video signal which shows the target outline 

 and prominent features, such as king posts, 

 exhaust stacks, etc. For a target to be imaged 

 it must have first been detected in a surface 

 search mode, as the imaging mode does not 

 operate independently of the other modes. 



The FLAR automatically tracks targets, 

 and calculates target course and speed. This 

 automatic process isfarsuperiorto the manual 

 method of determining the location and move- 

 ment of targets using the grid lines on the 

 SLAR film. However, this process is affected 

 by position errors from the aircraft's inertial 

 navigation system (INS). The planned imple- 

 mentation of the global positioning system 

 (GPS) navigation will substitutionally improve 

 the accuracy of course and speed estimates. 



In 1 991 and 1 993, Ice Patrol conducted 

 two tests of the FLAR's ability to detect ice- 

 bergs (Ezman et al, 1993; and Trivers and 

 f^urphy, 1994). Both tests focused solely on 

 the FLAR navigate mode. These tests indi- 

 cated that the FLAR failed to detect small and 

 medium icebergs at ranges the SLAR has 

 shown a high probability of detection. Pre- 

 sumably this is due to the head-on nature of 

 the FLAR. O'Brien et al. (1 993) demonstrated 

 that the best life-raft detection performance for 

 FLAR was between 350 and 010°R and that 

 the performance dropped off significantly at 

 relative bearings of greater than ±045°R. All 

 the data reported in O'Brien et al. (1993) were 

 collected with the radar in the periscope mode 

 and at altitudes (500 and 1 500 ft) much lower 

 than Ice Patrol altitudes. 



Trivers and Murphy (1994) indicated a 

 slight increase in FLAR iceberg-detection range 

 with altitude and hinted at a decrease in ice- 

 berg detection-range with sea state. 



In a test with HU-25 radars (AN/APS- 

 127 FLAR and AN/APS-131 SLAR), 



Lewandowski et al. (1989) computed much 

 smaller FLAR-only liferaft sweep widths than 

 SLAR-only liferaft sweep widths. This result 

 and HC-1 30 FLAR work all seem to indicate 

 that FLAR is far less efficient at poor radar- 

 reflective target detection. Presumably, this is 

 due in part to the spreading of FLAR power 

 over a much larger beamwidth. The AN/APS- 

 1 35 SLAR has twice as much peak power to 

 azimuthal beamwidth as the AN/APS-137 

 FLAR. The multiple "looks" of the FLAR does 

 not do much good if the radar cannot generate 

 enough power to get a return signal. 



However, Ezman et al (1993), Trivers 

 and Murphy (1994), and operational experi- 

 ence demonstrated that the FLAR is a very 

 strong discriminator, especially between ice- 

 berg and ship. No attempt was made to test 

 the ability of the FLAR to discriminate be- 

 tween various sizes of icebergs. 



Currently, the radar operators have 

 limited experience in imaging icebergs, al- 

 though they have much more experience in 

 imaging vessels. Stationary small fishing 

 vessel identification remains problematic even 

 with FLAR because of their small target area, 

 their lack of tnje motion, and their vertical 

 motion that mimics the wave motion. Much 

 more wori< needs to be done on gaining FLAR 

 identification experience. This would require 

 clear conditions (or surface truth) for operator 

 training feedback. 



COMBINED FLAR/SLAR OPERATIONS 



Ice Patrol could rely solely on FLAR- 

 equipped aircraft, using a as yet, undeter- 

 mined smaller track spacing than the current 

 25 nm spacing. However, a 15 nm track 

 spacing would result in a 40% reduction in 

 search area. This smallertrack spacing would 

 add an extra search day per biweekly period 

 to cover the entire limits of all known ice. 



64 



