157

Midlatitude Ionospheric Dynamics and DisturbancesGeophysical Monograph Series 181Copyright 2008 by the American Geophysical Union.10.1029/181GM15

Relating the Interplanetary-Induced Electric Fields With the Low-Latitude Zonal Electric Fields Under Geomagnetically Disturbed Conditions

Adela Anghel,1,2 David Anderson,1,2 Jorge Chau,3 Kiyohumi Yumoto,4 and Archana Bhattacharyya5

The overall ionospheric variability with periods ranging from long-term, secular changes to days, hours, and even minutes and seconds, is influenced by the solar activity, geomagnetic activity, and processes originating in the lower atmospheric layers. Using a wavelet transform approach, in this paper, we study the short-term (minutes to hours) and day-to-day variability of the ionospheric low-latitude zonal electric fields (LLZEF) at three longitude sectors, Peruvian, Philippine, and Indian, during time intervals of increased geomagnetic activity and relate the LLZEF variability to changes in the dawn-to-dusk component of the interplanetary electric field (IEF). Continuous Morlet wavelet and cross-wavelet amplitude spectra with reduced and increased frequency resolutions were obtained to analyze and compare the oscillation activity in the LLZEF and IEF spectra, in the 10-min to 10-h and 1.25- to 12-d period ranges. For the 1.25- to 12-d period range, periodicities in the LLZEF spectrum were compared with similar periodicities in the IEF spectrum over 9 February to 9 June 2001, with our wavelet results indicating the geomagnetic activity as an important driver of LLZEF variability in this period range. For the 10-min to 10-h period range, four case studies were examined when concurrent observations of Jicamarca incoherent scatter radar zonal electric field and IEF, as calculated from the ACE satellite solar wind velocity and interplanetary magnetic field data, were available. We show that the wavelet transform represents a powerful tool to study the frequency dependence of the two specific mechanisms of ionospheric electric field variability, which are dominant during geomagnetic storms, namely penetration and disturbance dynamo.

1 Cooperative Institute for Research in Environmental Sciences, University of Colorado, Boulder, Colorado, USA.

2 Space Weather Prediction Center, NOAA Boulder, Colorado, USA.

3 Radio Observatorio de Jicamarca, Instituto Geofisico del Peru, Jicamarca, Peru.

4 Space Environment Research Center, Kyushu University, Fuku-oka, Japan.

5 Indian Institute of Geomagnetism, New Panvel, Navi Mumbai, India.

1. INtROduCtION

Stemming from the need to understand and predict the ionosphere behavior and its deviations from the normal climatological mean, under both quiet and disturbed con-ditions, recent modeling and observational studies have shown an increased interest in the short-term (minutes to hours) and day-to-day variability of the upper atmosphere and ionosphere [e.g., Mendillo and Schatten, 1983; Parish et al., 1994; Forbes et al., 2000; Pancheva et al., 2002]. The ionosphere–thermosphere system is a complex system, its complexity being determined by (1) inherent internal inter-actions occurring inside the system, (2) interactions with the magnetosphere above, where space plasma processes in the magnetosphere caused by its coupling to the solar wind

158 RELAtING thE INtERPLANEtARy-INduCEd ELECtRIC FIELdS WIth thE LLZEF

provide an interface with highly variable inputs of electro-dynamic energy and energetic particles, (3) interactions with the middle atmosphere below, itself modulated by tropo-spheric weather and surface topology, and (4) variability of the external sources driving the system. As a result, the iono-sphere displays variations from its normal patterns that af-fect the ionospheric predictions on timescales ranging from secular to days, hours, and even minutes and seconds. These variations have been observed in different ionospheric pa-rameters [e.g., Parish et al., 1994; Rishbeth and Mendillo, 2001; Pancheva et al., 2002; Fagundes et al., 2005; Abdu et al., 2006] possibly induced by wave activity originating in the lower regions of the atmosphere, quasiperiodic oscil-lations in the geomagnetic activity, or other triggers such as the periodic variability of the solar radiation flux. the solar radiation influences mostly the long-term (months to years) variability, while the geomagnetic activity and the lower atmosphere processes can induce oscillations with periods ranging from about few seconds or minutes to several days or weeks [e.g., Laštovička, 2006].

In this paper, the main focus is on the variability of the ionospheric low-latitude zonal electric field (LLZEF) in re-sponse to changes in the dawn-to-dusk component of the interplanetary electric field (IEF) during time intervals of increased geomagnetic activity. During storm times, large ionospheric electric field and current perturbations travel from high to equatorial latitudes, changing the ionization distribution over large areas and controlling the storm-time dynamics and electrodynamics of the low- and mid-latitude ionosphere. The most important sources of low-latitude elec-tric field perturbations during storm times are (1) the prompt penetration electric fields of solar wind/magnetospheric origin and (2) the atmospheric disturbance dynamo electric fields due to auroral Joule heating and ion-drag acceleration. the penetration electric fields are associated with changes in the field-aligned current system responsible for shielding the inner magnetosphere and mid- and low-latitude ionosphere from the high-latitude magnetospheric convection electric fields and propagate instantaneously to equatorial latitudes in response to changes in the magnetospheric convection [e.g., Fejer and Scherliess, 1997; Huang et al., 2005]. On the other hand, the disturbance dynamo electric fields are as-sociated with enhanced deposition of energy and momentum in the auroral zone that causes winds to develop, producing predominantly westward electric fields on the dayside and eastward on the nightside at equatorial latitudes [e.g., Blanc and Richmond, 1980; Scherliess and Fejer, 1997].

