VLF Phase Research Articles

VLF propagation disturbances and electron precipitation at mid-latitudes

Observations of nighttime phase disturbances to long-distance mid-latitude (2 ≲ L ≲ 4) VLF transmissions during a magnetospheric substorm on January 2, 1971, are presented. The onset of the phase disturbances to a number of VLF paths began at the time of the substorm expansion. Whistler measurements during this event indicate that all the paths were below the plasmapause. Bremsstrahlung X rays and VLF emissions were observed above the plasmapause, near L = 4, simultaneously with the VLF propagation disturbances. A previous analysis of the X ray and VLF emission data indicates that electron precipitation at L = 4 was caused by cyclotron resonance between energetic trapped electrons and whistler mode VLF waves near the equatorial plane. The data presented here suggest that the mid-latitude ionospheric disturbances are due to energetic electron precipitation, which is possibly caused by this same mechanism. For this event it is estimated that a flux of electrons J(>40 kev) ≲ 300/cm² sec ster with e-folding energy equal to 45 kev will account for the measured phase change along a path nearly parallel to L = 3. Calculations illustrate that nighttime VLF phase is very sensitive to values of incident electron flux that are only marginally detectable, if at all, with the more conventional indirect indicators of electron precipitation, e.g., X rays and riometer absorption. This suggests that VLF propagation can serve as a useful tool for the detection of low-intensity electron precipitation.

D-region recombination coefficients and the short wavelength X-ray flux during a solar flare

VLF phase and amplitude measurements were made on five different frequencies at São Paulo, Brazil during a solar flare which occurred on 22nd January 1972. The phase and amplitude measurements during the decay phase of the flare were combined with the full wave solutions of Wait and Spies (1964) to calculate the recombination coefficient in the lower ionosphere. The values thus obtained are lower than those reported by Reid (1970), but are compatible with those reported by Montbriand et al. (1972) during Solar X-ray events. The effective loss rates have been utilized to calculate the ion-production at the maximum of the flare, which in turn has been utilized to calculate the incident X-ray flux as a function of wavelength at the maximum of the flare. Extensions to the calculations are discussed.

VLF Phase Changes Produced by Particle Precipitation into the Geomagnetic Anomaly During Solar Proton Events

Solar protons in the events of August 28 and September 2, 1966, produced phase disturbances on the 21.4–kHz VLF transmissions from station NSS, Annapolis, Maryland, to São Paulo, Brazil, and also on the 26.1‐kHz VLF transmissions from station NPM, Hawaii, to São Paulo, Brazil. Both these paths lie well below the polar zone, but partially traverse the geomagnetic anomaly. A quantitative connection between the observed phase advances and satellite observations of the proton flux is used to calculate the length of path affected and the height changes in the disturbed portion of the waveguide. The VLF variations are accounted for by exponential ionospheric models with a reference height equal to the effective reflecting height of the sharply bound model.

Ionospheric effects of solar flares—IV. Electron density profiles deduced from measurements of SCNA's and VLF phase and amplitude

A method is presented by which electron density profiles in the D -region of the ionosphere may be derived during solar flares from: 1. (1) SCNA measurements at two frequencies; 2. (2) simultaneous measurements of SCNA's and SPA's at VLF. A parabolic profile is assumed. At lower levels, the profile is estimated from the measurements of phase and amplitude changes at VLF. Profiles are obtained for a number of flare events, and are compared with those derived recently by cross-modulation, partial reflection and rocket techniques. It is observed that the enhanced electron density at 60 km is of the order of 100 cm −3 during an average flare effect, about 900 cm −3 around 70 km, but increases steeply with height to about 10 4 cm −3 around 85 km.

