AKR Triggered by Type III Bursts

Published in Geophysical Research Letters, vol. 8, number 10, pages 1091-1094, October 1981.

THE STIMULATION OF AURORAL KILOMETRIC RADIATION BY TYPE III SOLAR RADIO BURSTS

W. Calvert
Department of Physics and Astronomy
The University of Iowa
Iowa City, Iowa 52242

Abstract. It has been found that the onset of auroral kilometric radiation (AKR) frequently coincides with the arrival of type III solar radio bursts. Although the AKR onsets are usually abrupt and appear to be spontaneous, they sometimes develop from a discrete frequency near the leading edge of a type III burst or sometimes occur at progressively lower frequencies following that edge. From this, and the absence of the related solar electrons in specific cases, it was concluded that the incoming type III waves were sometimes responsible for stimulating auroral kilometric radiation. It was estimated that intense, isolated type III bursts were capable of stimulating AKR roughly one third of the time, and that at least ten percent of the observed AKR onsets could be attributed to these and weaker bursts, including some barely detectable by the ISEE plasma wave receivers.

INTRODUCTION

Auroral kilometric radiation, abbreviated "AKR", consists of intense radio waves, at frequencies of 50 to 700 kHz, emitted in the earth's magnetosphere above the auroral zone [Dunckel et al., 1970; Gurnett, 1974]. Typically ten megawatts when it occurs, the total power may occasionally reach a billion watts. This paper reports the discovery that such intense emissions can be stimulated by much weaker external waves, namely those of the type III solar radio bursts.

AKR consists of extraordinary-mode waves emitted near the electron cyclotron frequency, approximately perpendicular to the earth's magnetic field [Gurnett and Green, 1978; Kaiser et al., 1978; Calvert, 1981a]. It originates in a plasma cavity which extends from 1.3 to at least 3.3 earth radii (geocentric) at 70^o invariant magnetic latitude, in which the plasma density may be less than one particle-pair per cubic centimeter [Calvert, 1981b]. The occurrence of AKR is associated with the same energetic (1-5 keV) inverted-V electrons also believed to cause the visual auroral arcs [Gurnett, 1974; Benson and Calvert, 1979; Green et al., 1979; Benson et al., 1980].

AKR is believed to be generated by the doppler-shifted cyclotron-resonance instability [Melrose, 1976; Wu and Lee, 1979]. According to this theory, the source region acts like an enormous wave amplifier which draws its energy from energetic electrons. It has been proposed that the wave growth rates for this process would be sufficient to amplify thermal noise to the intense levels which are observed.

If the AKR source region consists of a wave amplifier, it should also be expected to amplify other, nonthermal waves which enter it with the appropriate wave direction, frequency, and polarization. Specifically, it might be capable of amplifying the waves of type III solar radio bursts, provided the conditions are correct when they encounter the auroral zone. In order to test this concept, the type III bursts observed with ISEE-1 were examined for intensifications which could be attributed to amplification. The surprising outcome of this search was not only that such intensifications were found, but that these intensifications would frequently develop into intense AKR which lasted long after the type III burst bad completed.

OBSERVATIONS

The ISEE-1 satellite includes a wave receiver [Gurnett et al., 1978] capable of detecting both auroral kilometric radiation and type III radio bursts. The instrument scans the frequencies from 100 Hz to 400 kHz with roughly six percent frequency resolution every 32 seconds. The received signals are displayed in frequency-versus-time spectrograms, on which the stronger signals appear as darker areas. These ISEE-1 spectrograms, recorded from an equatorial earth orbit out to a geocentric distance of 22 earth radii, were supplemented by the plasma-wave observations from ISEE-3 [Scarf et al., 1978], situated at the liberation point 0.1 AU in front of the earth. The ISEE-3 data are displayed as signal strength versus time for discrete channels up to 100 kHz (see Figure 2).

