SEARCH

SEARCH BY CITATION

Keywords:

  • disturbance electric fields;
  • ionospheric storm effects;
  • ionospheric plasma drifts

Abstract

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results
  5. 3. Discussion
  6. 4. Summary and Conclusions
  7. Acknowledgments
  8. References

[1] Low-latitude ionospheric electric fields and currents are often strongly disturbed during periods of enhanced geomagnetic activity. These perturbations can last for several hours after geomagnetic quieting. We use incoherent scatter radar measurements from Jicamarca and Arecibo during 19–21 October 1998 to study, for the first time, the low-latitude disturbance electric fields during the recovery phase of a large magnetic storm. On 19 October the Jicamarca data showed initially large and short-lived (time scale of about 10–20 min) upward and westward drift perturbations in the early afternoon sector, due to the penetration of strong magnetospheric electric fields probably driven by an increase in the solar wind dynamic pressure. Following the decrease of auroral activity, very strong afternoon and nighttime zonal and vertical disturbance dynamo drifts were observed over Jicamarca but, surprisingly, only small and sporadic perturbation drifts were present over Arecibo. The latitudinal variation of the daytime zonal disturbance dynamo drifts during this day is in good agreement with that previously observed for thermospheric disturbance winds. The disturbance dynamo drifts also indicate significantly different local, storm time, and possibly latitudinal dependences for their vertical/perpendicular and zonal drift components, suggesting large variations in both the amplitude and direction of the disturbance dynamo electric field vector. In the next day, during moderately disturbed conditions, the daytime perturbation drifts were small, but following transient increases in geomagnetic activity, large upward/perpendicular drifts were observed near dusk over Jicamarca and Arecibo, where simultaneous westward perturbation drifts were also seen. These perturbations are consistent with the occurrence of strong prompt penetration electric fields reaching the magnetic equator. The measured prompt penetration drift patterns are generally in good agreement with predictions from global convection models. Later at night, under moderately disturbed conditions, relatively large zonal and meridional disturbance dynamo electric fields were observed by the two radars. Our results illustrate the large variability of the low-latitude perturbation electric fields relative to their climatological values after large storms, probably due to the importance of additional disturbance processes. They also indicate that a much deeper understanding of solar-wind/magnetospheric/ionospheric processes is required for accurate predictions of these electric fields.

1. Introduction

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results
  5. 3. Discussion
  6. 4. Summary and Conclusions
  7. Acknowledgments
  8. References

[2] Large plasma drift and current perturbations in the low-latitude ionosphere during periods of strong geomagnetic activity have been the subject of numerous studies for several decades [e.g., Sastri, 1988; Abdu et al., 1995; Buonsanto, 1999; Richmond et al., 2003]. These electric field (electrodynamic plasma drift) disturbances occur globally, cover a broad range of time scales, and often significantly affect the plasma density distribution and the occurrence of plasma instabilities at low latitudes [e.g., Fejer, 1986; Abdu et al., 1991, 1997; Sobral et al., 1997; Sastri et al., 2000; Basu et al., 2001]. Their general morphology and climatological relationship to solar wind, magnetospheric, and high-latitude ionospheric processes are now fairly well established, but the sources of their large variability are still not well understood [e.g., Fejer, 2002].

[3] Sharp and short-lived (time scales smaller than about an hour) plasma drift perturbations are due mostly to the prompt penetration of solar-wind/magnetospheric electric fields to low latitudes [e.g., Kelley et al., 1979; Fejer, 1986; Fejer et al., 1990a; Kikuchi et al., 1996; Sastri et al., 1997, 2002]. These transient disturbances generally occur during periods of large and rapid changes in magnetospheric convection, when the inner edge of the plasma sheet, and the associated region-2 Birkeland currents, are temporarily configured to shield out a weaker (undershielding) or stronger (overshielding) cross-tail electric field [e.g., Wolf et al., 1982; Wolf, 1983; Senior and Blanc, 1984; Spiro et al., 1988; Peymirat et al., 2000; Kobea et al., 2000]. The average characteristics of equatorial prompt penetration F region vertical plasma drifts (zonal electric fields) are in good agreement with the predictions from global convection models [e.g., Fejer and Scherliess, 1997], which suggests that they do not depend on the details of magnetospheric physics. Prompt penetration equatorial zonal plasma drifts (vertical electric fields) have not been reported [Fejer, 1997].