Effects of X-ray stars on VLF signal phase

Stellar X-ray effects from Sco XR-1, Cen XR-2 and Cen XR-4 have, been found on several of the numerous VLF phase recordings collected on a world-wide basis since 1965. Only marginal VLF phase anomalies could be associated with Tau XR-1, and none with the Sco XR-1 flares of 3 June 1967. Stellar VLF phase effects are more pronounced at higher signal frequencies and on VLF paths with long nighttimes. Scattered L α radiation after sunset and before sunrise reduces the sidereal shift of VLF effects on E ↔ W paths; and the closeness of the signal path position to the solar terminator at local midnight can shift the date when a maximum stellar VLF phase anomaly is observed. A detectability factor is defined which compares expected stellar effects on arbitrary signal paths to the established effect of Sco XR-1 on the 22.3 kHz signal from NWC to Tananarive. VLF data show: In 1968, Sco XR-1 caused a maximum phase advance of ~0.6 μsec/Mm on NWC (22.3 kHz)—Tananarive; Cen XR-2 and Cen XR-4 increased their X-ray flux to about that of Sco XR-1 on or about 7 March 1967 and 6 July 1969, respectively; Cen XR-2 flux peaked ~27 March and possibly ~10 April and decreased rapidly between 20 April and 10 May 1967. It is concluded that 1. (a) presently known stellar X-ray sources are not detrimental to the accuracy of the OMEGA navigational system; 2. (b) the permanent X-ray stars may give rise to a seasonal variation of nighttime phase at frequencies above 15 kHz; 3. (c) continuous VLF phase tracking at frequencies above 20 kHz can provide some valuable astrophysical information complementing space probe data.

Observed and Predicted VLF Phase Behavior for the Solar Eclipses of September 11, 1969, and March 7, 1970

A VLF propagation and a D‐region aeronomy model have been used to predict the effect of two solar eclipses on VLF propagation to Aztec, Arizona. Transmissions at 12.2 kHz from Hawaii were monitored during the eclipse on September 11, 1969; and transmissions at 12.0 kHz from Trinidad, at 24 kHz from transmitter NBA, at 12.5 kHz from Forestport, New York, and at 17.8 kHz from transmitter NAA were monitored during the eclipse on March 7, 1970. The VLF phase predictions were found to agree with measurements except for the two nearly coincident northerly paths from Forestport and NAA.

Short path VLF phase and amplitude measurements during a stratospheric warming in February 1969

This note presents some VLF data from a short, high-latitude path recorded during a stratospheric warming event in February 1969. Based on ionosonde data and VLF results it is concluded that no large changes in D-region electron density occurred over North Norway during the event. The VLF observations, showing daytime phase retardation and constant amplitude during the stratwarm, can be explained by assuming a height increase of 3 km for the atmosphere at VLF reflection levels (about 60 km). Height data for constant-pressure levels in the stratosphere make such an assumption reasonable.

VLF Phase Disturbances, HF Absorption, and Solar Protons in the PCA Events of 1967

The ionization model developed in conjunction with VLF and HF propagation parameters from our analysis of the PCA's of August 28 and September 2, 1966, is applied in unmodified form to the 1967 events of January 28, February 2, May 23, and May 28. As for the 1966 events, proton intensities, energies, and cutoffs measured on the polar orbiting satellite 1963–38C are successfully connected to VLF phase disturbances and multifrequency riometer absorption increases. Considering all these events, the model provides the following rms fits between observed and computed values: 4.0 μ sec for 141 daytime and nighttime VLF phase values for seven paths at five frequencies during six proton events; 0.45 db for 181 riometer absorption values at 30 MHz during the six events; and 0.20 db for 36 riometer absorption values at 50 MHz during the two events that produced this absorption. The variability of the proton intensities and spectrums provides a rigorous test of the validity and internal consistency of the ionospheric model deduced through the complementary use of riometer absorption and VLF propagation changes occurring during PCA's. In the present analysis, as in the preceding one, the temporal variations in the proton cutoffs play an important role in the VLF analysis as a result of the lengths of the VLF transmission paths in relation to the polar‐cap area disturbed by the solar protons.

The correlation between sudden phase anomalies and solar microwave radio bursts

More than two hundred SIDs were observed as VLF phase anomalies (SPA) at times when measurements were being made of solar radio noise at 7 GHz. 99 per cent of the SPAs were correlated with microwave solar events. The SPAs are usually smooth and do not show structure corresponding to that in many of the associated radio bursts. A linear correlation coefficient of 0.6 was found between the mean flux of solar bursts and the ‘importance’ of the SPAs as measured in terms of the phase change divided by the length of the transmission path. The SPAs show only very small dependence on the solar zenithal angle.

