A. C. Fraser-Smith

Results of ULF magnetic field measurements near the epicenters of the Spitak (Ms = 6.9) and Loma Prieta (Ms = 7.1) earthquakes: Comparative analysis

The characteristics of the ULF magnetic field emissions measured at two magnetic observatories in the Republic of Georgia prior to and after the Ms = 6.9 earthquake that occurred near Spitak, Armenia, on December 7, 1988, are compared with the apparently similar emissions associated with the Ms = 7.1 earthquake that occurred near Loma Prieta, California, on October 17, 1989. The main features of the Spitak measurements, according to observations made at the Dusheti station (128 km to the Spitak epicenter), as compared with the Loma Prieta measurements, which were made at Corralitos, California (7 km to the Loma Prieta epicenter), are the following: (1) The intensity of ULF background activity started growing 3 to 5 days before the Spitak earthquake, whereas the corresponding increase in activity began 12 days before the Loma Prieta earthquake; (2) a substantial ULF emission burst was recorded at Dusheti starting 4 hours prior to the main shock; a similar large burst of ULF activity commenced 3 hours before the Loma Prieta event, and continued until the occurrence of the main shock; (3) ULF activity remained high for about two weeks after the Spitak earthquake, and for several months after the Loma Prieta earthquake; (4) ULF noise bursts were observed 1 to 6 hours before powerful aftershocks at Spitak during the period of enhanced activity, but there was no conclusive link between the ULF noise at Corralitos and the after-shocks. A major difference in the ULF activity preceding the two earthquakes is a difference in amplitude (0.2 nT at Spitak and 5 nT at Loma Prieta), but this is easily explained as being caused by the different distances of the observation stations from the epicenters.

ULF magnetic field measurements near the epicenter of the Ms 7.1 Loma Prieta earthquake

Abstract Anomalous measurements of the ultra-low frequency (ULF) magnetic field fluctuations prior to the magnitude 7.1 Loma Prieta earthquake of October 17, 1989, have been studied. For the past few years we have been monitoring fluctuations of the magnetic field of the Earth in the ULF range at Corralitos, California; our instruments were located 7 km from the epicenter of the earthquake. We have observed four anomalies in our data which may turn out to be precursors to the earthquake. First we observe narrowband noise fluctuations centered at 0.1 Hz and estimated to have a bandwidth of 0.00143–0.00167 Hz, i.e. a Q value of 60–70. The narrowband fluctuations appear to have a maximum equivalent amplitude of over 1400 pT/√Hz which is roughly 31 dB higher than the typical quiet average amplitude background levels. These fluctuations begin around September 12 and last until October 5. Next we observe the appearance of additive wideband noise fluctuations beginning around October 5 and continuing until the occurrence of the earthquake. These wideband fluctuations, which cover almost the entire 0.01–10 Hz frequency range of the system, have an average amplitude that is approximately 19 dB larger than typical levels in the 0.01–0.02 Hz band. Thirdly, we observe an atypical decrease in noise levels in the 0.2–5 Hz band throughout the day prior to the earthquake. The fourth anomaly is a jump in the power of magnetic field fluctuations, mostly in the 0.01–0.5 Hz band, in the three hours preceding the earthquake. This activity reached its highest level in the lowest 0.01–0.02 Hz band, and had a magnitude of roughly 60000 pT/√Hz, which is about 40 dB larger than typical background noise levels in the band. Our anomalous measurements do not appear to be the result of any magnetic field fluctuations generated in the upper atmosphere or to movement of our sensor caused by shocks preceding the quake. In describing these anomalous magnetic field fluctuations, both electrokinetic and piezoelectric theories of the generation of magnetic field fluctuations prior to earthquakes suggest wideband fluctuations. In our measurements, we have observed both wideband and narrowband fluctuations, so that our narrowband measurements do not seem to be easily explicable by the above theories. However, the wideband fluctuations could turn out to be attributable to the above mechanisms, since at the monitored frequencies, the skin depths of electromagnetic waves are comparable with the distances from our sensors to the earthquake focus, and therefore our measurements would not be unexpected.