Type III radio bursts on the ISEE-1 spectrograms appear as smooth bands sweeping downward in frequency and lasting up to a few hours. On the ISEE-1 spectrograms, they start at the highest frequency (400 kHz) and frequently extend down to the solar wind plasma frequency (~50 kHz). Auroral kilometric radiation, on the other hand, exhibits a quite different signature. It is bursty on a time scale of minutes and tends to appear and disappear simultaneously over a range of frequencies. The AKR bursts may persist for a few hours at a time, and there is a tendency for the bandwidth of AKR to narrow at both ends of a sequence of bursts [Kaiser and Alexander, 1977]. Since their signatures are so different, it is usually easy to distinguish between AKR and type III bursts on the same spectrogram. Aided by the ISEE-3 plasma-wave observations of type III bursts, a study of roughly two hundred, twelve-hour ISEE-1 spectrograms recorded during 1979 yielded a few dozen examples of coincident onset, including the two in Figure 1.

Figure 1. An ISEE-1 spectrogram showing two intense type III solar radio bursts which appeared to stimulate auroral kilometric radiation.

Three strong type III bursts occur in Figure 1, along with a few weaker ones. The earliest burst, which occurred just after ISEE-1 left the plasmasphere, was accompanied by a narrow band of enhanced signals at 200 kHz, beginning at its leading edge. A few minutes later both this band and the type III burst were intensified together, but this was probably an incidental result of the satellite's orbital motion. Over the next hour the AKR band expanded and wandered toward lower frequencies until, at 1730 UT, it became an intense, half-hour display of signals with the familiar characteristics of AKR. Afterward these signals contracted to a more narrow band at 130 kHz. It is unclear whether or not the intensification near 130 kHz which lasted until 2000 UT was related to the following type III burst at 1910 UT. Although the weaker type III burst at 2050 UT was free of detectable AKR signals, the strong one at 2150 UT was followed by a volley of AKR which lasted until midnight, aptly announcing the fourth of July. In this case the AKR appeared to begin near 100 kHz, again coincident with the leading edge of the type III burst.

The AKR intensification in Figure I between 1730 and 1810 UT was also detectable at ISEE-3 by its more sensitive radio astronomy wave receiver [Knoll et al., 1978]. On the radio-astronomy spectrogram, it appeared as spur near 200 kHz, extending after the type III burst and clearly associated with it. Additional spurs were found to accompany some of the other type III bursts detected with this instrument during January and July, 1979, as were superimposed temporal striations which also suggested AKR signals. The occurrence of such spurs and striations at ISEE-3 seemed to favor the solstices, presumably because one of the geomagnetic poles was then tipped more toward the satellite.

Two more ISEE-1 cases where the AKR onsets were coincident with type III bursts are presented in Figure 2, along with certain of the ISEE-3 plasma-wave signals received at the same time. The first was a diffuse enhancement, beginning at 0000 UT and 140 kHz, which slowly expanded and probably extended over the data gap until around 0400 UT. Although it failed to develop the impulsive, broadband structure which is typical of AKR, it seemed sufficiently similar to the tapered onset at 1620 UT in Figure 1 to tentatively assume it was the same. The type III burst in this case was quite weak, and hardly discernable in the ISEE-3 data. However, with the better sensitivity and visual integration of ISEE-1, it is visible in the spectrogram beginning at 2320 UT and extending downward to meet the diffuse AKR onset. The second case involved a much more intense type III burst beginning at 0530 UT, clearly evident in the ISEE-3 data and also visible as a smooth background signal extending down to 35 kHz in the ISEE-1 spectrogram. This time the corresponding enhancement was unmistakably AKR which began abruptly over the band of frequencies between 120 and 250 kHz and subsequently followed the leading edge of the type III burst down to 65 kHz. Although it is difficult to distinguish from the ongoing AKR, a second progressive onset also occurred at 0730-0800 UT near 100 kHz, which corresponded to the later strong type III bursts detectable at ISEE-3. Although such instances of progressive onset were relatively rare, they served as a convincing clue that the two wave signals were not always independent.

Figure 2. A strong type III burst (at 0530 UT) where the coincident onset of intense AKR occurred at progressively lower frequencies along its leading edge, plus a much weaker one (at 0000 UT) accompanied by diffuse AKR signals.