VLF phase disturbances, HF absorption, and solar protons in the events of August 28 and September 2, 1966

The VLF phase disturbances and HF absorption observed during the proton events of August 28 and September 2, 1966, were quantitatively connected to the proton intensities, energies, and cutoff latitudes observed by satellite 1963-38C, at 1100 km in a polar orbit. The connection is achieved through considerations that include a model of the ionospheric D region described in terms of an effective recombination coefficient related to the four generalized ionic reactions. The complementary use of the VLF and HF observations tends to reduce ambiguities often present in the formation of D-region models. The VLF is the more sensitive indicator of ionization effects due to solar protons in comparison to the HF absorption (especially at night) because the VLF responds to ionization enhancements over a relatively narrow altitude range varying with the proton energies and intensities.

Effects of polar cap absorption events on VLF transmissions

Following a brief review of Hakura's model of a polar cap absorption event (PCA), VLF instrumentation, and data reduction problems, eleven PCAs are studied for their effects on phase and amplitude of VLF transmissions. The eleven PCAs occurred between 24 March 1966 and 6 June 1967. VLF phase anomalies are always advances (Δφ), the magnitude of which increases with increasing geomagnetic latitude of the path, decreasing signal frequency, and decreasing electrical ground conductivity. Signal strength falls by more than 15 dB/1000 km for paths with very low electrical ground conductivity (Greenland and Antarctic Ice Cap), whereas it may increase by a few dB for polar paths over sea water. It is shown that normalization of (Δφ), with respect to frequency and affected length, leads to two distinct linear relationshipsbetween (Δφ) and (log log F/ F 0, where F is the proton flux as measured by satellites. The line of lower slope applies to paths over a highly conducting boundry (sea water), and the line of higher slope to a poorly conducting boundary (ice). The minimum detectable proton flux is 0·5 protons/cm 2 sec ster. The (Δφ)-(log log F/ F 0) relationship implies a decreased sensitivity of (Δφ) to changes in F for larger values of F, an almost uniform VLF reflection height over the polar cap, little influence on reflection height of the shape of the flux spectrum during peak flux, and an upper limit of approximately 25 μsec for a PCA phase anomaly on NPM (26·1 kHz)-Kiruna. Occasionally, irreversible cycle advances were observed on NPG-STO and WWVL-STO. The former are interpreted in terms of enhanced mode conversion and the latter in terms of the tracking of an interfering VLF signal. An event on 13 December 1966 is believed to have been due to electron precipitation from the outer radiation belt.

The position and height of auroral absorption deduced from VLF phase measurements

The main findings from VLF measurements during auroral absorption are as follows: The geographical extent of auroral absorption seems to be at least a factor of two larger than found by HF riometers. The VLF reflection layer is markedly depressed during auroral disturbances. At night large and rapid fluctuations in the reflection layer may occur simultaneously over distances of more than 500–1000 km. During day-time the depressed VLF reflection layer seems to be relatively homogeneous over the same distances.

VLF Phase Anomaly induced by a Nuclear Explosion

THE artificial disturbance of the upper atmosphere by nuclear explosions has been extensively studied. Investigations subsequent to the Argus Project1 have shown conclusively that the effects produced by such explosions are similar to those caused by solar flares2. The observations described here concern the single nuclear explosion which occurred at Lop Nor, Sinkiang Province, China (lat. 45° 15′ N., long. 82° 15′ E.) on June 17, 1967, at 00 : 19 : 08 GMT. It is conjectured that this was a megaton-range, high-altitude nuclear explosion, but little reliable unclassified information is available.