Global lightning and climate variability inferred from ELF magnetic field variations

We report characteristics of global lightning activity on a seasonal time scale as inferred from two independently operating instruments measuring magnetic field variations in the lower ELF range. In this picture, the strongest source of lightning activity is centered on the North American continent. We reduce the source proximity effect to the observing sites experimentally with a simple differential technique and reconfirm the conclusion of Holzer and Deal (1956) that wave propagation conditions in the lower ELF range are of minor importance. The resulting mean seasonal variations of continental lightning in mid- and tropical latitudes are found to be related to surface temperature variations in moderate- and tropical rain forest climates and they support the view of Williams (1992) that monitoring of global lightning activity may provide a thermometer-independent measure of temperature changes associated with climate variability.

Seasonal variations of globally measured ELF/VLF radio noise

Our laboratory has been conducting a global survey of extremely low frequency (ELF) and very low frequency (VLF) radio noise since February 1985. Eight measurement stations around the world record the instantaneous noise amplitude in each of sixteen narrow-frequency bands in the 10-Hz to 32-kHz frequency range, and we have calculated the monthly averages of these amplitudes for the four stations with the longest times of operation. The period, amplitude, and phase of temporal variations in the averages are important indicators of the sources and propagation characteristics of the noise in the various frequency bands. Furthermore, since the principal source of ELF/VLF radio noise is lightning, long-term variations of the noise must relate to changes in global thunderstorm activity. We find that the noise amplitudes vary seasonally by up to a factor of 4 in some of the sixteen frequency bands; in addition, many of the variations correlate quite well with global lightning flash rates.

Ultra‐low frequency magnetic field measurements in southern California during the Northridge Earthquake of 17 January 1994

Measurements of ultra-low frequency (ULF) magnetic field fluctuations by two independent monitoring systems in Southern California were in progress during January 1994 when the moderately-large M6.7 Northridge earthquake occurred on 17 January. Our two measuring systems are located at Table Mountain, on the other side of the San Gabriel mountains and at a distance of 81 km from the epicenter, and at Pinon Flat, south of Palm Springs and at a distance of 206 km from the epicenter. Both systems operated well throughout the month and without interruption due to the earthquake. As a result of the occurrence of a moderate magnetic storm on 11 January, which was followed by a period of enhanced ULF magnetic activity that persisted until after the time of the earthquake, the sensitivity of our measurements throughout California was reduced for roughly a week before the earthquake took place. Nevertheless, no large signals that could be associated with the earthquake were evident at any time, except for the usual co-seismic shaking response of the detectors. Subsequent removal of the upper atmosphere signals from the Table Mountain measurements, using the measurements from the more distant Pinon Flat location as reference, essentially left no significant residual. Thus, assuming that ULF magnetic fields were produced by the earthquake, their amplitudes were too small to produce obvious increases in the ULF background noise at 81 km from the epicenter, which is in agreement with our earlier estimate of a range of about 100 km for the ULF magnetic field fluctuations observed prior to the M7.1 Loma Prieta earthquake. These results imply that a network of conventional magnetic field detectors spaced less than 100 km apart would be required to detect ULF magnetic field fluctuations prior to earthquakes with magnitudes greater than 7. Under the same conditions, superconducting magnetic field gradiometers could offer greater sensitivity and range.

A search for ELF/VLF emissions induced by earthquakes as observed in the ionosphere by the DE 2 satellite

Satellite observations of ELF/VLF wave activity by groups from both the Soviet Union and France have indicated the possibility of ELF/VLF radio emissions generated by earthquakes. However, an examination of ELF/VLF wave data from the low-altitude (apogee ∼ 1300 km, perigee ∼300 km, inclination ∼90°) Dynamics Explorer 2 (DE 2) satellite showed no clearly distinguishable ELF/VLF signatures associated with earthquakes. After an initial survey of approximately 5000 DE 2 orbits, ELF and VLF wave data were selected from 63 satellite orbits, called earthquake orbits, in which the ionospheric footprint of the DE 2 crossed the geographic latitude while passing within ±20° geographic longitude of the epicenters of imminent or recent earthquakes of magnitude ≥5.0. ELF/VLF noise measured near the epicenters was analyzed for occurrence rates and average spectra, as well as for peak and mean electric field intensities in three spectrometers covering a frequency range of 4 Hz - 512 kHz in 20 channels. The same analysis was then repeated for 61 carefully matched control orbits when there were no imminent or recent earthquakes within ±20° geographic longitude or within ±10° geographic latitude of the satellite footprint. These control orbits resembled the earthquake orbits with respect to latitude, longitude, local time, and geomagnetic index Kp. Sixty-three percent of the earthquake orbits showed an ELF or VLF emission above 10 µV/m in at least one of the 20 channels when the satellite passed near an epicenter. The same analysis performed on control orbit data yielded a 62% chance of observing similar emissions. Moreover, these results did not change when geomagnetic latitudes, instead of geographic latitudes, were considered. Further analyses failed to indicate any significant differences between the ELF/VLF noise measured on earthquake orbits and control orbits with regard to the general nature of the spectra, the frequency of occurrence of emissions, and peak and mean values of the electric field of the emissions.

