( )f particular interest is the test-run made on a small iceberg (33 by 247 

 feet) during which six well documented fade areas were passed through. 

 The results of power measurements on this berg are shown in figure 34, 

 and the results of a Wiresonde sounding taken immediately after the 

 power measurements are shown in figure 35. From these observed tem- 

 perature and relative humidity distributions, and the resultant index of 

 refraction variation compared to that for standard conditions, it is im- 

 mediately apparent that subnormal conditions existed. To further in- 

 vestigate this phenomena, a ray diagram (fig. 36) was constructed graph- 

 ically using a method similar to that developed by Anderson and Abbott 

 [/£]. Accordingly, the maximum detection range to be expected for this 

 target is approximately 17,000 yards; whereas, the maximum range ex- 



(N- I) ■ I0 4 



Figure 35. — Vertical distribution of temperature, relative humidity, and refractive 

 index observed by Wiresonde in the vicinity of iceberg of figure '.\\. 



pected for the standard conditions for a target of this height is 30,000 

 yards. The range of iceberg detection was actually 19,600 yards and this 

 increased range of detection over that expected from the ray diagram is 

 attributed mostly to diffraction. Also shown diagrammatically in figure 

 36 is the space comparison between the actual fades (fig. 34) and the fades 

 expected from destructive interference phenomena deduced from the ray 

 diagram. The solid wedge areas indicate the approximate range of theore- 

 tical fades. The shape of the expected fade curves has not been com- 

 puted. Smooth sea conditions prevailed during the test-run. There was 

 indication of sharp fades during other test-runs when the sea surface was 

 smooth and this is in accordance with what one would expect by consider- 

 ation of the surface roughness and its effect on reflection. 



78 



