluxes based on ship measu
rements are systematically underestimated.