Further evidence for a global correlation of the Earth-ionosphere cavity resonances

The Stanford ELF/VLF Radio Noise Survey provides digital time series in the ELF range and VLF analog recordings of one horizontal magnetic field component at different locations around the globe. Three sites, Arrival Heights in Antarctica, Sondrestromfjord in Greenland and Stanford in California, are chosen to investigate the natural magnetic noise in the frequency range of the Earth-ionosphere cavity resonances (6-60 Hz). The resonances are interpreted as pure resonance phenomena and the associated parameters are derived in a robust manner by use of the complex exponential algorithm. The resulting amplitudes of the first and second resonances exhibit a pronounced diurnal variation and a distinct day to day variability of the same order during the time interval from January to April 1990. The day to day variability is well correlated between all of the three different sites. They also exhibit a 20-30 day variation that is related to the solar rotation period, expressed by means of sunspot numbers.

The Manifestation of the Madden–Julian Oscillation in Global Deep Convection and in the Schumann Resonance Intensity

Abstract This study determines the relationship between intraseasonal oscillations observed in two independent measures of global lightning activity: a global mean convective index (a proxy for deep convection) derived from the Goddard Television Infrared Observational Satellite (TIROS) Operational Vertical Sounder (TOVS) Pathfinder infrared cloud observations, and Schumann resonance magnetic intensity recorded at Arrival Heights, Antarctica. The study was initiated when previous results indicated a possible link between intraseasonal variations in Schumann resonances and variability of sunspot numbers on the timescale of the solar rotation period. The authors used seven years (1989–95) of daily records, though the Schumann resonance record had a number of gaps. Results of cross-spectrum and composite analysis show that intraseasonal oscillations in deep convection modulate the global variations in the Schumann resonance intensity. In the Tropics, the intraseasonal wave in deep convection has a wavenumber...

ULF magnetic fields generated by electrical storms and their significance to geomagnetic pulsation generation

The STAR Laboratory has been operating two independent electromagnetic noise monitoring systems in the vicinity of Stanford University for several years, one covering the ULF range (0.01–10 Hz) and the other the ELF/VLF range (10 Hz – 32 kHz). These two systems were operating during the relatively rare occurrence of a large local electrical storm on 23 September 1990 and their measurements (which cover 25 narrow frequency bands in the more than five-decade range 0.01 Hz – 32 kHz) provide a detailed picture of the ULF and ELF/VLF magnetic field fluctuations that accompanied the storm. As would be expected, considering the important role played by lightning in its generation, ELF/VLF radio noise is enhanced. However, the ELF/VLF enhancements are smaller than those that occur at ULF: the amplitude of the magnetic field fluctuations in a 100 Hz band centered on 2 kHz was approximately doubled, whereas in the range 0.2–0.5 Hz the amplitude was increased roughly 20 times, and the increase was about 50 times for the range 0.01–0.02 Hz. Some years ago I proposed that repetitive lightning in electrical storms could trigger Pc 1 geomagnetic pulsation events, but these new results suggest that there may exist a broader link between electrical storms and geomagnetic pulsations. Considering that there are perhaps 2000 electrical storms in progress around the world at any time, and that large ULF electrical current systems below the ionosphere can stimulate ULF hydromagnetic waves in the ionosphere, the new measurements suggest that the electrical currents associated with thunderstorms, taken as a whole, may be a source of ULF energy in the magnetosphere and thus influence the generation and amplification of some classes of geomagnetic pulsations.

The planetary rate of sprite events

We propose a new formula to calculate the planetary rate of sprite events, based on observations with sprite detectors. This formula uses the number of detected sprites, the detection efficiency and the false alarm rate of the detector and spatial and temporal effectiveness functions. The role of these elements in the formula is discussed for optical and non-optical recordings. We use the formula to calculate an average planetary rate of sprite events of similar to 2.8 per minute with an accuracy of a factor similar to 2 - 3 by use of observations reported in the literature. The proposed formula can be used to calculate the occurrence rate of any physical event detected by remote sensing.