This GOES SST product is produced and disseminated in close to real-time at the LSU Earth Scan Lab (www.esl.lsu.edu). The input file is GOES-16 and GOES-19 calibrated brightness temperatures in channel 7 (~4 micron) obtained from the Google Cloud CLI. All night-time images (with 5 minute frequency) over a ~10-hour period are subjected to a warmest pixel composite technique which screens out much cloud contamination. The resultant brightness temperature image is converted to SST in degrees C using a Gulf of Mexico (now America) regression model with R2 of 0.99 and RMS error of 0.1° C.
The ABI technique was developed by Alaric Haag and Nan Walker in 2019, using Gulf NDBC buoy temperatures (top 1m only) and single pixel point measurements, matching the buoy locations. The technique was developed using GOES GVAR and first described in Walker et al, Geophysical Research Letters, 2003.
One of the major advantages of this product is to monitor SST fronts, even in summer, and thus is a practical method for monitoring rapid changes within the Loop Current system. At times, residual cold clouds remain in the night-time image. Thus, a cloud-masked SST image was also produced using the cloud criteria provided by NOAA NESDIS within the GOES ABI file and described in their Algorithm Theoretical Basis Document entitled "ABI Cloud Mask" by Heidinger and Straka version 3.0, July 2012. The cloud mask we applied at each pixel location was coded "cloudy" or "probably cloudy".
For further information, contact Nan Walker at nwalker@lsu.edu