no 



N CLIMATOLOGY 



Figure 4.15. Radiosonde stations used in this study. 



increment was effectively removed and there remains a more reliable 

 estimate of the mean N at the center of the height increment. The final 

 value of A^'o, for example, at 1 km over Washington, D.C., would be 300 

 with a standard deviation of 15.7 N units for 152 observations. The 

 error in the mean is then determined to be 1.3 from 



s(N) = 



s(N) 

 Vk 



(4.10) 



with k = 152. An examination of similar data for all stations and levels 

 used shows that s{N) is generally less than 1.0 A" unit. 



In addition, when considering atmospheric cross sections, the mean 

 values of A were referred to sea-level by means of (4.9) thus further em- 

 phasizing climatic differences [28]. 



