In situ calibration of satellite sea surface temperature

The infrared SST signature is representative of only the upper few millimeters of the ocean. Thus, there is a fundamental problem in using in-situ SSTs to validate/calibrate satellite SST estimates. In spite of this difference drifting buoy SSTs continue to be the main source of calibration information used with the infrared satellite data. In this paper we try to understand the level of accuracy one can expect to have with the various candidate in situ SST measurements. Ignoring for now the basic difference between the satellite “skin” SSTs and the buoy “bulk” SSTs, we want to examine the characteristics of the buoy SSTs and evaluate their use as in situ calibration for satellite infrared SST measurements. We submit ship SSTs to the same analysis along with moored buoy SSTs. An important statistic is the SST difference as a function of distance/time to define just when/where the buoy SST could be used for calibration/validation. We then compare two different “buoy-calibrated” satellite SSTs with the drifting buoy SSTs along with a skin-SST algorithm