To be certain, such striking examples were rather uncommon. Often the type III bursts and the AKR either occurred alone, or else both occurred so frequently any correlation would have been obscured. Some of the typical intermediate cases which suggested a possible correlation are illustrated in Figure 3. In the first panel, a brief 80 kHz enhancement occurred at 1515 UT which appeared to straddle the type III leading edge, but it terminated before the type III burst had completely passed the earth. Then at 1900 UT a weaker type III burst was followed by a lasting sequence of strong, irregular AKR signals. The multiple type III bursts before 0400 UT on 15 August occurred without AKR, and it is uncertain whether the feature beginning at 0415 UT was a type III burst, AKR, or both. Finally, the signals intensified at 0510 UT, coincident with a weak type III signal, and they acquired more the character of AKR. These signals appeared to diminish just before the following strong type III burst at 0745 UT, and it is uncertain whether they reappeared thereafter. In the third panel of Figure 3 (which was free of data gaps), the first two AKR onsets, at 2220 and 0110 UT, occurred abruptly without detectable type III bursts. However, the later onset at 0345 UT occurred at progressively lower frequencies which again seemed to roughly follow the leading edge of the associated type III burst.

Figure 3. ISEE-1 spectrograms which illustrate the variability of AKR stimulation, including type III bursts which failed (before 0400 UT in the second panel) and apparently spontaneous AKR (at 2220 and 0110 UT in the third panel).

ANALYSIS

Although the evidence for AKR stimulation is largely subjective, it became quite convincing once enough cases had been examined. On the other hand, since AKR and type III bursts both occurred frequently during 1979, an objective test was devised to demonstrate that their alignment was not accidental. For this purpose, the ISEE-3 plasma-wave data were independently examined for intense, isolated type III bursts at 100 kHz (like that at 0530 UT in Figure 2), with peak signals exceeding 10^-18 W/m2Hz, preceded by no signals stronger than the instrumental threshold (~3 x 10^-20 W/m2Hz) for at least two hours. With roughly 90% coverage for the year, this yielded 104 occurrences. Excluding a few perigee cases, 27 of the corresponding ISEE-1 spectrograms were available, and this data set was then examined for AKR onsets.

The criterion for an onset was the occurrence of moderately strong AKR signals (greater than roughly 10-17 W/m2Hz) preceded by their absence for at least 30 minutes, excluding isolated AKR bursts lasting less than 30 minutes. In ten of the 27 cases, no AKR was found, and ongoing AKR occurred in nine. AKR onsets satisfying the criterion were found in the remaining eight, all within 30 minutes after the type III bursts had begun.

In order to assess the significance of this result, it was necessary to estimate the frequency of AKR onsets. Using the same onset criterion, 400 hours of the available ISEE-1 data between January and April 1979 were examined separately, and eighty onsets were found. Thus the average onset rate was 0.2 per hour, and the probability of an AKR onset during a random 30-minute interval was ten percent. If the AKR onsets were independent Bernoulli events (see Feller [1950], p. 137), the probability of finding eight in twenty-seven such intervals is 0.003 (one in three hundred). It is therefore quite unlikely that the observed alignment between the type III bursts and the onsets was coincidental.

It seems the most plausible conclusion is that the type III events were responsible for stimulating or triggering AKR. The reverse hypothesis (seriously proposed to the author in a related context) was discarded as improbable. It is well known that a type III radio burst is generated by flare-ejected solar electrons. Relative to the arrival of the pertinent radio frequencies, the more energetic flare photons arrive at the earth up to thirty minutes earlier, and the ions a day or so later. Only the type III waves, and sometimes also the attendant flare electrons, arrive at the appropriate times to account for the stimulation of AKR, and it is believed that the former is responsible.

This was first suggested by cases where the related solar flares occurred eastward on the solar disc, since the energetic electrons from such flares would have been less likely to reach the earth. It was confirmed by the observations of the ISEE-3 energetic electron detector, which revealed no flare electron events (2 keV to 1 MeV) during any of the AKR onsets in the figures presented here [R. P. Lin, private communication]. Although the electron plasma oscillations at 17.8 kHz in Figure 2 suggest the arrival of electrons in spite of this conclusion (see Gurnett et al. [1980]), they occurred too late to account for the AKR onset by more than an hour. The particular case on 15 August in Figure 3 was also testable with the ISEE-1 Lepedea, since the satellite was then in the magnetosheath (between 0200 and 0700), where the arrival of solar electrons should have been detectable. However, no significant fluxes of energetic electrons were observed between 0.3 and 45 keV, nor were enhancements of more energetic particles detected by the ISEE-1 geiger counter [T. E. Eastman, private communication].

