The performance of the Seasat-A Satellite Scatterometer (SASS) in two
semi-enclosed seas, the North Sea and the Mediterranean, is assessed b
y comparing its wind velocity data with measurements from coastal stat
ions, ships, buoys and from the initialization of an atmospheric forec
ast model. For this purpose data generated using the same algorithm as
that employed for routine production of global ocean winds have been
used. Although some of the comparisons shows much greater scatter than
in previous open-ocean validation, which may be due to the poorer qua
lity of the non-satellite data, there is no systematic biases. Open-oc
ean algorithms appear to be applicable to the two regions studied. Exa
mples of wind and wind-stress curl fields obtained from the scatterome
ter are presented for several meteorological situations-an atmospheric
front, flow distortion by orography, oceanic upwelling induced by the
Mistral, and a storm-surge in the Adriatic. These case studies show t
he value of having detailed spatial coverage of winds with high intern
al consistency. The paper concludes with comments on the future use of
scatterometers in semi-enclosed seas.