BACKGROUND 175 



approximately exponential decrease of n with respect to height. Schel- 

 leng, Burrows, and Ferrell [15] attempted to remove this systematic n 

 decrease with height by utilizing a correction factor of linear form, 

 although, as has been seen in chapter 3, this leads to serious overestima- 

 tion of refraction effects in the more modern problem of satellite tele- 

 communications. 



The work of Hay [6] confirmed and extended the observations Gerson 

 had made concerning air mass profiles. Measuring air mass character- 

 istics at Maniwaki, Quebec, Hay concluded that each large-scale air mass 

 type in central Canada has a distinctive refractive index profile. Hay 

 determined the height distribution of A'' for four basic air masses by fitting 

 a second-degree polynominal to each of the four sets of air mass data, 

 indicating that an average A'' profile cannot be approximated effectively 

 by a linear curve except over small height increments. Hay obtained 

 further discrimination by constructing a "dry-term" curve for each air 

 mass. The dry-term curves display a smaller standard deviation than 

 the total A'^ curves for all air masses except the continental arctic which 

 is, of course, itself very dry. The largest variations in total A'' are due 

 to fluctuations in the wet term. The saturation vapor pressure is approx- 

 imately an exponential function of the temperature, so that during the 

 warm season normal temperature changes cause the saturation vapor 

 pressure and, therefore, the wet term is vary sharply. 



Contained also in Hay's paper [6] is a discussion of the use of A^ pro- 

 files to estimate corrections for radio ray refraction by use of a table of 

 mean effective earth's radius factors for each of four air masses within 

 95 percent probability limits. 



Misme [5, 16] has had an interest in synoptic radio meteorology in con- 

 nection with telecommunications networks in France and North Africa. 

 Interests of this French radio engineer include vertical gradients of the 

 refractive index and atmospheric reflection of radio waves. 



A reasonable question at this point is, to what extent are model ex- 

 ponential atmospheres applicable to various climatic zones around the 

 earth? While it is readily apparent that individual profiles vary widely 

 from any sort of exponential norm, there is an increasing backlog of 

 experimental evidence [17] that shows that long-term averages of synoptic 

 refractive index variations do tend toward an exponential form with 

 respect to height. Indeed, the 5-year mean A'^ profiles of figures 5.1 and 

 5.2, representing arctic and tropical climates respectively, bear out this 

 contention by showing close agreement with models developed from mid- 

 latitudinal data. Indeed, these examples verify the results obtained by 

 Gerson [2] for various air masses. 



Randall [4] investigated the relationship of surface meteorological data 

 to surface A'^, N s, and field strength in the FM frequency band. The 

 results described were drawn from a very limited data sample that covered 