Equatorial electric field measurements are rather sporadic, but recently, Anderson et al. [2004] developed a neural net-work-based plasma drift model, for both quiet and disturbed conditions, that uses ground-based magnetometer obser-

vations from pairs of equatorial stations to infer realistic, daytime equatorial vertical E ´ B plasma drift velocities wherever appropriately placed magnetometers exist. Using their neural network drift model, Anderson et al. [2006] and Anghel et al. [2007] calculated seasonally averaged patterns of quiet-time vertical E ´ B drifts at the Peruvian, Philip-pine, and Indian longitude sectors, showing that there is a very good agreement with drift patterns obtained with the global vertical drift model developed by Scherliess and Fe-jer [1999]. In addition, Anghel et al. [2007] also conducted an investigation on the variability of the daytime LLZEF with respect to changes in the IEF conditions. Expanding on their studies, in our paper, we use wavelet and cross-wavelet spectral analyses to compare the oscillation activity in the LLZEF and IEF spectra in the 10-min to 10-h and 1.25- to 12-d period ranges. Therefore, the purpose of this paper is twofold: (1) to study the variability of the daytime LLZEF in the 1.25- to 12-d period range at three longitudes, Peruvian, Philippine, and Indian, over a time interval of relatively in-creased geomagnetic activity, 9 February to 9 June 2001, and relate this variability to similar changes in the IEF and (2) to analyze and relate the oscillation activity in the LLZEF and IEF in the 10-min to 10-h period range for three case studies characterized by enhanced geomagnetic activity, the 17–19 April 2001, 15–18 April 2002, and 9–12 November 2004 storm events, using concurrent observations of Jicamarca in-coherent scatter radar (ISR) zonal electric field and IEF data in a wavelet analysis approach. A fourth quiet-time case, 29 March to 2 April 2004, is also considered for comparison purposes.

The paper is organized as follows: in the next section, we briefly describe the data sets and the wavelet analysis ap-proach, then we present the wavelet results for the 1.25- to 12-d period range, followed by an examination of the four case studies for periodicities in the 10-min to 10-h range, and in the last section, we present succinct our conclusions.

2. dAtA SEtS ANd ANALySIS MEthOdS

2.1. Data Sets

A direct measure of the strength of the equatorial elec-trojet current and of the magnitude of the F region vertical E ´ B drifts is provided by the difference in the horizon-tal H components, DH, between a magnetometer placed on the magnetic equator and one displaced 6–9° away, af-ter subtracting the nighttime baseline at each station [e.g., Anderson et al., 2004]. For our study, we used magnetom-eter data with a 5-min time resolution from three pairs of equatorial stations located in Peru, Philippines, and India. Magnetometer observations at the Peruvian sector were

ANGhEL Et AL. 159

obtained from Jicamarca (geographic coordinates, 11.9°S, 283.1°E, geometric latitude 0.8°N) and Piura (geographic coordinates, 5.2°S, 279.4°E, geometric latitude 6.8°N), at the Philippine sector, from davao (geographic coordinates, 7°N, 125.4°E, geometric latitude 1.32°S) and Muntinlupa (geographic coordinates, 14.37°N, 121.02°E, geometric latitude 6.39°N), and at the Indian sector, from tirunel-veli (geographic coordinates, 8.7°N, 76.9°E, geometric latitude 0.5°S) and Alibag (geographic coordinates, 18.6°N, 72.9°E, geometric latitude 10°N). the neural network drift model described by Anderson et al. [2004] and the calcu-lated ΔH values at each longitude sector were then used to estimate the daytime vertical E ´ B drifts and, in turn, the equatorial zonal electric fields knowing that 1 mV/m cor-responds to a vertical drift of ~40 m/s at the Peruvian sector, ~28 m/s at the Philippine sector, and ~25 m/s at the Indian sector.

In our analysis, the interplanetary conditions are described using 64-s averages of merged ACE Magnetic Field Experi-ment (MAG)–Solar Wind Electron, Proton, and Alpha Mon-itor (SWEPAM) Level 2 interplanetary magnetic field and solar wind velocity data, where the two parameters are im-portant in establishing the dawn-to-dusk component of the IEF, IEF Ey. First, the calculated IEF-Ey at the spacecraft position L1 (~1.4 million km) is time-shifted to the mag-netopause position using the radial component of the solar wind velocity, in addition to a 10-min time delay. Then, the time-shifted IEF Ey is smoothed by using a sliding-window average procedure, which is equivalent to a low-pass filter-ing of the data. By using the time-shifting and smoothing procedures, we obtain a very good correlation between the IEF and the equatorial zonal electric fields at all three longi-tude sectors [e.g., Kelley et al., 2003; Anghel et al., 2007].

