Do Historical "Zero Depth" Hydrographic Data Contain Useful Information on Climate Trends?

The sea surface temperature observations in the archives of hydrographic surveys are compared with comparable observations from ships-of-opportunity as collected in the familiar COADS data set, with the aim of quantifying the possible biases in the COADS results. The focus is on three compact regions of the North Atlantic. It is found that there is generally a significant cold bias in the reported hydrographic surface temperatures and that the magnitude of this bias varies greatly with location. It is possible that this reflects extrapolated temperatures from depth being reported as surface observations. Until the surface hydrographic data can be sifted to eliminate such suspect values they are unlikely to be of use in defining climatic trends.