748 



AIRBORNE NAVIGATION AND GROUND SURVEILLANCE 



described in reference 7 of this chapter's bibliography, along with recent 

 measurements of radar scattering at X band over land and sea. For 

 example, in that sea scale the Douglas sea state 1 is subdivided into four 

 separate sea states.) The curves in Fig. 14-1 1 show that for a beamwidth of 



3 4 5 



BEAMWIDTH (deg) 



Fig. 14-11 Doppler Velocity Water Bias Error vs. One-Way Beamwidth for 

 7 Angle of 69° and j8 Angle of 0°. (Courtesy, Laboratory for Electronics) 



5°, which requires an antenna diameter of approximately 18 inches at 

 X band (3 cm) and 13.5 inches at K^ band (2.2 cm), the uncorrected bias 

 error is 2.3 per cent at Douglas sea state 4 and 3.7 per cent at Douglas sea 

 state 1. This is an error spread of 1.4 per cent between the two extremes. 

 For a beamwidth of 3° (corresponding to antenna diameters of 30 inches 

 at X band and 22 inches at K^) the uncorrected bias errors for the two 

 extremes of sea state reduce to 0.9 per cent and 1.4 per cent respectively, 

 or a spread of 0.5 per cent. 



Since this error is a systematic or bias error, it is possible to eliminate it 

 either by careful calculation or (still better) by in-flight calibration for any 

 one particular sea state. Most modern doppler systems are equipped with a 

 land-sea switch^ with one sea position, which should be calibrated for the 

 average or most frequently occurring sea state. The error which remains is 

 the spread between the errors for the sea state extremes. As can be seen 

 from Fig. 14-11 and the examples cited above, this error spread again 

 decreases rapidly with decreasing beamwidth (or increasing antenna 

 aperture). For the two above examples of 5° and 3° beamwidths, the total 

 error spread would be 1 .4 per cent, or an actual peak error of dzO.7 per cent, 

 and 0.5 per cent, or an actual peak error of ±0.25 per cent if the sea switch 

 were assumed to be adjusted for the average of these sea states. These, 