2.2. Continuous Morlet Wavelet

Over the last few years, the continuous wavelet transform [e.g., Kumar and Foufoula-Georgiou, 1997; Torrence and Compo, 1998] has become a favored tool to analyze peri-odicities that occur in the atmosphere [e.g., Pancheva, 2000; Abdu et al., 2006]. The wavelet method targets nonstation-ary signals with variable frequency content like the ones we deal with here. For our analysis, we favored the continu-ous Morlet wavelet transform to relate the variability in the LLZEF and IEF. The Morlet wavelet is a complex-valued function consisting of a plane wave modulated by a Gaus-sian envelope:

t1

2 1 4exp

t 2

2 2exp j ot , (1)

where the parameters σ and ωo control the tradeoff between the time and frequency resolution. here, we chose σ = 1, and for ωo, we selected different values throughout the paper, where large ωo values correspond to increased frequency resolution in the wavelet domain. An important advantage of the wavelet method described here consists in its abil-ity to extract amplitude information about the periodici-ties present in the signal spectrum. The method can also be used to calculate “instantaneous” frequency response func-tions in the wavelet domain, defined in a similar way like in the Fourier domain at each time instant. The “classical” frequency response function in the Fourier domain can then be obtained by averaging over time the “instantaneous” fre-quency response functions, with the specific consequences of averaging. Also, to obtain information about the simul-taneous presence of similar periodicities in different sig-nals, we used a type of cross-wavelet analysis defined as a geometric mean of the wavelet amplitude spectra of the signals [e.g., Manson et al., 2005]. In all our wavelet plots, the statistical significance levels were calculated based on a first-order autoregressive parametric spectral estimate of the power spectrum [Roberts and Mullis, 1987] consid-ered as a background spectrum, multiplied by the desired percentile value for the χ2 distribution with two degrees of freedom.

In the following, three examples of simulated wave-let spectra are presented to familiarize the reader with the method, for a better understanding of our results. the first example refers to Plate 1a and shows the wavelet amplitude spectra of a signal obtained by superimposing harmonics of amplitude 1 at different time intervals. As shown in Plate 1a (left), for wo = 6 there is a good time resolution but reduced frequency resolution in the wavelet domain, with strong beatings between relatively closed spectral components, ap-pearing as amplitude modulations. In Plate 1a (right), ωo is 12, and the spectral components appear as distinct spectral lines well localized in time. The contour lines in the wavelet spectra represent the 95% significance levels, and the two slant lines define the cone-of-influence where the edge ef-fects become significant.

Our second example refers to an amplitude modulation case and is presented in Plate 1b. In this case, a 30-d-long signal of 1-d period is modulated by a 5-d period:

x t cos 2 t Tp1 1 cos 2 t Tp2

cos 2 t Tp1 0.5cos 2 t Ts 0.5cos 2 t Td (2)

In equation (2), the primary periods are Tp1 = 1 d and Tp2 = 5 d, and the sum and difference secondary periods are Ts=Tp1Tp2/

160 RELAtING thE INtERPLANEtARy-INduCEd ELECtRIC FIELdS WIth thE LLZEF

(Tp1+Tp2) = 0.833 d and Td=Tp1·Tp2/(Tp2 - Tp1) = 1.25 d. For ωo = 6 (Plate 1b, left), the two secondary periods beat with the primary 1-d period producing a 1-d period modulated in amplitude by a 5-d period. For ωo = 30 (Plate 1b, middle), distinct spectral lines corresponding to the two secondary periods and to the primary 1-d period can be distinguished. Plate 1 (right) shows the three spectral components at day 15. In general, in a wavelet domain with reduced frequency resolution, two simultaneous and relatively closed periods, T1 and T2, may beat with each other with the beating period T = T1T2/|T1 - T2| and produce an amplitude modulation ef-fect that may appear as distinct and unrelated bursts of oscil-lation activity.

In our third example, illustrated in Plate 1c, the signal is generated as:

x t 2cos 2 t Tp1 cos 2 t Tp2

cos 2 t Ts cos 2 t Td , (3)

where Tp1= 1 d, Tp2= 10 d, Ts= 0.9091 d, and Td = 1.11 d. For ωo = 6 (Plate 1c, left), the two secondary periods beat with each other with a 5-d beating period, producing, like in our second example, a 1-d period modulated in amplitude by a 5-d period, although different spectral components are present in this case. For ωo = 30, two distinct spectral lines of amplitude 1 are clearly noticed in the wavelet spectrum (Plate 1c, middle) and in the corresponding plot for day 15 (Plate 1c, right).

We conclude that, in most cases, a time domain analy-sis of nonstationary signals by itself, which is very similar with our wavelet analysis for ωo = 6, might not be suf-ficient. At the other extreme, a Fourier analysis might not help much either since the time information in this case is completely lost. however, as shown here, the wavelet analy-sis can provide an accurate way of analyzing nonstationary signals in the time–frequency domain over large frequency bands.

3. LLZEF PERtuRBAtIONS OF GEOMAGNEtIC ORIGIN

3.1. LLZEF Perturbations in the 1.25- to 12-d Period Range

Previous studies have shown that the geomagnetic activity is an important driver of planetary wavelike oscillations (pe-riods in the 2- to 30-d range) in the ionosphere [Altadill and Apostolov, 2001; Forbes et al., 2000; Rishbeth and Men-dillo, 2001; Pancheva, 2002]. Therefore, in this section, we relate periodicities in the 1.25- to 12-d range that are present in the LLZEF spectra corresponding to the Peruvian, Philip-

