NIMBUS II HRIR DATA 



Figure 2 is an analog display of the NIMBUS II HRIR data from 

 day and night orbits of 22 Jiane 1966. Local times were approxi- 

 mately 0100 and 1300. The data were obtained on orbits 504 and 

 509/510 (composite) . Both the day and night coverage show a 

 bright cloud bank lying just to the east of the Gulf Stream 

 boimdary. The warm edge of the Gulf Stream water indicated by 

 arrows appears as the dark feature at the western edge of the 

 cloud bank. The western protrusion of the meander of the Stream 

 as shown in the ship survey in Figure 1 can be seen in the 

 central portion of both displays. The colder Slope Water appears 

 as a grey scale intermediate between the warm Gulf Stream water 

 and the cold cloud bank. 



For analysis, the HRIR data were digitized and rectified for 

 display at a scale of 1 : 1,000,000. The data were filtered to 

 remove system noise, and the basic 8km x 9km resolution elements 

 were combined to provide radiation temperature data points in a 

 10 X 8 grid within each 1° latitude- longitude square. The radi- 

 ation temperature data from the grid-print map were manually 

 contoured for surface temperature values. The SST analysis from 

 the HRIR data is depicted in Figure 3. 



Five ranges of radiation temperature are indicated in Figure 3. 

 Within the Gulf Stream/Slope Water mass bounded by the east coast 

 and the cloud bank, radiation temperatures between 18°C and 22°C 

 are unshaded, temperatures between 22°C and 24°C are light shaded; 

 the Gulf Stream boxmdary (24°C to 27°C) is unshaded, and two 

 darker shades of grey are used for the 27°C to 31°C and the 31°C 

 to 33°C. All of the unshaded areas within the less than 24°C 

 water mass of the Slope Water are 22°C. 



CALIBRATION OF NIMBUS II HRIR DATA 



Figure 4 shows that portion of the Gulf Stream boundary 

 defined by ship observations of 15-18 June 1966 and the sea 

 surface temperatures contoured from the ART flight of 21 June 

 1966 that were used to calibrate the NIMBUS II HRIR data of 

 22 June 1966. Sea surface temperatures from the ship track and 

 the ART data were compared with the respective HRIR grid-point 

 data to determine an average bias between the radiation tempera- 

 tures and the ground-truth data, and mean absolute errors were 

 computed between the radiation temperatures and the ship and 

 aircraft data. Because of the geographical uncertainty of the 

 location of the NIMBUS II HRIR scan elements (up to 1° of 

 latitude and longitude according to the NIMBUS II USER'S GUIDE) 

 and of the time lapse between the ship observations and the 

 data orbit, both the ship track and the aircraft flight track 

 were shifted slightly to obtain the best agreement with the 



367 



