An Improved In Situ and Satellite SST Analysis for Climate

A weekly 1° spatial resolution optimum interpolation (OI) sea surface temperature (SST) analysis has been produced at the National Oceanic and Atmospheric Administration (NOAA) using both in situ and satellite data from November 1981 to the present. The weekly product has been available since 1993 and is widely used for weather and climate monitoring and forecasting. Errors in the satellite bias correction and the sea ice to SST conversion algorithm are discussed, and then an improved version of the OI analysis is developed. The changes result in a modest reduction in the satellite bias that leaves small global residual biases of roughly −0.03°C. The major improvement in the analysis occurs at high latitudes due to the new sea ice algorithm where local differences between the old and new analysis can exceed 1°C. Comparisons with other SST products are needed to determine the consistency of the OI. These comparisons show that the differences among products occur on large time- and space scales with monthly rms differences exceeding 0.5°C in some regions. These regions are primarily the mid- and high-latitude Southern Oceans and the Arctic where data are sparse, as well as high-gradient areas such as the Gulf Stream and Kuroshio where the gradients cannot be properly resolved on a 1° grid. In addition, globally averaged differences of roughly 0.05°C occur among the products on decadal scales. These differences primarily arise from the same regions where the rms differences are large. However, smaller unexplained differences also occur in other regions of the midlatitude Northern Hemisphere where in situ data should be adequate.