[4] Preliminary impulses of geomagnetic sudden commencements and sudden changes in the solar wind dynamic pressure also produce transient ionospheric zonal electric field and current perturbations, but with generally smaller amplitudes than those resulting from large changes in magnetospheric convection. The local time variation of equatorial ionospheric zonal electric fields resulting from preliminary impulses of geomagnetic sudden commencements was derived by Tsunomura and Araki [1984]. This pattern closely resembles that associated with sudden convection increases [Fejer et al., 1990a]. Sibeck et al. [1998] showed that the large majority of north/south perturbations with amplitudes between 10 and 40 nT seen in dayside equatorial ground magnetograms can be explained in terms of either substorm onset or changes in the solar wind dynamic pressure and that many southward IMF turnings produce no observable effects. However, large fluctuations in the solar wind electric field are often highly correlated with corresponding ionospheric electric field and current disturbances from high to equatorial latitudes [e.g., Gonzales et al., 1979; Kelley et al., 2003].

[5] More slowly varying (time scales from a few to several hours) low-latitude disturbance electric fields and plasma drifts during and up to about a day or two after geomagnetically disturbed conditions are generally due to ionospheric disturbance dynamo electric fields resulting from enhanced energy deposition (mostly through Joule heating) into the auroral ionosphere [Blanc and Richmond, 1980; Fejer et al., 1983; Sastri, 1988; Mazaudier and Venkateswaran, 1990; Fejer, 1997]. These electric field perturbations are westward during the day and eastward at night and have largest amplitudes near dusk. Scherliess and Fejer [1997] showed that the climatology of the equatorial vertical plasma drifts measured over Jicamarca is in excellent agreement with the predictions of the Blanc-Richmond ionospheric disturbance dynamo model. The equatorial average zonal disturbance drifts are also consistent with results from this model [Fejer, 1997], but their relationship to the vertical disturbance dynamo drifts has not been established.

[6] The October 1998 storm has been the subject of several experimental and numerical studies, which so far have focused on the development of the storm time ring current [e.g., McAdams et al., 2001; Liemohn et al., 2001a, 2001b], on the timing of geomagnetic disturbances [Anderson et al., 2000], and on the occurrence of rapid subauroral ion drifts (SAIDs) [Anderson et al., 2001]. In the following sections, we study the low latitude ionospheric electric fields during the recovery phase of this storm. Our data consist mostly of F region electrodynamic (E × B) plasma drifts measured by the Arecibo (18°N, 67°W, dip latitude 30°N) and Jicamarca (11.9°S, 76.8°W, dip latitude 1°N) incoherent scatter radars. We will present the first detailed measurements of both zonal and meridional prompt penetration and disturbance dynamo drifts over Jicamarca and Arecibo. Our results will show that the zonal and meridional disturbance electric fields can have significantly different storm-time dependence, strong latitudinal dependence, and very large departures from their climatological patterns. We also show that zonal and meridional prompt penetration electric fields resulting from an increase in the solar wind dynamic pressure are consistent with perturbation drifts associated with convection enhancements. These results illustrate the large and complex variability of the low-latitude disturbance electric fields resulting from major geomagnetic storms.

2. Results

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results
  5. 3. Discussion
  6. 4. Summary and Conclusions
  7. Acknowledgments
  8. References

[7] The October 1998 geomagnetic storm started at about 0500 UT when a magnetic cloud with a southward (Bz < 0) interplanetary magnetic field (IMF) originating from a coronal mass ejection, encountered the magnetopause. Detailed descriptions of the solar wind, magnetospheric, and high latitude ionospheric conditions during this storm were given by Anderson et al. [2000, 2001], McAdams et al. [2001], and Liemohn et al. [2001a]. Figure 1 presents the AE indices derived from 84 magnetometer stations and the Dst indices for 18–21 October 1998. In this two-phase decay storm, the Dst exhibited a brief steep initial phase at about 0300 UT on 19 October and a minimum of −139 nT at about 1500 UT, which occurred shortly after the period of largest AE values. Then, the Dst index recovered and the AE indices showed decreasing geomagnetic activity up to about 1500 UT on 20 October, when there was a moderate increase in geomagnetic activity. The Kp indices (not shown) had a maximum of 7- during 0300–0600 UT, and a secondary maximum of 6+ between 1200 and 1500 UT on 19 October. Later, geomagnetic activity decreased up to about 1800 UT on 20 October, when it increased for about 6 hours, reaching Kp = 5+, before decreasing to moderately disturbed levels. In this work, we will study the response of the low latitude ionospheric electric fields in the American sector from 1640 UT on 19 October to 1200 UT on 21 October.

image

Figure 1. AE and Dst indices during 18–21 October 1998.