pine, and Indian longitude sectors, to similar periodicities in the IEF spectrum using the continuous wavelet approach de-scribed above. the zonal electric fields at the three longitude sectors were obtained from magnetometer ΔH observations via a neural network-based vertical drift model [Anderson et al., 2004]. Plate 2a shows the daytime ΔH observations from the Peruvian, Philippine, and Indian sectors as a func-tion of local time and day of the year, for the entire year 2001, indicating a large day-to-day variability in DH and seasonal changes, with peaks at equinox. The wavelet am-plitude spectra of the ΔH observations at the three sectors, in the 0.2- to 1.8- and 1.5- to 33-d period ranges, are shown in Plates 2b and 2c. Similar wavelet amplitude spectra of the ΔH-inferred LLZEF at the three sectors are displayed in Plate 3. It can be seen in Plates 2 and 3 that the main spec-tral features of the ΔH data in the 0.2- to 33-d period range are in general preserved by the neural network processing procedure of inferring the LLZEF data. Referring to Plate 3b, some general characteristics are worth mentioning: (1) oscillations with periods in the 1.5- to 33-d range are present in the LLZEF spectra at all three longitudes over the entire year 2001, (2) continuous and strong bursts of oscillation ac-tivity with periods less than about 5 d are observed through-out the year at all three locations, (3) there is an enhanced oscillation activity over a large range of periods, mostly at equinoxes, and especially at the Indian sector, (4) overall, in this period range, the oscillation activity is more intense in the Indian sector whereas the least activity is recorded at the Philippine sector, and (5) referring to Plate 3a, there are stronger 8-, 12-, and 24-h periodicities at the Philippine sec-tor than at the other two sectors.

Plate 4 shows the wavelet amplitude spectrum of the IEF/15 for ωo = 6, over the entire year 2001. A scaling factor of 15 was used to obtain comparable amplitudes for the IEF and LLZEFs in the wavelet plots. The main features distin-guished in Plate 4 are (1) the presence of continuous and strong bursts of oscillation activity with periods less than about 5 d present throughout the year, (2) enhanced oscil-lation activity over a large range of periods during the No-vember and December months and at equinoxes, (3) during January, February, and summer solstice, the periodicities are less than about 5 d and have smaller amplitudes than for the rest of the year, and (4) the significant peaks in the spectrum have periods less than about 10 d.

Subsequently, we focus our discussion on periods in the 1.25- to 12-d range in the LLZEF and IEF spectra, over the 9 February to 9 June 2001 (40–160) interval characterized by relatively increased geomagnetic activity. The wavelet amplitude spectra of the LLZEF at the three longitudes and of the IEF/15 are shown in Plate 5 (left, ωo = 6) and Plate 5 (middle, ωo = 30) and their cross-spectra in Plate 5 (right,

ANGhEL Et AL. 161

Plate 1. Wavelet amplitude spectra for the (a) first, (b) second, and (c) third examples.

162 RELAtING thE INtERPLANEtARy-INduCEd ELECtRIC FIELdS WIth thE LLZEF

Plate 2. (a) ΔH observations at the three longitude sectors over the entire year of 2001 as a function of local time, (b) wavelet amplitude spectra for the 0.2- to 1.8-d period range, and (c) wavelet amplitude spectra for the 1.5- to 33-d period range (ωo = 6).

Plate 3. Wavelet amplitude spectra (ωo = 6) of the ΔH-inferred LLZEF for the 0.2- to 1.8- and 1.5- to 33-d period ranges, at the three longitude sectors.

ANGhEL Et AL. 163

ωo = 30). In Plate 5a (left), in the 70- to 120-d time interval, the IEF spectrum displays ongoing bursts of oscillation ac-tivity, which appear as an amplitude modulation of the 1.25- to 5-d periods by a 10-d period. By representing the IEF spectrum for ωo = 30 (Plate 5, middle), distinct spectral lines can be distinguished, revealing a relative constant spectral content of the IEF over this entire time interval.

Plates 5b, 5c, and 5d show the LLZEF wavelet amplitude spectra corresponding to the Peruvian, Philippine, and In-dian longitude sectors, respectively. The LLZEF spectra (ωo = 30) at the three sectors present some common features, with more pronounced similarities between the Philippine and Indian sectors, which are separated by 0300 UT hours. To relate the oscillation activity in the LLZEF and IEF, we used a cross-wavelet analysis as described in the previous section. the cross-wavelet spectra are shown in Plate 5 (right) and indicate similar periodicities that are present si-multaneously in the IEF and LLZEF spectra. In Plate 5 (top), the global cross-wavelet spectrum of the IEF/15 and of the three LLZEFs is displayed. The periodicities present in the cross-wavelet spectra are periodicities in the equatorial zonal electric fields most probably of geomagnetic origin and most probably associated with the electric field penetra-tion mechanism.

3.2 LLZEF Perturbations in the 10-min to 10-h Period Range

In this section, we analyze the oscillation activity in the LLZEF and IEF wavelet spectra in the 10-min to 10-h pe-riod range for three storm case studies and a quiet-time case. The three storm cases examined are the 17–19 April 2001, 15–18 April 2002, and 9–12 November 2004 storm events, when concurrent observations of Jicamarca ISR zonal elec-tric field and IEF data were available for at least 3 days. The quiet-time example 29 March to 2 April 2004 is only included for comparison purposes. Time and frequency do-main analyses of the three storm events have been reported previously in the literature [e.g., Kelley et al., 2007; Nicolls et al., 2007; Maruyama et al., 2007; Anghel et al. 2007], but this is the first time, to our knowledge, when wavelet analysis in the form presented here has been ever used to study the relationship between the equatorial zonal electric field and IEF data.

