General VLF phase variations observed at College, Alaska

One of the original objectives of the VLF program at College was to determine how geomagnetic and auroral disturbances affected long-distance high-latitude VLF propagation. Phase and amplitude recordings of VLF signals at College, obtained primarily during 1963 from several of the stabilized VLF transmitters, were used in this study. Using both case-by-case and statistical correlation studies, we found no strong connection between geomagnetic or auroral disturbances and the large, frequent, short-term VLF phase variations observed at College. A consideration of the physical processes involved in the production of geomagnetic disturbances shows that, in general, little or no direct effect on long-distance VLF signals would be expected. Auroral precipitations producing negative-bay disturbances are generally relatively limited in extent. Thus, an aurora could affect only those signals which pass near the precipitation region. An examination of our VLF records indicates that the second mode is not negligible at great distances during the night. The nighttime second mode amplitude may almost equal that of the first on 4000- to 5000-km paths. A strong second mode at the sunrise line seems necessary to produce the sunrise effect. For example, by using the sunrise fading pattern on VLF signals, we show that the second guided-wave mode propagated strongly to 7500 km at night on NBA-Europe paths during June 1962. This result is used to show that mode interference is important when considering VLF signal variations. A model involving the production of a disturbance wave by the conversion of energy from the primary field in relatively limited areas of the D region is proposed to explain the commonly observed 30 to 90 degree nighttime phase anomalies recorded at College. Suggested mechanisms for the production of the disturbance wave are mode conversion or diffraction from steep horizontal gradients in the D region. The propagation of the second mode to great distances poses a serious difficulty to the use of VLF data for D-region research. Second mode interference effects in the 4000- to 5000-km range can lead to errors in computed effective reflection height changes of 30%; interpretations based on such computations would be quite misleading if the erroneous values were interpreted as actual heights or height changes in the D region.