Comparison of the NCDC sea surface temperature datasets: ERSST and OISST

The monthly Extended Reconstruction of Global Sea Surface Temperature (ERSST) analysis is produced based on the ICOADS and sea ice concentrations. The analysis covers a long time span from 1854 to present. Although it includes uncertainty from sampling and other errors, the smoothing methods employed in ERSST have the advantage of filtering out noise so that we can focus on long term temporal trends for climate applications.

The Optimally Interpolated SST (OISST) is generated from both in situ and satellite data including AVHRR infrared satellite SST data and AMSR data using the optimum interpolation method. The OISST product is best used for short time scale applications.

The differences in the purpose and construction these operational SST products are significant and will be emphasized. ERSST is a 2 degree monthly dataset while the OISST is 0.25 degree daily dataset. The ERSST version 3b uses in situ data only while the OISST incorporates both in situ and satellite data. Due to the differences in their input data and spatial-temporal resolution, the two datasets have different advantages and disadvantages in their application to climate change studies. In this presentation, we will compare the two datasets and contrast their application to phenomena of varying temporal and spatial scales.