Download figure to PowerPoint

[8] Figure 2 shows the AE indices and the F region electrodynamic plasma drifts measured over Arecibo and Jicamarca during the first part of the October 1998 coordinated incoherent scatter radar campaign. At F region heights, an eastward (southward) electric field of 1 mV/m corresponds to an upward/perpendicular (eastward) drift velocity of 27 m/s over Arecibo and 40 m/s over Jicamarca. Solar local time (LT) lags UT by about 4.5 hours at Arecibo and 5 hours at Jicamarca. The smooth plasma drift plots denote the average quiet time (Kp < 3) equinoctial patterns for the average decimetric solar flux index (F10.7) of 125 of this period.

image

Figure 2. AE indices and Arecibo and Jicamarca plasma drift measurements during 19–20 October 1998. The smooth curves denote the quiet time drift patterns. Noon corresponds to about 1630 and 1700 UT over Arecibo and Jicamarca, respectively.

Download figure to PowerPoint

[9] The Jicamarca drifts were obtained using the experimental procedure described by Woodman [1970, 1972], and the data acquisition and signal processing techniques developed by Kudeki et al. [1999]. The Jicamarca data analysis technique now provides unprecedented accuracy, time resolution, and altitudinal coverage for electrodynamic drift measurements. We averaged these equatorial drift measurements over the altitudinal range of 235–570 km, where they do not change much with altitude [e.g., Woodman, 1970]. In this case, the Jicamarca vertical drifts have an accuracy better than 1 m/s for our integration time of 5 min. We further smoothed the less accurate zonal drifts using a 3-point running average, resulting in an accuracy of 5–10 m/s and an effective integration time of 10 min.

[10] The Arecibo measurements shown in Figure 2 were made using the standard azimuthal scanning at a fixed elevation of about 25° [e.g., Behnke and Harper, 1973] with a cycle time of about 16 min and processed assuming constant and spatially uniform drifts over the whole scan. We averaged these measurements over the altitudinal range from 256 to 404 km, resulting in a typical accuracy of 5–10 m/s during the day, and 10–20 m/s at night, when the signal-to-noise ratio is much smaller.

[11] Following the peak of geomagnetic activity, the daytime vertical/perpendicular drifts presented in Figure 2 were initially smaller than the quiet time values, underwent shorted-lived increases at about 1740 UT, and then closely matched the average values up to about 2000 UT. In the evening and nighttime periods, large vertical/perpendicular perturbation drifts were observed over Arecibo and Jicamarca under very quiet geomagnetic conditions. In this case, the Arecibo drifts showed downward excursions in the afternoon (2000–2300 UT) and late night-early morning (0700–0800 and 1000–1200 UT) periods, and the Jicamarca vertical perturbation drifts were downward to about 0100 UT (2000 LT), and then upward to about 0700 UT. The Arecibo and Jicamarca vertical perturbation drifts turned downward after 1000 UT.

[12] The Jicamarca zonal drifts presented in Figure 2 show large eastward (∼50 m/s) perturbations during the day, and westward perturbations at night, with largest magnitudes (∼75 m/s) occurring at about 0900 UT (0400 LT). The daytime eastward and the evening downward disturbance dynamo drifts were the largest ever measured at Jicamarca in these local time sectors. Notice that the equatorial vertical and zonal disturbance drifts were often out of phase on 19–20 October, indicating different storm-time dependences for the zonal and meridional perturbation electric fields. The large disturbance drifts during the geomagnetically quiet period following 2100 UT were due to disturbance dynamo electric fields. In contrast, the Arecibo daytime and nighttime zonal drifts generally followed their average quiet time pattern, indicating a strong latitudinal dependence of the disturbance drifts. As mentioned earlier, the Jicamarca drifts have smaller errors than the Arecibo drifts.

