Observations of the oceanic thermal skin made during SUDOFEX have demonstra
ted the effectiveness of operationally using new shipborne radiometers to s
how the correspondence between in situ radiometric skin sea surface tempera
ture (skin SST) and the present AVHRR sea surface temperature (pseudo bulk
SST) products. A recommendation is made that algorithms should be derived t
hat produce skin SST by regression against in situ skin SST rather than sub
surface thermometer data. This will eliminate SST errors arising from both
the oceanic thermal skin effect and the diurnal thermocline.