VLF phase perturbations produced by solar protons in the event of February 5, 1965

Solar protons in the event of February 5, 1965, caused day and nighttime phase disturbances in the 18.6-khz VLF transmission between station NLK in Jim Creek, Washington and Thule, Greenland. The phase anomalies persisted for three days with a maximum deviation of 20 μsec, representing 67% of the total normal diurnal variation. A quantitative connection is made between the observed VLF phase variations and the solar protons measured by satellites in the near-earth environment by means of specific ionosphere and VLF propagation models. The lower ionosphere is described by effective recombination coefficients, αeff's, appropriate to day and nighttime ionization conditions. The effective VLF reflecting height of a sharply bounded ionosphere is located at the altitude where the ionospheric conductivity parameter ωr equals 2.5 × 105 sec−1. The VLF variations are also explicable by an exponential ionosphere with a reference height equal to the effective reflecting height of the sharply bounded model. An integral flux of 150 protons/cm² sec with E ≥ 20 Mev produced a daytime lowering of the VLF reflecting height by 10 km. The ionization model of the lower ionosphere derived in relation to the VLF is used to compute daytime riometer absorption that fits well to the absorption measured at McMurdo Sound, Antarctica, at 30 Mhz, but not as well at 50 Mhz.

Heat conduction waves in the upper atmosphere

In the thermosphere a temperature and pressure disturbance of the neutral air can be transported vertically by a wave mode that mainly depends on the heat conductivity of the air. In the low-frequency region (ω < 10−2 sec−1 or periods of τ > 10 min) these heat conduction waves propagate with phase and group velocity proportional to the reciprocal density of the air. The mean velocity is of the order 10–100 m/sec within the ionospheric F region. The traveling time of a disturbance from the exosphere down into the lower F region is about 10 hours and about 2 days down into the E region. These waves may be excited by hydromagnetic waves in heights above 1000 km. Their large propagation time can explain the time delay of ionospheric storms and of the geomagnetic activity effect of the neutral air, as well as the after-storm effect of VLF phase heights in the nighttime E layer related to the time of impact of an enhanced solar wind as the beginning of a geomagnetic storm. Moreover, heat conduction waves may be the cause of ionospheric traveling disturbances excited within the exosphere and propagating downward with a phase velocity equally directed to their group velocity.

Phase Measurements of VLF Transmissions Over an 11,000‐km Transequatorial Path

Phase measurements of the 18.6‐kHz transmissions from Jim Creek were recorded during 1964 and 1965 at São Jose dos Campos (SP). Brazil. The diurnal signal phase variation presents an average time shift delay of 80 µs, which is rather regular throughout the year. The behavior of the phase rate of change during sunrise and sunset is discussed in terms of the interference pattern of propagation modes and in terms of the angle between the sunrise (and sunset) line and the transmitter‐receiver great circle path. Experimental evidence permits one to conclude that the phase‐path reflection height is a strong controlling factor of the second‐order mode propagation. Observations also show that the angle mentioned above is an important parameter in VLF phase studies.

VLF phase perturbations produced by the Soviet high-altitude nuclear explosion of November 1, 1962

The Soviet high-altitude nuclear burst of 0912 UT on November 1, 1962, produced delayed perturbations of the VLF transmissions monitored at APL/JHU over five paths remote from the burst region. The burst-related VLF effects varied from 6 to 55% and averaged 32% of the normal diurnal variation. It is suggested that the VLF phase perturbations are produced by particles from the burst that drift to the remote regions and produce an ionization enhancement in the lower ionosphere in their first, and sometimes in their second, global orbit. The temporal histories of the VLF phase perturbations are compatible with a model wherein it is assumed that the trapped particles are electrons from the radio-active decay of neutrons, the artificial belt extending from an outer shell of L ≈ 3.4 to at least L = 1.9. Satellite particle data do not show an artificial belt of electrons from the decay of neutrons, a belt which in this case may have been obscured by the background flux of electrons from the natural Van Allen belt and from the artificial belts from the earlier Soviet explosions. Satellite data do, however, show a narrow belt of fission electrons ≳1 Mev with maximum flux at L ≈ 1.77 and a lower boundary at L = 1.73. These fission electrons did not affect the VLF paths intercepted by these shells. The VLF and satellite data agree that fission electrons ≳1 Mev were not present on shells above L ≈ 1.9.