[13] Figure 3 shows the temporal variations of IMF Bz and solar wind dynamic pressure measured by the WIND satellite, and the equatorial magnetic field horizontal component and plasma drifts measured at Jicamarca between 1700 and 1830 UT. During this period, the WIND satellite was located at about (93, 33, 6) RE, and measured a solar wind speed of about 380 km/s. Therefore we shifted these satellite measurements by 23 min to account for the propagation time to the front of the magnetopause. Anderson et al. [2000] showed that during this period the IMF data from WIND were in very good agreement with those from IMP 8, which was located a few minutes upstream of the magnetopause. The solar wind perturbations take a few minutes to propagate through the magnetosheath.

image

Figure 3. IMF, north-south magnetic field component and solar wind dynamic pressure measured by the WIND satellite and shifted by 23 min and Jicamarca horizontal magnetic field and plasma drifts observations near 1730 UT.

Download figure to PowerPoint

[14] The radar measurements presented in Figure 3 show upward and westward perturbation drifts of about 25 m/s (i.e., eastward and upward electric fields of ∼0.5 mV/m) nearly at the same time (1730 UT) as the first sudden increase in the solar wind dynamic pressure. Notice the excellent agreement between the Jicamarca magnetic field and vertical drift (i.e., zonal electric field) data. The Arecibo measurements shown in Figure 2 also suggest the simultaneous occurrence of eastward and northward electric fields perturbations at about 1740 UT, but these data have much longer integration times and larger uncertainties. As we will discuss later, the vertical and zonal plasma drift signatures are consistent with the occurrence of a short-lived prompt penetration electric field. We believe that these prompt penetration electric fields were triggered by the first increase in the solar wind dynamic pressure. The IMF data show a northward turning of Bz from −12 to 4 nT also at about 1730, which triggered a weak global response, probably a substorm, at about 1742 UT [Anderson et al., 2000]. The large upward drift perturbation in Figure 3 cannot be explained by this northward IMF turning and ensuing substorm because northward IMF turnings routinely drive daytime equatorial downward F region plasma drifts and westward equatorial electrojet currents [e.g., Fejer, 1986; Kikuchi et al., 2000]. On the contrary, we believe that the northward turning contributed to the rapid decrease of the prompt penetration drifts generated by the first increase in the solar wind dynamic pressure at 1730 UT. The solar wind dynamic pressure increases at 1850 and 1905 UT led to smaller equatorial upward plasma drift and magnetic field perturbations. The absence of observable corresponding westward velocity perturbations following these latter increases in the dynamic pressure is probably due to the larger uncertainties and longer integration times of the Jicamarca zonal drift measurements. The two brief, small-amplitude upward velocity perturbations over Jicamarca between 2300 and 2400 UT shown in Figure 2 were also correlated with sudden increases in the solar wind dynamic pressure.

[15] The AE and plasma drift data during the second day of the October 1998 storm period are plotted in Figure 4. In the morning sector, during a geomagnetically quiet period, the upward/perpendicular drifts were consistently smaller than the average quiet time values, possibly due to disturbance dynamo electric fields. In the afternoon, the meridional drifts over Arecibo showed downward perturbations between 1930 and 2330 UT (1500 and 1900 LT) that were quite similar to those measured in the previous day (see Figure 2). It is possible that these perturbations were due to disturbance electric fields generated either by the enhanced high-latitude currents a few hours earlier or by localized disturbance winds. On the other hand, since the Arecibo quiet times drifts often show significant departures from their average values even during quiet times [Fejer, 1993], we cannot rule out quiet time variability as the source of these downward perpendicular perturbation drifts. The Jicamarca afternoon vertical drifts are basically similar to their average quiet time values, except for brief upward drift enhancements associated with the increase in AE near 1900 UT (1400 LT). Notice that here again the corresponding Jicamarca zonal drifts show small short-lived westward perturbations, which is suggestive of prompt penetration electric field effects. The Arecibo and Jicamarca daytime zonal drifts did not show other clear indications of disturbance electric field effects.

image

Figure 4. Same as Figure 2 for 20–21 October 1998.

