`Wild¿ bearings in high-frequency direction finding

In high-frequency direction finding there occur, not infrequently, bearing errors which substantially exceed those normally experienced. It is customary to refer to bearings that are found to be very wide of the mark as `wild¿. Since one exceptionally large error in a small group of bearings apparently has a disproportionate effect on the overall result, attempts have been made from time to time to justify the rejection of such `wild¿ bearings on the basis of the size of the error. The justification for such a course is examined in the paper, and an alternative more systematic approach is proposed. Misidentification, i.e. observing the wrong signal, has often been a cause of `wild¿ bearings in the past. Every effort must be made to eliminate this cause. In this connection, it should be noted that propagation conditions rarely produce large errors with signals that are strong, not subject to severe fading and exhibiting sharp steady bearings. A large error associated with such conditions almost certainly points to misidentification. Having excluded misidentification, other cases of wild bearings should be examined against the propagation conditions obtaining. If no true ray path was likely, the signals are probably due to scattering and bearings should be rejected. Light may be shed on the propagation conditions by the signal waveform and the nature of the fading. The rejection of bearings solely on the basis of magnitude of the error is deprecated.