It is known that the two most important sources of iono-spheric electric fields are (1) electric fields induced by solar tidal and thermospheric winds through the E and F region dynamo mechanisms and (2) electric fields of solar wind/magnetospheric origin generated at high altitudes by cir-culation of the magnetospheric plasma as a result of solar wind interaction with the Earth’s magnetic field [Gonzales et al., 1979]. While the atmospheric electric field sources

are dominant at low and mid latitudes during quiet times, the magnetospheric sources become dominant at these latitudes during disturbed magnetic conditions. Thus, during storm times, the most important sources of ionospheric electric field disturbances are the prompt penetration electric fields and the disturbance dynamo electric fields, although signifi-cant contributions from other sources, such as substorms, neutral atmospheric waves, or spread F enhanced electric fields during nighttime, might also be involved. Previous studies have used different Fourier-based spectral analy-sis techniques to identify the sources of ionospheric elec-tric field variability during disturbed magnetic conditions. Earle and Kelley [1987] performed Fourier analyses in the 1- to 10-h period range to identify the sources of ionospheric electric field at equatorial latitudes and study the frequency dependence of the penetration of the high-latitude magnet-ospheric convection electric fields to low latitudes. they re-ported that for Kp larger than 3 and periods less than about 5 h, the magnetospheric electric field sources dominate the atmospheric sources, the entire system acting as a high-pass filter with a peak in the 3- to 5-h period range and a roll-off near the 10-h period [Vasyliunas, 1972]. More recently, Nicolls et al. [2007], using large data sets of concurrent IEF and ΔH-inferred LLZEF data, studied the frequency de-pendence of the penetration mechanism using a frequency response function approach, showing that their filter, with IEF as input and LLZEF as output, peaks near the 2-h pe-riod, passes periodicities in the 30-min to 5-h range, attenu-ates the longer periodicities, and drops off for periodicities shorter than about 30 min. In this section, we show that, to some extent, the wavelet analysis allows us to better identify different sources of periodicities in the equatorial zonal elec-tric fields, and that the filtering process is more complex than predicted by a time or a Fourier domain analysis, the wavelet analysis revealing in each of our case studies some peculiari-ties regarding the relationship between the IEF and LLZEF.

Plate 6a shows the wavelet results for the 18 April 2001 storm event. Plate 6a (top) displays the LLZEF (red line) and IEF (blue line), scaled by a factor of 5, as a function of local time for 17–19 April 2001 (107–109), when concurrent Jicamarca ISR zonal electric field and IEF data were avail-able. The storm event was characterized by a daily Ap of 50 and commenced and developed mostly during the nighttime at Jicamarca on 17–18 April, which explains the strong anti-correlation between the two time series. The days before and after the event were quiet with a daily Ap of 7. This storm event was analyzed in more details by Anghel et al. [2007], and here we compare the spectra of the IEF and LLZEF in a wavelet domain with increased frequency resolution. In Plate 6a (middle and bottom), the wavelet amplitude spectra (ωo = 30) of the IEF and LLZEF, respectively, for periods

164 RELAtING thE INtERPLANEtARy-INduCEd ELECtRIC FIELdS WIth thE LLZEF

ranging from 10 min to 10 h, are plotted as functions of local time and period, with the contour lines representing the 95% significance levels. the ratio between the maximum ampli-tudes in the two spectra is about 7, but for plotting purposes, we used a scaling factor of 5. A strong and highly oscillating IEF was recorded by the ACE satellite during the nighttime hours at Jicamarca. The wavelet amplitude spectrum of the IEF shows (1) strong periodicities less than about 1 h con-fined to a time interval of increased IEF, (2) significant perio-dicities in the 1.5- to 3-h range developing few hours prior to the onset of increased IEF activity, persisting over the entire time interval of increased IEF, and then fading away about 4 h after the increased activity in the IEF ceased, and (3) a less significant 4-h period of long time extent. As seen in the IEF wavelet spectrum, periodicities longer than about 1.5 h form a background spectrum on which high-frequency compo-nents, with periods less than about 1 h, superimpose during the interval of increased IEF activity. The LLZEF spectrum shows some significant frequency components with peri-ods less than about 3 h that occur simultaneously in the IEF spectrum. Significant in the LLZEF spectrum are two spec-tral lines at about 1-h period and one spectral line near the 2-h period, each having a corresponding periodicity in the IEF spectrum. Of less significance are five spectral lines around the 3-, 4-, 5-, and 8-h periods. The 2.75-h period seems to be associated with a strong similar periodicity in the IEF spectrum, while the 3.2- and 5-h periods cannot be directly linked with similar periodicities in the IEF spectrum. They extend over long time intervals and might be associated with a disturbance dynamo effect. Although very weak and nar-row-banded, the 4-h period has a corresponding periodicity in the IEF spectrum, while the 8-h period in the LLZEF spec-trum has a relatively constant amplitude over the entire time interval and might be attributed to the terdiurnal tide.

