Satellite and Skin-Layer Effects on the Accuracy of Sea Surface Temperature Measurements from the GOES Satellites

Abstract

The latest Geostationary Operational Environmental Satellites (GOES) have facilitated significant improvements in the ability to measure sea surface temperature (SST) from geostationary satellites. Nonetheless, difficulties associated with sensor calibration and oceanic near-surface temperature gradients affect the accuracy of the measurements and the estimation and interpretion of the diurnal cycle of the bulk SST. Overall, measurements of SST from the GOES imagers on the GOES-8–10 satellites are shown to have very small bias (<0.02 K) and rms differences of between 0.6 and 0.9 K relative to buoy observations. Separate consideration of individual measurement times, however, demonstrates systematic bias variations of over 0.6 K with measurement hour. These bias variations significantly affect both the amplitude and shape of estimates of the diurnal SST cycle. Modeled estimates of the temperature difference across the oceanic cool skin and diurnal thermocline show that bias variations up to 0.3 K can result from variability in the near-surface layer. Oceanic near-surface layer and known “satellite midnight” calibration effects, however, explain only a portion of the observed bias variations, suggesting other possible calibration concerns. Methods of explicitly incorporating skin layer and diurnal thermocline effects in satellite bulk SST measurements were explored in an effort to further improve the measurement accuracy. While the approaches contain more complete physics, they do not yet significantly improve the accuracy of bulk SST measurements due to remaining uncertainties in the temperature difference across the near-surface layer.