Download figure to PowerPoint

[16] The low-latitude disturbance drifts are usually largest and most important at night, when they can strongly affect the low-latitude plasma density distribution and generation of ionospheric plasma waves. Figure 4 shows strong increases followed by decreases in the upward/perpendicular perturbation drifts over Arecibo and Jicamarca. These features are associated with the increase and decrease of AE near dusk (between about 2330 and 0100 UT). Figure 1 indicates that during this period Dst was nearly constant at about −70 nT. At later times, the Arecibo data show upward/perpendicular perturbations starting at about 0700 UT and downward perturbation drifts after 0900 UT, although in this late-night local time sector the errors on these Arecibo measurements are fairly large, due to low signal-to-noise ratios. The Jicamarca nighttime data show downward perturbation drifts at 0300 UT and upward perturbations at 0700 and 0900 UT.

[17] The Arecibo and Jicamarca daytime zonal drifts on 20 October were generally close to their average quiet time values. The Arecibo zonal perturbation drifts near dusk were first westward and then eastward. We will show later that the signatures of the Arecibo and Jicamarca disturbance drifts in this local time sector are consistent with prompt penetration electric fields resulting from an increase and then a decrease in the cross polar cap potential. On the other hand, the possible importance of other disturbance processes cannot be ruled out. In the post midnight sector, the Arecibo data show large westward disturbance drifts from 0600 to 1100 UT, with a sharp peak between 0800 and 0900 UT. Jicamarca zonal drift measurements were sparse in the early night sector, due to the occurrence of strong equatorial spread F echoes either driven or enhanced by the large upward drift velocity near dusk. Later at night, the equatorial zonal drifts were initially comparable to their quiet time average values and then had large westward perturbations between 0700 and 1200 UT, when geomagnetic activity was moderate. In this local time sector, the upward/perpendicular perturbation drifts are consistent with the occurrence of prompt penetration eastward electric fields. The smoothly varying Arecibo and Jicamarca westward perturbation drifts are consistent with disturbance dynamo electric fields, whereas the brief westward (and upward) perturbation drifts over Arecibo at about 0830 UT are indicative of undershielding prompt penetration electric field effects, as discussed later.

[18] The strong equatorial disturbance dynamo and prompt penetration drifts observed during 19 and 20 October, respectively, produced large changes in the height of the nighttime F layer over Jicamarca. Figure 5 presents the temporal variation of hF (bottomside of the F layer) during these nights. On 19 October the large downward perturbation drifts have precluded the development of the prereversal enhancement, which normally lifts the height of the F layer near dusk. The disturbance dynamo drifts reversed to upward between about 0100 and 0700 UT (2000 and 0200 local time), as shown in Figure 2, largely canceling the quiet time downward nighttime drifts. The large rise in hF during this period, when the radar measurements still showed downward plasma motions, was most probably due to strong recombination effects in the bottomside of the F layer. This highlights the inherent limitation in the derivation of the equatorial nighttime vertical drifts from the time rate of change of hF when these heights are below 300 km. In the evening of 20 October, hF initially showed a strong uplift as a result of the combined effects of the prereversal enhancement and of the eastward prompt penetration electric field. Strong spread F echoes developed at about 2000 local time and rapidly covered a large range of altitudes, which briefly precluded measurements of plasma drifts using the incoherent scatter technique. At later times, the large downward plasma motions pushed hF to low altitudes and prevented the further occurrence of equatorial spread F.

image

Figure 5. Temporal variations of the bottomside of the F layer near dusk over Jicamarca during 19 and 20 October 1998. The smooth curve denotes the quite time pattern.

Download figure to PowerPoint

3. Discussion

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results
  5. 3. Discussion
  6. 4. Summary and Conclusions
  7. Acknowledgments
  8. References

[19] In this section we will discuss in more detail the characteristics of the disturbance plasma drifts presented in the previous section and then compare them with results from previous experimental studies, as well as with predictions from the Blanc-Richmond disturbance dynamo theory and from global convection models.