The results for the 17 April 2002 storm event are pre-sented in Plate 6b. Plate 6b (top) shows the Jicamarca zonal electric field (red line) and the IEF (blue line), scaled by a factor of 10, as a function of local time for 15–18 April 2002 (105–108). During this time interval, the daily Ap var-ied from 7, on 15 April, to 41 and 54, on 17 and 18 April. The storm commenced and developed on 17 April during the daytime hours at Jicamarca, which explains the strong correlation between the two signals. In this case study, we have a scenario when both the IEF and LLZEF evolved from quieter to more fluctuating values. the wavelet spectra for the IEF and LLZEF are shown in Plate 6b (middle and bot-tom, respectively). The ratio between the maxima of the two spectra is about 7. The IEF spectrum shows periods less than about 1 h during daytime on 17 April, when the IEF was highly fluctuating. Some significant spectral components with periods longer than 1 h can also be distinguished in

the IEF spectrum: 1.5-, 2-, 3.5- to 4-, 6-, and 9-h periods. The 1.5-h period was present in the spectrum for about 24 h, between the midnights on 17 and 18 April and the 2-h pe-riod persisted between noontime on 16 April and midnight on 18 April, both periodicities being very strong in these time intervals. The other periodicities in the IEF spectrum have smaller amplitudes and extend from early morning on 16 April to noontime on 18 April. The LLZEF spectrum is dominated by a strong 8-h period over the entire time in-terval and, on 17 April, periodicities less than 1 h correlate very well with similar periodicities in the IEF spectrum. In addition, 1.5-, 2-, and 4-h periods also appear as significant in the LLZEF spectrum having very narrow bands compared with similar periodicities in the IEF spectrum but extending over the same time intervals.

the third case study, presented in Plate 6c, refers to the 9 November 2004 storm event, which is part of a more com-plex event that commenced on 7 November 2004. Plate 6c (top) displays the LLZEF (red line) and IEF (blue line), scaled by a factor of 10, as a function of local time for 9–12 November 2001 (314–317). daily Ap values of 120 and 181 were registered on 9 and 10 November, respectively. The ratio between the maxima of the two spectra is about 7. The wavelet amplitude spectrum of the IEF is displayed in Plate 6c (middle) and shows significant periodicities less than about 2 h mostly confined to an interval of strong and highly fluctuating IEF. A strong 3-h period is also observed developing earlier on 9 November and persisting till about noon on 10 November. Other spectral lines with significant amplitudes are also observed in the IEF spectrum extending over long time intervals of more than 4–5 days, which possi-bly developed even earlier, maybe on 6 or 7 November. the LLZEF spectrum for this storm event is not as spectacular as the IEF spectrum, and there are no significant periodicities in the 3- to 8-h period band. Some significant periodicities less then about 2 h and a strong 3-h period similar with those observed in the IEF spectrum are distinguished, and two less significant and narrow-banded 4- and 6-h periods can also be noticed. The strong 9.5-h period might be associated with a disturbance dynamo effect since it does not have a direct correspondent in the IEF spectrum. During the entire inter-val, a strong 8-h period can be observed in the equatorial zonal electric field spectrum. the 8-h periodicity, most prob-ably attributed to the terdiurnal tide, is present in all three storm events and also in the quiet-time example in Plate 6d but with different amplitudes in each case, suggesting that its magnitude might be dependent on the strength of the storm.

For two of the storm events presented here, 18 April 2001 and 17 April 2002, with daily Ap values of about 50, the IEF wavelet spectra are characterized by distinct period bands

ANGhEL Et AL. 165

Plate 4. Wavelet amplitude spectrum of the IEF for the 1.5- to 33-d period range.

Plate 5. Wavelet amplitude spectra of the LLZEF at the three longitude sectors and of the IEF/15 in the 1.25- to 12-d period range and over 40- to 160-d time interval for (left) ωo = 6 and (middle) ωo= 30 and (right) the cross-wavelet spectra for ωo= 30.

166 RELAtING thE INtERPLANEtARy-INduCEd ELECtRIC FIELdS WIth thE LLZEF

Plate 6. Time series of the (red line) LLZEF and (blue line) IEF and their wavelet amplitude spectra for the 10-min to 10-h period range, for (a) 17–19 April 2001, (b) 15–18 April 2002, (c) 9–12 November 2004, and (d) 29 March to 2 April 2004.

ANGhEL Et AL. 167

less than about 4 h that emerged from a noiselike back-ground spectrum. Also, during the 9 November 2004 event, characterized by daily Ap values greater than 120 for three consecutive days, distinct period bands of large amplitudes are observed in the IEF spectrum, over the entire 10-min to 10-h period range, superimposed on a noiselike background spectrum. In each of the three storm cases, some of the sig-nificant periodicities in the IEF wavelet spectrum, especially those less than about 4 h, are also observed in the LLZEF spectrum as very narrow bands relative to their counterparts in the IEF spectrum, extending over the same time interval, but with a certain degree of attenuation [Earle and Kelley, 1987; Nicolls et al., 2007]. In the 9 November 2004 case, the LLZEF wavelet spectrum does not show any significant periods in the 4- to 8-h range, only two weak and narrow spectral bands, although there is a quite strong oscillation activity in the IEF spectrum in this period range. We also showed that periodicities longer than about 1.5 h are present in both IEF and LLZEF spectra hours in advance of a visible onset of the storm, as in the case of the 17 April 2002 event, and many hours after the main phase of the storm, as in the 18 April 2001 and 9 November 2004 events.

We conclude that in the 10-min to 10-h period range, the system, with IEF as input and LLZEF as output, behaves like a highly nonlinear time-varying filter affecting more or less each periodicity in the spectrum, with possibly more at-tenuation on periodicities longer than about 4 h, as seen in the 9 November 2004 event. Similar and simultaneous pe-riodicities present in both LLZEF and IEF wavelet spectra are most probably associated with penetration effects, while periodicities longer than about 3 h, which are present in the LLZEF spectrum but do not have a correspondent in the IEF spectrum, might be attributed to disturbance dynamo effects, although some other processes may also be accounted for.

