SST processing system
The SST processing procedure include five components: SST retrieval everyday 8-time, calculating daily mean, five-day mean, ten-day mean monthly mean SST, products verification, and output of products. Processing procedure is shown Fig.1.
Fig.1 FY-2 SST Processing Procedure
SST retrieval everyday 8-time
SST is retrieved at each 0.5° latitude / longitude grid point within the area from 50° S to 50° N and from 55° E to 155° E every 3 hours. This component includes, 1) image coordinate transform, 2) Calculating satellite zenith and solar zenith angle, 3) Converting digital count values of IR to brightness temperatures with calibration table, 4) Performing Cloud detection and Atmospheric Attenuation correction.

Calculating daily mean SST
The step is SST comparison test everyday 8 times, 8-time SST Merged and calculating daily mean SST

Calculating five-day mean, ten-day mean and monthly mean SST
Five-day mean SST are calculated from daily mean SST, ten-day mean SST are calculated from five-day mean values, and monthly mean SST are calculated from ten-day mean values.

Products verification
Product verification is to determine the accuracy of satellite-derived SST products. The SST retrievals from the above processing are compared with the conventional SST measurements from GTS. Calculating root mean square (RMS) and bias.

Output of products
Archival tape, print, image display, contour chart, and Distributing

Calculation result
Daily mean, five-day mean, ten-day mean SST products were obtained, covering the area from 50° N to 50° S and 55° E to 155° E. Fig.2 shows Ten-day SST pseudo color image in April, 1999.


Fig.2 Ten mean SST pseudo color image of FY-2 in April, 1999
Discussion
FY-2 SST processing system can obtain more frequently SST, so that probability of obtaining cloud-free data is enhance. But the radiometer aboard FY-2 satellite has only one IR channel, the accuracy of Atmospheric Attenuation correction is not higher. The same results show RMS is about 0.9° C-2.2°C, we will improve algorithm.

References

• McClain, E. P., Pichel, W. G., and Walton, C. C., 1985, Comparative performance of AVHRRbased multi-channel sea surface temperature, Journal of Geophysical Research, Vol. 90, No. C6, 1158711602.
• Qiu Kangmu, 1987, statistical estimation method for retrieval sea surface temperature, Atmospheric Radiation progress and prospects, p.699.
• Chen Weiying, 1998, Sea Surface Temperature Retrieval From Geostetionary Meteorological Satellite (GMS-5) Infrared Split-Window Channels, Proceedings of SPIE Optical Remote Sensing of the Atmosphere and Clouds. , Vol.3501.p484.