The Solar Wind Control of the Power and Occurrence Rate of ULF Waves Observed at AFGL Ground Magnetometer Network

P. J. Chi, C. T. Russell, G. Le
Institute of Geophysics and Planetary Physics
University of California, Los Angeles

R. M. Bloom
Pacific-Sierra Research Corporation
Santa Monica, CA

H. J. Singer
NOAA Space Environment Laboratory
Boulder, CO

INTRODUCTION

The relationships between the Ultra-Low-Frequency (ULF) wave activity in the dayside magnetosphere and interplanetary parameters provide clues as to the generation mechanisms of these waves. The solar wind speed and the interplanetary magnetic field (IMF) direction have been found to be the two key interplanetary parameters affecting dayside ground magnetic activity. Waves generated in the foreshock region due to the ion cyclotron instability are thought more easily coupled to the magnetosphere as the IMF becomes more aligned with the earth- sun line. Increases in the solar wind speed lead to higher plasma flow along the magnetopause which can excite the Kelvin- Helmholtz instability at the magnetopause. Both waves described above can excite the field line resonance in the magnetosphere and increase magnetic activity on the ground.

Here we study statistically how the ULF wave activity recorded at low-latitude ground stations relate to interplanetary conditions.

DATA

Power density spectra (PSD) for all seven ground stations of the Air Force Geophysical Laboratory (AFGL) magnetometer network (Figure 1) have been calculated by Bloom and Singer [1995] in six frequency bands: 2-4 mHz, 4-8 mHz, 8-16 mHz, 16-32 mHz, 32-64 mHz, and 64-128 mHz for the years of 1978 and 1980.

The table below shows the six frequency bands used by Bloom and Singer [1995] versus the traditional classifications of geomagnetic pulsations.

                   Bloom & Singer [1995]  Pc classifications
                   -----------------------------------------
                    2 -   4 mHz           Pc 5
                    4 -   8 mHz           Pc 4, Pc 5
                    8 -  16 mHz           Pc 4
                   16 -  32 mHz           Pc 3, Pc 4
                   32 -  64 mHz           Pc 3
                   64 - 128 mHz           Pc 2, Pc 3

                                    Table 1

In this study, twenty-minute averages of the dayside wave power in the H-component recorded at Mt. Clemens and the solar wind parameters recorded by the IMP 8 spacecraft are used. There are a total of 2957 20-minute intervals.

WAVE POWER

Figure 2 shows the scatter plots of the integrated magnetic power spectral densities, represented as log power (nT^2), versus solar wind speed. For all six frequency bands, wave power increases when solar wind speed increases. The lower the wave frequency, the greater the solar wind speed affects the wave power. The lines in the scatter plots show a robust and local smoothing of the data.

Figure 3 shows similar plots for the wave power and the IMF cone angle, arc cos(BX/BT). For the two lowest frequency bands (2-8 mHz), wave power is stronger when the IMF direction is away from the earth-sun line. For the next three higher frequency bands (8-64 mHz), wave power is stronger when the IMF cone angle is small. The wave power of the highest frequency band (64-128 mHz) is approximately constant for all IMF cone angles.

Figure 4 shows the plots of the wave power and the IMF clock angle, arc tan(BY/BZ). Except for the two highest frequency bands, wave power is stronger when IMF is southward, i.e., when the IMF clock angle is close to ±180 degrees.

Figure 5 summarizes the results of Figures 3, 4, and 5.

OCCURRENCE RATES OF WAVE EVENTS

The ULF waves in the dayside magnetosphere usually appear in a form of quasi-sinusoidal continuous pulsations. These pulsations have apparent peaks in the power spectra. We use the PSD data at Mt. Clemens again to select wave events from different frequency bands. The events for each frequency band are selected by identifying peaks in the spectrum by choosing times when

    Pi > Pi+1     i = 2, 3, 4, 5.
AND Pi > Pi-1     Pi: wave power for the ith frequency band.

When compared with simultaneous solar wind data, these selected events allow us to examine the interplanetary conditions favorable for the occurrence of these events. Figure 6 shows the distribution of the IMF cone angle for all the data with or without wave events. The number on the top of each bar represents the number of data points for the bar. The same cone angle distributions are shown for times when wave events occurred in each frequency band (e.g., 32-64 mHz) in Figure 7. To determine which are the favorable conditions for the occurrence of wave events, the normalized occurrence rate is calculated as:

(Normalized Occurrence Rate)i
                  Occurrence Rate for the ith Frequency Band
                = ------------------------------------------
                   Occurrence Rate for all the Data Points

where
                          Number of Data in the Bin
       Occurrence Rate = ---------------------------
                         Number of Data for all Bins
Figure 8 shows the normalized occurrence rates for the four frequency bands with respect to the IMF cone angle. The number of events for each bin is also shown on the top of the bar. For the three frequency bands, 8-16 mHz, 16-32 mHz, and 32-64 mHz, overlapping with the Pc 3, 4 band, the occurrence of wave events was favored by low IMF cone angles. This is consistent with the model that the source of Pc 3, 4 waves is the upstream waves in the foreshock region.

Figure 9 is the distribution of IMF magnitude for all the data points, and Figure 10 shows the normalized occurrence rates for the four frequency bands with respect to the IMF magnitude. Notice that some bars near the both sides of the plot only contain few events so the errors corresponding to them could be large.

The predicted frequency of upstream waves can be expressed as a simple formula:

f (mHz) = 6 * B_imf (nT).

This is also the expected frequency for many Pc 3, 4 waves in the dayside magnetosphere if they are simply driven by the upstream waves. In Figure 10, the ranges of predicted frequencies of upstream waves for the four frequency bands are shown under the horizontal axes and the peaks of the normalized occurrence rates are marked if identified. For 4-8 mHz waves, the IMF magnitude does not affect the occurrence of wave events. For the three higher frequency bands, the peak of the normalized occurrence rate shifts toward stronger IMF magnitudes as the frequency of waves is higher. The peaks for 16-32 mHz and 32-64 mHz waves (~Pc 3) are within the predicted frequencies of upstream waves. This is also consistent with the upstream wave generation model. The peak for 8-16 mHz waves (~Pc 4) is higher than the predicted frequencies of upstream waves. A possible explanation is that many Pc 4 waves are in fact harmonics of fundamental field line resonance and these harmonic resonances are driven by a broad band upstream wave.

CONCLUSIONS

1. Strong solar wind speed enhances the ULF wave activity at low-latitude ground stations.

2. For the lower frequency part of ULF waves, the wave activity is stronger when the IMF is southward and the IMF cone angle is large. The higher frequency part of ULF waves is less sensitive to the IMF clock angle and the wave activity is stronger when the IMF cone angle is small.

3. The results of the occurrence rate study of the selected wave events support the upstream wave generation model for Pc 3, 4 waves.

REFERENCE

Bloom, R. M., and H. J. Singer, Diurnal trends in geomagnetic noise power in the Pc 2 through Pc 5 bands at low geomagnetic latitudes, J. Geophys. Res., in press, 1995.


Back to the GEM online poster page

Last modified: July 20, 1995

Peter Chi
pchi@igpp.ucla.edu