14-8] DOPPLER NAVIGATION SYSTEM PERFORMANCE DATA 757 



smoothing afforded by these longer flights decreases the effect of certain 

 slowly varying errors, such as compass and variation errors, and the errors 

 in fix measurement and lack of knowledge of the flight end points. An 

 example of this is the data shown in Table 14-1, taken from the test flights 

 of the AN/APN-66(XA-2) discussed above, as broken down into three 

 different ranges of flight length. It is interesting to note that the mean error 

 for the longer flights of 1422 miles average length is approximately half as 

 large (0.45 per cent) as the mean error for the shorter flights of 145 miles 

 average length (0.86 per cent). 



Table 14-1 ACCURACY CHARACTERISTICS OF AN/APN-66 

 FOR 59 OVERLAND RUNS 



Range of Distance, Miles Average Length, Miles Mean Position Error, % 

 0-200 145 0.86 



201-650 486 0.64 



651-2500 1422 0.45 



Another series of flights, consisting of 766 flight legs with the same 

 equipment, provided position error data which were broken down into 

 range and transverse errors, as defined in Equations 14-8 through 14-11 of 

 Sec. 14-3. On the basis of these data the constants which relate these errors 

 to the total distance traveled and which were defined as Kr and Kt in 

 Equation 14-11 were determined (for that equipment) to be 8 and 16, 

 respectively, for D expressed in nautical miles. 



One further demonstration of the significant effect of the heading 

 reference and magnetic variation information on system performance is 

 shown by the results obtained by Wright Air Development Center, Dayton, 

 Ohio, with Radar Navigation Set AN/APN-82, which is composed of the 

 same doppler radar as the AN /APN-66 (the AN /APN-81) and Navigational 

 Computer AN/ASN-6 (with only a manual variation capability) and which 

 was tested with a standard N-1 magnetic compass. As discussed earlier in 

 this paragraph, the mean position error over land of the AN/APN-66 (with 

 automatic variation and an improved N-1 heading reference) was 0.51 per 

 cent of distance traveled, whereas it was 1.5 per cent of distance traveled 

 for the AN/APN-82. The difference can be attributed largely to the 

 variation and heading errors, and, perhaps to a lesser degree, to the differ- 

 ence in the type of computer used on those two systems. 



Modern doppler radars are capable of accuracy performance of the order 

 of less than 0.1 to 0.2 per cent probable and approximately 0.5 per cent 

 maximum error over land. Over-water performance should be improved by 

 the use of land-sea switches and new automatic techniques, and the remam- 

 ing water bias error should increase the maximum velocity error to some- 

 where less than 1 per cent, depending largely on antenna size. Computers 