4. CONCLuSION

In this paper, we used the continuous Morlet wavelet trans-form to relate the oscillation activity in the LLZEF and IEF spectra, in the 10-min to 10-h and 1.25- to 12-d period ranges, during time intervals of increased geomagnetic activity. The wavelet method described here can be easily tuned to dif-ferent frequency resolutions in the wavelet domain and can provide accurate amplitude values of the periodicities present in the spectrum. Three examples of simulated wavelet spectra were presented to familiarize the reader with the method.

For periods in the 10-min to 10-h range, we showed that the wavelet method represents a powerful tool to study the frequency dependence of the two specific mechanisms of equatorial electric field variability, which are dominant during disturbed conditions, namely penetration and atmo-

spheric disturbance dynamo, without the need to subtract the quiet-time components from the LLZEF and IEF data. In general, in separating out the two contributions, we associ-ate a periodicity longer than about 3 h in the LLZEF wavelet spectrum with disturbance dynamo effects when it is not si-multaneously present in the IEF wavelet spectrum, although some other processes may also be accounted for, and con-sider that a periodicity in the LLZEF wavelet spectrum is due to penetration effects most probably when it is simul-taneously present in both LLZEF and IEF wavelet spectra. Previous studies associated a periodicity in the LLZEF Fou-rier spectrum with the penetration effects if it were present in both LLZEF and IEF Fourier spectra [e.g., Earle and Kel-ley, 1987; Nicolls et al., 2007], although there is no time information about the occurrence of the periodicities in the two spectra. Also, since the wavelet method provides accu-rate amplitude values of the periodicities in the spectrum, it makes the wavelet transform suitable for studying the pen-etration efficiency and its time dependency by developing “instantaneous” frequency response functions in the same way like in the Fourier domain [Kelley et al., 2003; Huang et al., 2007; Nicolls et al., 2007].

For periods in the 1.25- to 12-d range, the wavelet and cross-wavelet analyses of the IEF and LLZEF data over the 9 February to 9 June 2001 (40–160) time interval, indicate that there are significant periodicities in the LLZEF wave-let spectrum of possible geomagnetic origin, which are well correlated with similar and simultaneous periodicities in the IEF wavelet spectrum. We consider that these periodicities might be associated with penetration effects since we do not impose a limit on the maximum period that can be passed by the system, although it is not excluded that other processes may also be involved.

It is possible that the wavelet analysis in conjunction with physics-based models and ground-based and satellite data sets will bring more insight about the sources of periodici-ties in the LLZEF and about the system that links the IEF and the equatorial zonal electric fields. here, we provide a new method of analysis and present a few case studies, but separating out different sources of ionospheric electric field variability still remains a task that requires further investigations.

Acknowledgment. Funding to carry out this study came from an NSF Space Weather grant (AtM#0207992).

REFERENCES

Abdu, M. A., T. K. Ramkumar, I. S. Batista, C. G. M. Brum, h. takahashi, B. W. Reinisch, and J. Sobral (2006), Planetary wave signatures in the equatorial atmosphere–ionosphere system

168 RELAtING thE INtERPLANEtARy-INduCEd ELECtRIC FIELdS WIth thE LLZEF

and mesosphere-E- and F-region coupling, J. Atmos. Sol. Terr. Phys., 68, 509–522.

Altadill, D., and E. Apostolov (2003), Time and scale size of plan-etary wave signatures in the ionospheric F region: Role of the ge-omagnetic activity and mesosphere/lower thermosphere winds, J. Geophys. Res., 108(A11), 1403, doi:10.1029/2003JA010015.

Anderson, d., A. Anghel, J. L. Chau, and O. Veliz (2004), day-time vertical E ´ B drift velocities inferred from ground-based magnetometer observations at low latitudes, Space Weather, 2, S11001, doi:10.1029/2004SW000095.

Anderson D., A. Anghel, J. L. Chau, and K. Yumoto (2006), Glo-bal, low-latitude, vertical E ´ B drift velocities inferred from daytime magnetometer observations, Space Weather, 4, S08003, doi:10.1029/2005SW000193.

Anghel, A., d. Anderson, N. Maruyama, J. Chau, K. yumoto, A. Bhattacharyya, and S. Alex (2007), Interplanetary electric fields and their relationship to low-latitude electric fields under dis-turbed conditions, J. Atmos. Sol. Terr. Phys., 69, 1147–1159.

Blanc, M., and A. D. Richmond (1980), The ionospheric distur-bance dynamo, J. Geophys. Res., 85(A4), 1669–1686.

Earle, G. D, and M. C. Kelley (1987), Spectral studies of the sources of ionospheric electric fields, J. Geophys. Res., 92(A1), 213–224.

Fagundes, P. R., V. G. Pillat, M. Bolzan, y. Sahai, F. Becker-Guedes, J. R. Abalde, S. L. Aranha, and J. A. Bittencourt (2005), Observations of F layer electron density profiles modulated by planetary wave type oscillations in the equatorial ionospheric anomaly region, J. Geophys. Res., 110, A12302, doi:10.1029/2005JA011115.

Fejer, B. G., and L. Scherliess (1997), Empirical models of storm- time equatorial zonal electric fields, J. Geophys. Res., 102, 24,047–24,056.

Forbes, J. F., S. E. Palo, and X. Zhang (2000), Variability of the ionosphere, J. Atmos. Sol. Terr. Phys., 62, 685–693.

