Have mid-latitude ocean rain-lenses been seen by the SMOS satellite?

Abstract In winter in mid-latitudes the ocean mixed layer is typically a few hundred metres deep because of intense surface cooling and wind mixing. However, it is shown here that the SMOS (Soil Moisture and Ocean Salinity) satellite has detected 3–4 instances per day (averaged over the globe) of anomalous bands of surface fresh water under atmospheric fronts. One typical case shows a fresh surface anomaly of 4 psu located southwest of Australia under an atmospheric front with rainfall of 5.6 mm/day. In this case the size of the salinity anomaly and the ECMWF rainfall rate along the front imply that the rainwater is staying within the upper 15 cm of the ocean, despite mixing from the frontal winds. If these lenses are confirmed, they are significant because a surface that is 4 psu fresher, can be 1 K cooler and still be stable, and this could reduce sea to air heat fluxes by 35 W/m 2 . If an air–sea coupled model has a low vertical ocean resolution, and is unable to model these shallow lenses, this may result in incorrectly high sea to air heat fluxes, and the model troposphere would warm unrealistically by 1.3 K over a 5 day forecast. For a global climate model the error would be 4.4 K of tropospheric heating per decade.