3.1. Disturbance Dynamo Plasma Drifts

[20] Our observations during the recovery phase of the October 1998 storm showed highly unusual and complex plasma drift perturbation patterns. The slowly varying daytime and evening vertical and zonal perturbation drifts during magnetic quieting period of 19 October have signatures consistent with the effects of ionospheric disturbance dynamo drifts [Blanc and Richmond, 1980]. The temporal evolution of disturbance dynamo vertical drifts over Jicamarca was modeled by Scherliess and Fejer [1997] using AE as a measure of geomagnetic activity. This seasonally averaged climatological model predicts evening (1700–0100 UT) downward velocity perturbations of only about 5 m/s for 19 October, but Fejer [2002] pointed out that the downward disturbance dynamo drifts near dusk are largest during solar maximum equinoctial conditions. These effects would increase the predicted downward perturbations near dusk on 19 October to about 15 m/s, which is still much smaller than the observed values (about 40 m/s). These evening disturbance dynamo drifts are in fact the largest ever observed by the Jicamarca radar in this local time sector. The upward perturbation drifts over Jicamarca between 0200 and 0700 UT on 20 October occur earlier and are also much larger than predicted by the Scherliess-Fejer model. On the other hand, the upward disturbance dynamo drifts during 0600–1000 UT on 21 October, under moderately disturbed conditions, are in much better agreement with the empirical disturbance dynamo model. During this latter period there were also upward perpendicular prompt penetration drifts over both Jicamarca and Arecibo, which will be discussed later. These results illustrate the large variability of the disturbance dynamo electric fields following large storms, particularly in the evening sector [Scherliess and Fejer, 1997; Fejer, 2002]. Clearly, other processes not accounted for by AE also play important roles in the efficiency of the equatorial ionospheric disturbance dynamo mechanism.

[21] The zonal drifts also showed large departures from their quiet time values during the recovery phase of this storm. The daytime Jicamarca eastward perturbation drifts during October 19th were also the largest ever measured by the radar. Large westward perturbation drifts were observed at Jicamarca between about 1900 and 0800 LT on 19–20 October and during 0100–0700 LT on 21 October, which is consistent with the average zonal disturbance drift pattern [Fejer, 1997] and with results from the Blanc-Richmond theory. The data shown in Figure 2 indicate that the vertical and zonal disturbance drifts had noticeably different local and/or storm time variations.

[22] The zonal disturbance drift pattern showed large latitudinal effects in the first day of observations, when large eastward and westward perturbations were present over Jicamarca but not over Arecibo. The latitudinal dependence of the daytime zonal disturbance drifts is consistent with the predictions of the Blanc-Richmond theory and with climatological disturbance drift patterns [Fejer, 1997]. Recent studies of thermospheric winds measured by the Wind Imaging Interferometer (WINDII) on board the Upper Atmosphere Research Satellite (UARS) showed that the daytime zonal disturbance winds change from eastward close to the magnetic equator to westward at midlatitudes and are close to zero near the latitude (about 30° magnetic) of Arecibo [Emmert et al., 2002]. Figure 6 shows the latitudinal variation of the average thermospheric disturbance wind vectors derived from WINDII data. The low-latitude perturbation winds have largest amplitudes above 200 km. These results indicate a similar daytime zonal disturbance dynamo drifts and the upper thermospheric perturbation winds. This is surprising since the F region winds are inefficient drivers of daytime plasma drifts, due to the large daytime E layer conductivity.

image

Figure 6. Latitudinal and local time dependence of daytime thermospheric disturbance winds measured by UARS (adapted from Emmert et al. [2002]).

Download figure to PowerPoint

[23] Figures 2 and 4 showed very different relationships between the Arecibo and Jicamarca zonal disturbance dynamo drifts in the postmidnight sector. In the first night, under geomagnetically quiet conditions, there were large westward perturbation drifts over Jicamarca but not over Arecibo. In contrast, in the following post midnight period and under moderately disturbed conditions, the Arecibo and Jicamarca disturbance dynamo westward perturbation drifts were comparable, which is in better agreement with climatological results (as we will discuss later, the sharp westward perturbation over Arecibo near 0800 UT on 21 October was due to prompt penetration electric fields). This result again indicates the large variability of the disturbance drifts, particularly during and after large geomagnetic storms when additional parameters not accounted for by the AE index play important roles in the local time, storm time, and latitudinal dependence of the disturbance drifts.