Gonzales, C. A., M. C. Kelley, B. G. Fejer, J. F. Vickrey, and R. F. Woodman (1979), Equatorial electric fields during magneti-cally disturbed conditions, 2. Implications of simultaneous au-roral and equatorial measurements, J. Geophys. Res., 84(A10), 5803–5812.

huang, C.-S., J. C. Foster, and M. C. Kelley (2005), Long-duration penetration of the interplanetary electric field to the low-latitude ionosphere during the main phase of magnetic storms, J. Geo-phys. Res., 110, A11309, doi:10.1029/2005JA011202.

huang, C.-S., S. Sazykin, J. Chau, N. Maruyama, and M. C. Kelley (2007), Penetration electric fields: Efficiency and characteristic time scale, J. Atmos. Sol. Terr. Phys., 69, 1135–1146.

Kelley, M. C., J. J. Makela, J. L. Chan, and M. J. Nicolls (2003), Penetration of the solar wind electric field into the magneto-sphere/ionosphere system, Geophys. Res. Lett., 30(4), 1158, doi:10.1029/2002GL016321.

Kelley, M. C., M. J. Nicolls, d. Anderson, A. Anghel, J. Chau, R. Sekar, K. Subbarao, and A. Bhattacharyya (2007), Multi-lon-gitude case studies comparing the interplanetary and equatorial ionospheric electric fields using an empirical model, J. Atmos. Sol. Terr. Phys., 69, 1174–1181.

Kumar, P., and E. Foufoula-Georgiou (1997), Wavelet analysis for geophysical applications, Rev. Geophys., 35(4), 385–412.

Laštovička, J. (2006), Forcing of the ionosphere by waves from below, J. Atmos. Sol. Terr. Phys., 68, 479–497.

Manson, A. h., C. E. Meek, t. Chshyolkova, S. K. Avery, d. thorsen, J. W. Macdougall, W. hocking, y. Murayama, and K. Igarashi (2005), Wave activity (planetary, tidal) throughout the middle atmosphere (20–100km) over the CuJO network: Satel-lite (tOMS) and medium frequency (MF) radar observations, Ann. Geophys., 23, 305–323.

Maruyama, N., S. Sazykin, R. Spiro, d. Anderson, A. Anghel, R. Wolf, F. Toffoletto, T. J. Fuller-Rowell, M. Codrescu, and A. D. Richmond (2007), Modeling storm-time electrodynamics of the low-latitude ionosphere–thermosphere system: Can long lasting disturbance electric fields be accounted for?, J. Atmos. Sol. Terr. Phys., 69, 1182–1199.

Mendillo M., and K. Schatten (1983), Influence of solar sector bound-aries on ionospheric variability, J. Geophys. Res., 88, 9145–9153.

Nicolls, M. J., M. C. Kelley, J. Chau, O. Veliz, d. Anderson, and A. Anghel (2007), The spectral properties of low latitude day-time electric fields inferred from magnetometer observations, J. Atmos. Sol. Terr. Phys., 69, 1160–1173.

Pancheva, d. (2000), Evidence for nonlinear coupling of planetary waves and tides in the lower thermosphere over Bulgaria, J. At-mos. Sol. Terr. Phys., 62(2), 115–132.

Pancheva, d., N. Mitchell, R. Clark, J. drobjeva, and J. Lastovicka (2002), Variability in the maximum height of the ionospheric F2-layer over Millstone hill (September 1998–March 2000): Influ-ence from below and above, Ann. Geophys., 20(11), 1807–1819.

Parish, h. F., J. M. Forbes, and F. Kamalabadi (1994), Planetary wave and solar emission signatures in the equatorial electrojet, J. Geophys. Res., 99, 355–368.

Rishbeth, h., and M. Mendillo (2001), Patterns of F2-layer vari-ability, J. Atmos. Sol. Terr. Phys., 63, 1661–1680.

Roberts, R. A., and C. T. Mullis (1987), Digital Signal Processing, Addison-Wesley, Boston, Mass.

Scherliess, L., and B. G. Fejer (1997), Storm time dependence of equatorial disturbance dynamo zonal electric fields, J. Geophys. Res., 102(A11), 24,047–24,056.

Scherliess, L., and B. G. Fejer (1999), Radar and satellite glo-bal equatorial F-region vertical drift model, J. Geophys. Res., 104(A4), 6829–6842.

torrence, C., and G. P. Compo (1998), A practical guide to wavelet analysis, Bull. Am. Meteorol. Soc., 79, 61–78.

Vasyliunas, V. M. (1972), the interrelationship of magnetospheric processes, in Earth’s Magnetospheric Processes, edited by B. M. McCormac, pp. 29–38, d. Reidel, dordrecht, Netherlands.

D. Anderson and A. Anghel, Cooperative Institute for Research in Environmental Sciences, University of Colorado, Campus Box 216, Boulder, CO 80309, uSA. (adela.anghel@noaa.gov)

A. Bhattacharyya, Indian Institute of Geomagnetism, Kalamboli highway, New Panvel, Navi Mumbai 410218, India.

J. Chau, Radio Observatorio de Jicamarca, Instituto Geofisico del Peru, Jicamarca, Peru.

K. Yumoto, Space Environment Research Center, Kyushu Uni-versity, 53 6-10-1 hakozaki, higashi-ku, Fukuoka 812-8581, Japan.