Cosmic noise absorption and VLF phase measurements on the eclipse of May 30, 1965

Measurements of the response of the D-region absorption of cosmic noise to the passage of totality during the eclipse of May 30, 1965, were accomplished at several frequencies. A site in the Cook Islands at Aitutaki, which was near the D-region totality path, was instrumented with riometers operating at 20, 30, and 60 Mc/s on wide beam antennas, and at 30 Mc/s with a narrow beam array directed toward the center of D-region totality. The phase of a VLF signal transmitted from Jim Creek, Washington, on 18 kc/s was also monitored. The absorption decreases observed during the eclipse on 20 and 30 Mc/s, and the phase changes, caused by the effective VLF signal reflection height changes are reported here.

Very low frequency phase perturbations and the Soviet high-altitude nuclear bursts of October 22 and 28, 1962: 2. Comparison with satellite particle data

Energy and spatial characteristics of artificial radiation belts were deduced in part 1, where we examined the temporal histories of the VLF phase perturbations produced by the two Soviet high-altitude nuclear bursts of October 1962. For both bursts the trapped electrons have energies ≲4 Mev, the electron spectrum ≲0.78 Mev resembling that from the radioactive decay of neutrons. This dichotomy of fission and neutron-decay β particles in the electron spectrum has not been achieved in the satellite particle data so far analyzed, where the energy ranges are either too broad or above 0.78 Mev. For the burst of October 22, the VLF characteristic that electrons ≲4 Mev are trapped in the region encompassing at least 1.75 ≲ L ≲ 3.76 is compatible with the trapped particle flux observed by Explorer 14, Alouette, and Telstar 1. For the burst of October 28, the VLF shows that fission β particles ≳0.78 Mev are not present in L > 2.32 and are confined on shells 1.6 Mev or the Explorer 15 results for electrons >1.9 Mev. The Alouette results also show that the energy spectrum extends to about 10 Mev, but that there is a rapid decrease in flux for E > 4 Mev. For the burst of October 28, the VLF shows that neutron-decay β particles have an outer trapping boundary at 2.32 ≤ L 230 kev but not with the Telstar data for E > 390 kev, where the outer boundary extends to L ∼ 4. The VLF result that protons 2.3 500 kev at L = 2.8 and 3.2. The characteristics of protons from nuclear explosions are reviewed, and the electron results are consolidated to form at least a model for additional investigations of the artificial radiation belts formed by the Soviet bursts of October 22 and 28, 1962.

Very low frequency phase perturbations and the soviet high-altitude nuclear bursts of October 22 and 28, 1962: 1. Observations and inferred radiation belts

On October 22 and 28, 1962, nuclear explosions over central Asia produced ionospheric D-layer perturbations as monitored on various VLF paths in the northern hemisphere. Delayed VLF effects due to trapped particles averaged 53% of the normal diurnal change for the burst of October 22, and 29% for the burst of October 28. The types of particles and their energies were the same for both detonations, but differences existed in the spatial extent of the radiation belts producing the VLF phase perturbations. The trapped particles were electrons with E ≲ 4 Mev and protons in the range 2.3 to 4.6 Mev. The electron spectrum is composed of fission and neutron-decay β particles, the latter predominating for E < 0.78 Mev. Protons <2.3 Mev may have also been present, but those with E < 1 Mev cannot penetrate into the D region and are of no direct VLF interest. With the dipole field approximation, for the detonation of October 22, fission and neutron-decay β particles were trapped in a region bounded by field lines intersecting the earth's surface in a range extending at least from 36.5° to 56° geomagnetic (for the real field the shells are 1.75 ≤ L ≤ 3.76). For October 28, fission β particles were confined to areas <49°N (L < 2.32) and probably to field lines around that passing through the burst point (at 36.5°N and L = 1.75); this type of confinement also existed for 2.3- to 4.6-Mev protons after both nuclear bursts. For October 28, effects due to neutron-decay β particles were found at 49°–51°N (2.32 ≤ L ≤ 2.79), a region which formed the outer boundary for these trapped particles. For the burst of October 22, a prompt effect was observed on the GBR to APL transmission.