A mode-conversion program has been developed which allows for the vertical inhomogeneity of the ionosphere as well as its anisotropy. As an illustration of the potential use of the program for diagnostic studies at VLF, the sunrise amplitude study of Walker [1965] is numerically modelled with allowance for four modes. Exponential profiles of varying reference height and scale height are used to simulate the terminator. Of the models studied, best agreement is obtained with a terminator thickness of 1000 kin. Despite the multimode nature of the models, Crarnbie's [1964] model is found to be valid for the path studied. This is because of weak conversion from the third and fourth nighttime modes to the first daytime mode. INTRODUCTION Crombie [1964] utilized mode-conversion concepts to explain phase steps and amplitude fading at dawn and dusk. His explanation is based on two modes being present in the nighttime portion of the path and only one mode in the daytime portion of the guide. In addition to fitting most experimental data, Crombie's explanation had consequences which were capable of verification by further experimental evidence. Walker [1965] reported results of an experiment designed especially for testing the consequences of Crombie's explanation. Involved in the experiments were a mobile monitor station on board ship and a permanent ground monitor station both of which recorded phase and amplitude of 18.0-kHz signals from Balboa (NBA). Walker's results fully supported Crombie's explanation. Bicket et al. [1970] in a series of in-flight nighttime VLF measurements along radials from Hawaii have convincingly demonstrated that more than two modes are significant during nighttime at 23.4 kHz. Using a nighttime profile (• = 0.5 km -• h -86 km in the notation of Wait and Spies [1964]) similar to that proposed by Bickel et al . [1970] and propagation conditions closely approximating those of Walker's experiment, a mode sum was generated for 18.6 kHz (this frequency rather than 18.0 kHz is used throughout this paper because of the availability of eigenangles from a previous study), and at least four modes were found to be significant.The waveguide program used to generate the mode constants was that reported by Pappert et al. [1967].Recently a mode-conversion program has been developed which allows fully for the vertical inhomogeneity of the ionosphere as well as its anisotropy. As an example of .the application of the program, the sunrise amplitude study of Walker, specifically the occurrence of simultaneous minima during sunrise, is numerically modelled with allowance for four modes. Several terminator models are used in the study. Despite the multimode feature of the nighttime portion of the path we find, using a daytime profile in which • = 0.3 km -x and h -70 km and a terminator thickness of 1000 km,

Weakly attenuated components of the VLF mode spectrum associated with propagation below highly anisotropic ionospheres are presented as a function of frequency and azimuth. In particular, results for the phase velocity, attenuation, polarization mixing ratio, and excitation factor for vertical dipole excitation are presented in the frequency range of 10–30 kHz for midlatitude paths, whereas azimuthal dependencies are presented for a frequency of 19.8 kHz and dip angles of 0° and 60°. The ionospheres are described by exponential electron density and collision frequency profiles, and the primary region of wave‐plasma interaction falls in the highly anisotropic region of the ionosphere. It is shown that polarization mixing is much more pronounced for a westerly propagation path than for an easterly path at midlatitudes, and, as a result, principally TE modes can be expected to influence the mode sum for frequencies at least as low as 20 kHz for propagation to the west. On the other hand, as has been shown previously, contribution to the mode sum by principally TE modes for propagation to the east becomes significant only near 30 kHz. Azimuthal anomalies include drastic polarization changes in going from easterly to westerly paths. For example, in the case of transverse propagation at the magnetic equator, it is shown that modes that are pure TM for propagation to the east may be pure TE for propagation to the west. Tantamount to this is the statement that modes that have dominant excitation for propagation to the east may have vanishing excitation for propagation to the west. Azimuthal dependencies are shown to be very often characterized by rapid variation of the mode constants in the neighborhood of north‐south or south‐north propagation. These variations manifest themselves in marked differences of the mode sum for azimuthal changes at least as small as 10°. Unfortunately, the azimuthal dependencies of the mode constants do not appear to lend themselves to any simple analytical approximations.

No abstract

Measurements of the field intensity of VLF transmissions from NPM at 19.8 kHz were made at four fixed ground sites located in Bechuanaland, Mrica, during July and August 1963. The sites were in the vicinity of the antipode of the transmitter in Hawaii. Each site employed both a vertical whip and a vertical, rotating loop antenna. The data indicate that a standing‐wave interference pattern is present in the antipodal region most of the time. During local sunrise and sunset, such a pattern appears to be absent and the signal appears to be received primarily from one general direction. Oscillations of 1‐ to 2‐hr periods in the diurnal amplitude data are explained as being due to interference of signals arriving from various directions, with phase changes associated with changes of the terminator position along the paths. The difference between east‐to‐west (E‐W) and west‐to‐east (W‐E) attenuation rates for the first waveguide mode is examined. A value of about 0.6 dBfMm is consistent with the observations.

No abstract