The eight coincident onsets in 27 also implied that intense, isolated type III bursts were capable of stimulating AKR roughly one third of the time, Since the average rate for such bursts was around 0.013 per hour, they would account for 0.004 AKR onsets per hour, or two percent of the total. The subjective study, on the other hand, included all detectable type III bursts, down to a signal level roughly one hundred times weaker. In that study it was estimated that more than one tenth of the AKR onsets could be explained by type III stimulation and no threshold was found below which stimulation obviously failed to occur (see, for example, Figure 2 at 0000 UT). In view of the increased occurrence frequency of the weaker type III bursts (estimated to often exceed one per hour) a sizeable fraction of the observed AKR onsets could be attributed to stimulation even if the efficiency for the weaker bursts were substantially less than one third. Furthermore, with the lack of a threshold, the possibility that all AKR is externally stimulated cannot be ruled out.

DISCUSSION

The following model is envisaged for AKR stimulation: The type III wave is generated moderately close to the sun (e.g., at 0.2 AU). It travels to the earth where it is refracted inward by the reduced magnetospheric plasma density, passes over the pole, and enters the auroral zone roughly perpendicular to the magnetic field. Alternatively, the wave could be reflected poleward by the dayside plasmasphere to reach the auroral zone, but a nighttime source is suggested by most of the current observations. Initially unpolarized, the wave undergoes mode splitting in the geomagnetic field and that produces the extraordinary wave probably required for AKR stimulation. The stimulation occurs where the wave finds the simultaneous presence of electron free energy and a low plasma density, since both are required for substantial wave growth.

The stimulation of AKR by an external wave strongly suggests that some sort of signal feedback is involved in the AKR source mechanism. An analog, of course, is the auditorium amplifier which can be set into oscillation by a boisterous speaker. The AKR theories thus far have dealt only with amplification, and feedback not considered. For a complete AKR theory it will be necessary to identify the feedback paths, since they are likely to dominate the emission behavior. Moreover, nonlinear effects or other gain-modifying behavior will also be important, since they will control both the triggering process and the ultimate emission level.

The feedback concept could also explain the complex discrete structure which has been observed in the AKR spectrum [Gurnett et al., 19791. A feedback oscillator functions only at certain frequencies where the loop phase shift is an integral number of cycles and the loop gain is sufficient. It will usually oscillate at only one such frequency, the others being quenched by nonlinearities. However, in the case of AKR, different feedback paths, at different locations in the source region, could produce the different discrete frequencies which are observed.

It might also be feasible to attempt the artificial stimulation of AKR with a satellite transmitter, since the signal level of a type III burst (10^-16 W/m²Hz) could be reproduced (over a 1 kHz bandwidth) with only fifty watts at one earth radius. With such an instrument one might be able to examine AKR stimulation under controlled conditions, and perhaps even trigger a substantial AKR burst and monitor the consequences. It might even be possible to use the AKR source region as a gigantic amplifier with which to generate sufficiently intense pulses to produce detectable echoes from the distant magnetotail, and thus to resolve some quite important questions about magnetospheric structure.

CONCLUSIONS

There is evidence that type III solar radio bursts can stimulate auroral kilometric radiation, and that the waves rather than the attendant solar electrons are responsible. The stirnulated signals sometimes begin at a discrete frequency, between 100 and 200 kHz, coincident with the leading edge of the type III burst. The signals tend to expand and vary in frequency, and later develop into intense bursts of AKR. At other times the stimulated AKR signals begin abruptly over a range of frequencies near the type III leading edge, and cases have been found where they begin at progressively lower frequencies following the leading edge.

The stimulation of AKR by an external wave signal suggests that the source mechanism must involve feedback, and be analogous to an electrical oscillator rather than an open-loop amplifier. This model could also explain the discrete AKR frequency structure. It might be feasible to stimulate AKR with the artificial waves from a satellite transmitter and thus learn a great deal more about AKR and its role in auroral phenomena.