3.2. Prompt Penetration Plasma Drifts

[24] Although disturbance dynamo drifts occurred most of the time during the recovery phase of the October 1998 storm period, there were also a number of daytime and nighttime short-lived perturbation drifts, which can easily be identified as due to prompt penetration electric fields. Figure 7 shows the low latitude prompt penetration electric field patterns for an increase in the cross polar potential by 45 kV, derived from a Rice Convection Model (RCM) run described by Spiro et al. [1988] and Fejer et al. [1990b]. The equatorward/perpendicular electric fields were obtained by dividing the equatorward (horizontal) electric fields, given by Fejer et al. [1990b], by the sine of the dip angle. Other convection models [e.g., Senior and Blanc, 1984; Peymirat et al., 2000; Tsunomura, 1999], and more recent RCM runs [e.g., Sazykin, 2000] give very similar prompt penetration electric field patterns. Figure 7 indicates that following an increase in the polar cap potential, the prompt penetration electric fields are eastward and poleward during the day, westward and poleward at night, and equatorward near dawn. These model perturbation electric fields decay on time scales of ∼15 min. The disturbance electric field patterns associated with a sudden decrease in the cross polar cap potential have comparable amplitudes and opposite polarities [Fejer and Scherliess, 1997]. The results from this convection model are in good agreement with climatological disturbance electric fields from equatorial to midlatitudes [Fejer and Scherliess, 1997, 1998].

image

Figure 7. Temporal variations of RCM prompt penetration electric field patterns at two storm times for invariant latitudes of 15° and 30°, following an increase in the polar cap potential drop by 45 kV (adapted from Fejer et al. [1990b]).

Download figure to PowerPoint

[25] Large equatorial upward and westward penetration drifts and magnetic field perturbations occurred near 1740 UT (1240 LT) over Jicamarca on 19 October, as shown in Figures 2 and 3. The magnetic field perturbation (∼150 nT) was much larger than the typical values (∼10–40 nT) reported by Sibeck et al. [1998]. These disturbance drifts and currents are consistent with the occurrence of eastward and downward prompt penetration electric fields resulting from a sudden increase in the polar cap potential, as seen in RCM results of Figure 7. The RCM also predicts comparable amplitudes for the eastward and downward prompt penetration electric fields near noon, which is in good agreement with the Jicamarca data (over this site, eastward and upward electric fields of 1 mV/m correspond to upward and westward penetration drifts of about 40 m/s). As discussed later, Hall conduction effects induced by the eastward penetration electric field near 1730 UT could have contributed to the corresponding westward perturbation drift. Figure 2 indicates that, as expected, upward and westward prompt penetration drifts also possibly occurred over Arecibo (where 1 mV/m corresponds to about 27 m/s) during this period, but these features are less clear due to the larger errors and longer integration time of the Arecibo data.

[26] The occurrence of low-latitude prompt penetration vertical/perpendicular plasma drifts has been known for over 4 decades [e.g., Gonzales et al., 1979; Fejer, 1986; Sastri et al., 1997], but until now there has been no evidence for equatorial prompt penetration zonal drifts. This lack of evidence can now be readily explained. First, these zonal disturbance drifts are not easily observable because they are short-lived and have much smaller magnitudes than the quiet time equatorial values. Second and most importantly, only after the introduction of the most recent data processing techniques [Kudeki et al., 1999] have the Jicamarca zonal drift measurements attained the necessary accuracy for the detection of these small perturbations.

[27] While the equatorial vertical and zonal drift perturbations at about 1730 UT on 19 October have clear characteristics of prompt penetration electric fields during undershielding conditions, the identification of the corresponding magnetosphere-ionosphere current system is considerably more difficult and is outside the scope of the present study. It is interesting to note, however, that Clauer et al. [2001] suggested that the complex electrodynamic phenomena resulting from a sudden compression of the magnetosphere, with the simultaneous northward turning of the IMF, has intrinsic features connected with the formation of a new field aligned current system which, over a characteristic time, evolves to the steady current system characteristic of northward IMF.

[28] Prompt penetration events with time scales of about 1 hour were also observed in the afternoon and nighttime periods during 20–21 October. The Jicamarca data on the afternoon of 20 October show small upward and westward perturbations at about 1800–1900 UT associated with increased auroral activity (see Figure 4), as indicated by AE. This small prompt penetration event cannot be inferred from the Arecibo data.

