SST ANALYSIS PRODUCTS 



Figures 5, 6, and 7 present the SST analysis products for 

 22 June 1966 as prepared by FNWC, FWC (Norfolk), and NAVOCEANO, 

 respectively. The Gulf Stream boundary, as defined by the ship 

 data during 15-21 June, is superimposed on each of the figures 

 to provide a frame of reference. All of the analysis products 

 are based upon the standard ship reports of sea surface tempera- 

 ture received during a 3 1/2- to 7- day interval preceding the 

 analysis time of 1200Z, 22 June 1966. 



The FNWC analysis is produced from two adjacent 10 x 10 

 degree latitude- longitude fields derived from ZOOM analyses of 

 63 X 63 point grids, using ship data received during a 3 1/2-day 

 interval. The FWC analysis is a manual subjective analysis of 

 the input data field. The NAVOCEANO analysis was done by a 

 developmental, large-scale numerical technique, using 1180 

 standard ship observations received during a 7-day interval. 



The sea surface temperature gradients associated with the 

 Gulf Stream boundary are much weaker in the standard analysis 

 products than those determined from the satellite data and 

 confirmed by the ship and flight data. The limitation of the 

 standard products lies in the data density in the input data 

 field. The grid-point density of the FNWC analysis field (approxi- 

 mately 10-mile spacing) is very close to that of the satellite 

 HRIR grid-print map (approximately 6-mile spacing). The major 

 difference is that there are not data entries at each of the FNWC 

 grid points as is the case for the clear-sky portions of the 

 grid-print map. Table 3 presents a comparison of the SST gradients 

 associated with the Gulf Stream boundary as determined from the 

 satellite and the FNWC, FWC, and NAVOCEANO analyses, 



PLANS FOR FUTURE WORK 



This study has demonstrated the potential of utilization of 

 satellite SST data for depiction of major features of the ocean 

 surface temperature structure. As the problems of instrument 

 noise, geographic position error, atmospheric absorption, and 

 cloud contamination are addressed, significant quantities of 

 HRIR SST data will become available as data input for SST 

 analyses. 



The next step in this investigation will be, with the cooper- 

 ation of FNWC, Monterey, to submit the corrected digitized HRIR 

 grid-print data as the input data field for FNWC analysis of the 

 SST pattern for 22 Jime 1966. It is expected that this test will 

 determine the capability of the FNWC analysis program for 

 characterization of the true SST field when provided with a 

 dense input of satellite data. 



369 



