COMPARISON WITH EXPERIMENTAL RESULTS 351 



U.S. Weather Bureau personnel at the Puunene Airport weather station. 

 The agreement between observation and prediction is fairly good, espe- 

 cially when one considers that only 32 of the 86 points lie outside of the 

 predicted standard error of estimate limits, while chance would indicate 

 that 29 points would exceed these limits. Also, it should be kept in mind 

 that in this case, as for all except the Collins data, the target beacon is 

 located on the surface of the earth, whereas the predictions from the 

 CRPL Standard Profile Sample are derived for targets in the free atmos- 

 phere; there is undoubtedly some bias introduced in this way. 



As a part of a continuing investigation into the atmospheric limitations 

 imposed on electronic distance measuring equipment, some measure- 

 ments have been made recently by the Tropospheric Physics Section, 

 NBS, of both range errors and range difference errors (across a phase- 

 differencing baseline) over a propagation path near Boulder, Colo. 

 Figures 8.25 and 8.26 are based on some of the preliminary results of 

 these measurements [24]. Figure 8.25 shows the results of measurements 

 of the fluctuations in apparent range, made at half-hour intervals on 

 May 9-11, 1961, over a 15.5-km path between a transmitting beacon 



r 



LINEAR REGRESSION 



r =0.86 



SLOPE : 0,865 ppm/N 



PREDICTED SLOPE, 0879ppm/N 



THE STANDARD ERROR OF ESTIMATE 

 OF EITHER LINEAR FIT IS 5.4ppm, 

 THE PREDICTED VALUE = 5.8ppm. 



E 



Q. 



a. 



Figure 8.25. Range error fluctuations observed over the Boulder Creek-Green Mountain 



path, Colo. 



3-day run, half-hourly readings, range = 15.5 km, target height 088 m, approx elevation angle 26° May 

 9-12, 1961 



