PRESENTATION OF A^ DATA 13 



Assuming errors of it 0.1 mbar for P and ±0.25 °C for T and T' (de- 

 rived from considering Wadley's field data as being accurate to ±5 units 

 of the next significant figure beyond the tabulated whole degrees Fahren- 

 heit for wet and dry bulb temperatures and hundredths of an inch of 

 mercury for total air pressure) one finds that errors in A^ may be as large 

 as 2.75 N units, which is much greater than the 0.5 N unit difference given 

 by the two formulas for N. Again it is seen that the difference in the 

 constants used in the two most widely accepted expressions for N yields 

 an error of the same size or smaller than that produced by errors of meas- 

 urement of the necessary meteorological parameters. 



1.5. Presentation of N Data 



There is now, in the literature, an almost bewildering choice of modi- 

 fications to basic N data when presented as a function of height. The 

 underlying principle is always to remove the syste?natic decrease of N with 

 height, h, in an assuincd standard atmosphere. This arises from the fact that, 

 the curvature of a radio ray, C, is given by ^ 



C = - "^rcos0 (1.27) 



n dh 



where 6 is the local elevation angle of the ray. Since the curvature of a 

 radio ray is proportional to the gradient of the refractive index, specifica- 

 tion of a model for dn/dh specifies the curvature of radio rays in that model 

 atmosphere. For example, it is customary for radio engineers to think in 

 terms of effective earth radius factors [20]. This convenient fiction, 

 discussed in chapter 3, makes straight the actual curved path of a radio 

 ray in the atmosphere by presenting it relative to an imaginary earth 

 larger in radius by a factor k than the radius of the real earth, a, thus 

 maintaining the relative curvature between earth and radio ray. 



This is expressed as 



1 I dn 1 



+ - -,7 cos d = y- -\- , 



a n dh ka 



curvature of curvature of curvature of curvature of 



earth radio ray effective earth straight ray 



^Equation (1.27) as well as general ray curvature considerations are elegantly 

 derived by Millington [21] for the effective earth's radius model. The same equation 

 is also derived in chapter 3. 



