Spectral Studies of Ionospheric and Solar Wind Electric Fields: Case Studies and Long-Duration Analysis Using Magnetometer Data M. C. Kelley, M. J. Nicolls, and G. Sachdeva Cornell University, Ithaca, NY USA. J. L. Chau and O. Veliz Jicamarca Radio Observatory, Peru. D. Anderson and A. Anghel NOAA, Boulder, CO USA.
Penetrating electric fields: How do they work? Rapid changes in magnetospheric convection can create electric field changes inside the plasmasphere. This allows the interplanetary medium to control the low-latitude electric field.
Historical observations of penetrating fields Kelley et al. [1979] Gonzales et al. [1979] Earle and Kelley [1987]
Questions. 1. What controls the variability and onset of PPE? 2. What frequency scales do PPE affect? 3. How do the fluctuations in the controlling solar wind affect the fluctuations observed at equatorial stations? 4. What role does the system itself play in this process? 5. Are the effects global? Low-latitude daytime magnetometer observations can provide a tracer for PPE. We use 5 years of magnetometer observations to investigate these questions, looking at: (a) (b) Case studies. Long-duration characteristics.
Magnetometer observations Magnetometer observations can be used to study longer term effects of solar wind and magnetospheric electric fields, with nearly continuous observations over many years. Using an on-equatorial magnetometer (e.g., Jicamarca) and an offequatorial magnetometer (e.g., Piura), the difference in the H component can be computed which removes the ring current effect. Dave Anderson has shown that H is a tracer for the vertical drift at the equator, and has constructed a neural network to predict such drifts given a H measurement. We remove the quiet-day contributions to the magnetometer-deduced E-field using the Scherliess and Fejer [1999] empirical model.
Jicamarca-Piura Data Coverage - Data from 2000 to 2004. - Moderate-high solar flux conditions, good Kp coverage
Fluctuation power versus Kp The fluctuation spectra show a clear dependence on Kp at both the equator and in the solar wind. Here we show the integrated spectrum from 0.15 to 10 cycles/hour versus Kp, along with the line of best fit. The dashed lines are the empirical results of Mozer [1971]. The horizontal line is the quiet day average (<Kp> < 3).
Average fluctuation spectra The average spectra (binned by Kp) show a -2 power spectrum for the mag data and a -5/3 power spectrum for the SW data (characteristic of MHD Hall turbulence), which was also observed in measurements of mid/high latitude e-fields by Mozer [1971]. The fluctuation power increases exponentially with Kp in both the solar wind and at the equator.
Transfer Function The observations imply a system response to the fluctuating solar wind. To investigate this system response, we can estimate the transfer function as the ratio of the estimate of the cross power spectral density to the estimate of the power spectral density. We also estimate the coherency. The results show at minimum a high-pass response and possibly a resonant response.
Case Study: March 31, 2001 Application of the average transfer function can be used to quantify the role of penetration electric fields. This example, the March 31, 2001 storm event (DST < -400 nt) has been investigated using global models [Maruyama et al., 2005]. This event had a step-like input for our time of interest. Penetration event Increased power, perhaps DD
Case Study: April 17, 2002 The April 17, 2002 event was characterized by an oscillatory IEF. The TF model captures the temporal variation, but underestimates the magnitude of the oscillations. I.e., the TF model understimates the power at the oscillation frequency, which is near 1 hour.
Case Study: July 27, 2004 A final case study is shown for July 27, 2004. This study again shows that the TF model captures some of the oscillations in the equatorial e-field fairly well. Some sharp changes in the equatorial e-field are not due to PPE. There are about 14 days in our dataset (2000-2004) that show penetration effects. Not due to penetration? Penetration event?
Resonant Resonance? - RLC Circuit Model Defining b as the L shell: From the formula for the inductance of a ring: L = 1000 b 5 Henrys, From the formula for the capacitance of a sphere: C = 165 b 5 7 Farads. The resonant period is then τ = 0.7 b 7 4 hours The Q of the system is, Q = 2RCω 0 = 2 I 1 C L 1.5 Compare to the Vasyliunas time constant, 8 hours, which gives L= 28800 H
Summary Summary - Spectral analysis of five years of zonal electric field estimates using the Anderson method are in good agreement with an earlier study [Earle and Kelley, 1987] using about 15 events. - Comparison with middle and high latitude electric field spectra show that in the frequency range 0.15-1 hour the equatorial electric field is essentially uniform all the way to the equatorial ionosphere. - Evidence is presented for a resonant response rather than a simple high pass filter. - Application of the transfer function to the IEF yields good agreement for a step-like function input on March 31, 2001. - Some features not reproduced may be due to mid-frequency gravity waves related to magnetic activity. This research was supported by NSF grant ATM-0000196.