Acknowledgments. R. R. Anderson, D. A. Gurnett, W. S. Kurth, M. L. Kaiser, and R. G. Stone deserve my sincere gratitude for productive discussions and enthusiastic encouragement. I am also grateful to T. E. Eastman and R. P. Lin for the particle analyses, and to F. L. Scarf and J. L. Steinberg for their ISEE-3 data. This work was supported by National Aeronautics and Space Administration Grant NGL-16-001-043 with NASA Headquarters, and by NASA contract NAS5-20093 with GSFC.

REFERENCES

Benson, R. F., and W. Calvert, ISIS-1 observations at the source of auroral kilometric radiation, Geophys. Res. Lett., 6, 479-482, 1979.

Benson, R. F., W. Calvert, and D. M. Klumpar, Simultaneous wave and particle observations in the auroral kilometric radiation source region, Geophys. Res. Lett., 7, 959-962, 1980.

Calvert, W., The signature of auroral kilometric radiation on ISIS-1 ionograms, J. Geophys. Res., 86, 76-82, 1981a.

Calvert, W., The aurora] plasma cavity, Geophys. Res. Lett., 8, 919-921, 1981b.

Dunckel, N., B. Ficklin, L. Borden, and R. A. Helliwell, Low-frequency noise observed in the distant magnetosphere with OGO-1, J. Geophys. Res., 75, 1854-1862, 1970.

Feller, W., An Introduction to Probability Theory and Its Applications, (Wiley, New York) 1950.

Green, J. L., D. A. Gurnett, and R. A. Hoffman, A correlation between auroral kilometric radiation and inverted-V electron precipitation, J. Geophys. Res., 84, 5216-5222, 1979.

Gurnett, D. A., The earth as a radio source: Terrestrial kilometric radiation, J. Geophys. Res., 79, 4227-4238, 1974.

Gurnett, D. A., F. L. Scarf, R. W. Fredericks, and E. J. Smith, The ISEE-1 and ISEE-2 plasma wave investigation, Geoscience Elect., GE-16, 225-230, 1978.

Gurnett, D. A., and J. L. Green, On the polarization and origin of auroral kilometric radiation, J. Geophys. Res., 83, 689-696, 1978.

Gurnett, D. A., R. R. Anderson, F. L. Scarf, R. W. Fredericks, and E. J. Smith, Initial results from the ISEE-1 and -2 plasma wave investigation, Space Sci. Revs., 23, 103-122, 1979.

Gurnett, D. A., R. R. Anderson, and R. L. Tokar, Plasma oscillations and the emissivity of type III radio bursts, Radio Physics of the Sun, M. R. Kundu and T. E. Gergely, eds., Int. Astron. Union (Reidel, Dordrecht, Netherlands), 367-379, 1980.

Kaiser, M. L., and J. K. Alexander, Relationships between auroral substorms and the occurrence of terrestrial kilometric radiation, J. Geophys. Res., 82, 5283-5286, 1977.

Kaiser, M. L., J. K. Alexander, A. C. Riddle, J. B. Pearce, and J. W. Warwick, Direct measurements by Voyagers I and 2 of the polarization of terrestrial kflometric radiation, Geophys. Res. Lett., 5, 857-860, 1978.

Knoll, R., G. Epstein, S. Hoang, G. Huntzinger, J. L. Steinberg, J. Fainberg, F. Greva, S. R. Mosier, and R. G. Stone, The 3-dimensional radio mapping experiment (SBH) on ISEE-C, Geoscience Elect., GE-16, 199-204, 1978.

Melrose, D. B., An interpretation of Jupiter's decametric radiation and the terrestrial kilometric radiation as direct amplified gyroemission, Astrophys. J., 207, 651-662, 1976.

Scarf, F. L., R. W. Fredericks, D. A. Gurnett and E. J. Smith, The ISEE-C plasma wave investigation, Geoscience Elect., GE-16, 191-195, 1978.

Wu, C. S., and L. C. Lee, A theory of the terrestrial kilometric radiation, Astrophys. J., 230, 621-626, 1979.

(Received April 7, 1981; revised July 13, 1981; accepted August 4, 1981.)