J.L. POSCH and M.J. ENGEBRETSON, Department of Physics, Augsburg College, Minneapolis, Minnesota 55454
R.L. ARNOLDY, Space Science Center, University of New Hampshire, Durham, New Hampshire 03824
Several recent studies of magnetospheric substorms have attempted to use impulsive ultra-low-frequency (ULF) wave bursts, known as Pi 1 and Pi 2 pulsations (with periods of 1-40 and 40-150 seconds, respectively) as markers of the time of substorm onset. Bösinger and Yahnin (1987), for example, attempted to ascertain whether Pi 1 is a better timer for substorms than Pi 2 because their shorter periods might make it easier to get an onset or maximum time. To study the temporal and spatial characteristics of Pi 1 bursts as a substorm signature, a large-scale array of magnetometers is needed. The U.S. AGO (automatic geophysical observatories) network is one such array. Another array of stations very useful for this type of study is the MACCS (Magnetometer Array for Cusp and Cleft Studies) array in Arctic Canada. Two pairs of nominally conjugate stations are Amundsen-Scott South Pole Station and Pangnirtung and AGO P1 and Clyde River.
Magnetometer data from antarctic stations including four AGOs, South Pole Station, and McMurdo Station, which are all equipped with search coil magnetometers, were studied in search of Pi 1 events as described by Heacock (1967). In this article, we focus on data in the frequency range from 0 to 1,000 millihertz (mHz) from 18:00 universal time (UT) on 7 May 1995 to 04:00 UT on 8 May 1995. Many Pi 1 bursts occurred throughout this 10-hour interval, which was characterized by large values of the planetary Kp index (4-5) and a high solar wind velocity [700-800 kilometers per second (km/sec)]. The most powerful burst, which occurred near 23:30 UT, was seen at all stations except McMurdo, which is located much farther west than the other stations. Since this burst is the most powerful and extends well beyond 500 mHz at the AGOs and South Pole, it will be the focus of further investigation.
Many stations in the Northern Hemisphere observed similar activity as that seen in Antarctica on 7-8 May 1995. Six of these are MACCS stations that are equipped with fluxgate magnetometers. The stations near 75° magnetic latitude include Cape Dorset, Coral Harbour, and Pangnirtung. Those near 79° magnetic latitude include Gjoa Haven, Pelly Bay, and Clyde River. All of these stations recorded the Pi 1 burst near 23:30 UT, but it was clearly stronger at the 75° stations. The 79° stations observed a much weaker burst, which became more faint as one moved westward across the chain of stations.
Figure 1 shows summed power over 300-500 mHz vs. universal time. Each graph summarizes 2 hours of data centered around midnight UT on 7 May 1995 for the five antarctic stations that see the Pi 1 activity near 23:30 UT. The time of the first large, sharp peak is indicated on each graph. The timing between stations is shown below the AGO P4 graph. One can see, pairs with similar latitude near 80° occur within 1 minute of each other. The pairs near 70° magnetic latitude occur more than 6 minutes apart. The pairs with similar longitude show that the Pi 1 burst occurs first at the 70° stations and then propagates poleward.
Figure 2, which is of the same format as figure 1, shows the Arctic MACCS stations for the same interval. The pattern that is seen in the Antarctic is seen here as well. The Pi 1 bursts seen at the MACCS stations near 80° MLAT occur within 10 seconds of one another. These 80° stations range over 55.9° of magnetic longitude. The bursts at the 75° stations, ranging over 30.6° longitude, occur over a much larger timescale, with a delay of on the order of 1-4 minutes.
Although the timing of this Pi 1 burst is interesting among stations within a hemisphere, it is also very important to study the timing between conjugate stations. To get the timing needed, we compare figure 1 and figure 2. The first conjugate pair, Pangnirtung and South Pole, occur about 45 seconds apart. The second pair, Clyde River and AGO P1 occur about 30 seconds apart. This example day with nightside Pi 1 activity indicates that such activity occurs nearly simultaneously at conjugate sites near both 75° magnetic latitude (MLAT) and 80° MLAT.
Previous studies of midlatitude stations using the Air Force Geophysics Laboratories (AFGL) midlatitude chain across the United States starting in the 1970s have shown that the Pi 1 bursts had the same onset time at many different nighttime longitudes (Knecht and Singer 1981). This simulataneous onset time does not appear to be the case at high-latitude arrays such as MACCS and the U.S. AGO network where, as shown above, onset times at stations with similar latitudes and different longitudes may vary from 10 seconds to 6 minutes, depending on the latitude of the chain of stations. Onset times at stations with similar longitudes and different latitudes varied by as much as 24 minutes (figure 1). This latitudinal effect of the Pi 1 burst onset time could just be the auroral motion poleward after break-up as suggested by Kokubun et al. (1988). This example has also shown that activity does not occur simultaneously at sites with different latitudes or longitudes. Important to note, however, is that stations near 80° MLAT varied less in time (approximately 1 minute or less) than stations near 70°-75° MLAT, which varied by as much as 6 minutes.
Possibly, Pi 1 could serve as a ground signature of ion cyclotron waves that are believed to be the source of the anomalous resistivity of the auroral acceleration region. Further study of the conjugate Pi 1 bursts to look for seasonal effects will be needed to understand more fully the importance of the simultaneous nature of Pi 1 bursts.
We thank W.J. Hughes of Boston University and H. Fukunishi of the Tohoku University for their contributions to this study.
This research was supported by National Science Foundation grants OPP 95-29177, OPP 96-13683, ATM 96-10072, and ATM 97-04766.
Bösinger, T., and A.G. Yahnin. 1987. Pi1B type magnetic pulsations as a high time resolution monitor of substorm development. Annales Geophysicae , 5A, 231-238.
Heacock, R.R. 1967. Two subtypes of type Pi micropulsations. Journal of Geophysical Research , 72, 3905-3917.
Knecht, D.J., and H.J. Singer. 1981. Multistation observations of simultaneous Pi 1 and Pi 2 pulsations at substorm onsets. EOS, Transactions of the American Geophysical Union , 62, 358.
Kokubun, S., T. Yamamoto, K. Hayashi, T. Oguti, and A. Egeland. 1988. Impulsive Pi bursts associated with poleward moving auroras near the polar cusp. Journal of Geomagnetism and Geoelectricity , 40, 537-551.