PRESENTATION OF BASIC DATA 91 



Data from 306 weather stations were obtained in order to give reason- 

 able geographical coverage. In general, 5 years of records were obtained 

 for each station from the period 1949 to 1958, preference being given to 

 the years 1954 through 1958. A noticeable exception, however, was 

 Russia, for which only 1 year of data (IGY) is reported in Climatic Data 

 for the World; thus all charts are drawn with dashed contours for Russia. 

 There are vast expanses of ocean for which there are no meteorological 

 observing stations. Climatic atlases were utilized in order to present 

 estimates of world climate in these locales. A reasonable coverage of 

 the sparse data areas of the world was made by estimating temperature 

 from sea surface isotherms [2] and humidity from charts of seasonal 

 average depression of the wet bulb temperature [3]. Pressure was esti- 

 mated for these locations from average winter and summer pressure 

 charts. 



When these data were converted to A^ [4] and charts prepared, a pro- 

 nounced altitude dependence could be seen, as in figure 4.1. Figure 4.1 

 and the following charts of N variations across the United States are from 

 an extensive N climatology now being prepared at the Central Radio 

 Propragation Laboratory. Although the present study is primarily 

 aimed toward worldwide variations, it is felt that the U.S. data better 

 illustrate the height dependence of Ng and the subsequent reduction 

 process employed. It is noted that the coastal areas display high values of 

 N s, while the inland areas have lower values. There are low values of 

 N s corresponding to the Appalachian and Adirondack Mountains and a 

 decrease with increasing elevation of the Great Plains until the lowest 

 values are observed in the Rocky Mountain region and the high plateau 

 area of Nevada. A corresponding gradient is observed from the west 

 coast eastward. Crosshatching encloses areas where the terrain changes 

 so rapidly that it was felt the data were inadequate to obtain realistic 

 contours of N s. 



The altitude dependence of N can be studied in terms of the "dry" and 

 "wet" components of N. These components are those of the two-term 

 expression in (1.17). The dry term, D, 



i) = ^^ , (4.2) 



is proportional to air density and normally constitutes at least 60 percent 

 of A. 



The average variation of density with altitude in the atmosphere may 

 be expressed in the first approximation as 



p = pocxp {-z/H\ (4.3) 



