350 



REFRACTION AND REFRACTIVE INDEX MODELS 



C/5 



ttr 



LlI 



1.5 



TARGET HEIGHT =3.046 km 

 RANGE =25, 1 km 

 ELEVATION ANGLE = 7° 



STANDARD ERROR 

 OF ESTIMATE 



J I \ \ L 



340 



350 360 370 



Ng-PUUNENE AIRPORT U.S.W.B. 



380 



Figure 8.24. Range error fluctuations observed on Maui path. 



The small discrepancy between the intercepts (i.e., between the mean 

 refraction) of the observed and predicted e versus A'^ « lines may be perhaps 

 attributed to, for example, antenna lobe pattern distortion caused by 

 differential refraction, or defocusing [22]. 



The remaining data which are examined were of necessity taken in 

 such a manner as to have a rather high degree of autocorrelation (trends). 

 Such data are not as suitable for confirming the accuracy of a regression 

 prediction process as are independent data. A discussion of this is in- 

 cluded at the end of this section. 



Turning to examination of radio range errors, figure 8.24 shows the 

 results of some measurements of apparent radio range fluctuations over a 

 25-km path on the island of Maui, Hawaii, on November 9-11, 1956 [23]. 

 These measurements were made at 1-hr intervals, and are essentially 

 "instantaneous" values. The target beacon was situated on the summit 

 of Mount Haleakala at an elevation of 10,025 ft, while the "ground" 

 station was near Puunene Airport at an elevation of 104 ft, thus yielding 

 a target height of 3.046 km, in a region of critical target heights for pre- 

 diction of radio range errors in tropical climates [19]. The measured 

 range fluctuations (absolute errors not measured) are plotted against 

 values of N s taken at about the same time (mostly 15 to 20 min later) by 