[29] Strong (∼40 m/s) upward drift perturbations, again associated with the increased geomagnetic activity, were observed over Jicamarca and Arecibo near dusk on 20 October, as shown in Figure 4. Over Arecibo, westward perturbation drifts also occurred at the same time as the upward/perpendicular drifts. The Arecibo and Jicamarca eastward disturbance electric fields were about 1.5 and 1 mV/m, respectively. The ratio of these zonal electric fields is only slightly smaller than L3/2, which is the value obtained by assuming a uniform dawn-to-dusk electric field in the equatorial plane for centered-dipole equipotential geomagnetic field lines. The Jicamarca data presented in Figure 4 also shows large downward disturbance drifts at 0100–0300 UT, with amplitudes comparable to the earlier upward perturbation drifts. Upward and downward disturbance drifts with comparable amplitudes seem to be a typical feature during storm recovery phase, when the ring current is symmetrical. These evening and early night perturbation drifts are consistent with the results shown in Figure 7, although the RCM significantly underestimates their amplitudes and lifetimes, particularly for the zonal component. More recent RCM runs [e.g., Sazykin, 2000] and other new convection models [e.g., Tsunomura, 1999; Peymirat et al., 2000], using more realistic low-latitude evening conductivities, predict larger equatorial zonal prompt penetration electric fields near dusk, in better agreement with our data.

[30] The electrodynamics of the evening quiet-time equatorial ionosphere was modeled in detail by Haerendel et al. [1992]. They have shown that the altitudinal profile of the F region zonal drift velocity is determined by the combined effects of the neutral wind dynamo, Hall conduction, and equatorial vertical current divergence. The Hall conduction term is proportional to the ratio of the field line integrated Hall to Pedersen conductivity times the zonal electric field, with an eastward electric field driving a westward plasma drift. This process seems fundamentally important in explaining the westward drifts in the bottomside of the equatorial F layer near the evening prereversal enhancement of the upward drift. Based on these results, we expect that during disturbed conditions the equatorial zonal plasma drifts can undergo sudden changes not only due to the prompt penetration vertical electric fields but also as a result of Hall conduction effects induced by zonal prompt penetration electric fields. This latter process should be most important near dusk, when the low-latitude meridional/perpendicular prompt penetration electric fields are expected to be very small (see Figure 7). Although these evening disturbance Hall conduction effects should not be easily observed with equatorial incoherent scatter observations (due to the rapid occurrence of spread F echoes following enhanced evening eastward electric fields), other Jicamarca drift observations (not shown here) seem to support their importance. It is possible that disturbance Hall conduction effects could also contribute to the zonal disturbance drifts at other local times (e.g., near 1730 UT on 19 October), but their full importance can be assessed only after they are properly modeled in other local time sectors.

[31] Very large amplitude upward/perpendicular and westward perturbations, typical of overshielding conditions, occurred over Arecibo again in the post midnight sector (at about 0800 UT) on 21 October. These disturbance drifts are consistent with the upward perturbations observed at Jicamarca, where noticeable zonal prompt penetration drifts are not expected near 0300 LT (see Figure 7). Even if we take into account the relatively large errors in the Arecibo drift measurements in the late night sector, we cannot fully explain the very large amplitudes of these prompt perturbation drifts by the small corresponding changes in AE. These results illustrate that, although AE provides good qualitative and climatological quantitative information on the disturbance electric fields, it often leads to inaccurate estimates of plasma drifts, particularly during and shortly after strongly active periods.

4. Summary and Conclusions

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results
  5. 3. Discussion
  6. 4. Summary and Conclusions
  7. Acknowledgments
  8. References

[32] We have presented the first detailed study of low-latitude disturbance plasma drifts during the recovery phase of a major geomagnetic storm. Our data show large latitudinal variability in the disturbance dynamo drifts and also in the relationship between their vertical/perpendicular and zonal components at low latitudes. These results indicate very different storm time and local time dependences of the zonal and meridional disturbance dynamo electric fields. The daytime zonal disturbance dynamo drifts are essentially zero over Arecibo, which is in excellent agreement with the latitudinal variation of F region disturbance winds derived from UARS observations. The nighttime measurements suggest longer lifetimes for the equatorial disturbance dynamo drifts than for the Arecibo drifts. Although the characteristics of the measured disturbance dynamo drifts are generally consistent with the climatological experimental and theoretical patterns, their amplitudes at the equator are initially much larger than the expected values. This suggests that the disturbance dynamo electric fields are sometimes significantly underestimated by the time history of AE, and that other solar wind, magnetospheric, and/or high-latitude ionospheric parameters can sometimes strongly affect the disturbance dynamo process.

[33] We have also observed daytime and nighttime short-lived vertical/perpendicular and zonal plasma drift perturbations that are clearly due to prompt penetration electric fields. Large upward and westward daytime prompt penetration drifts were observed on 19 October, most likely due to an increase in solar wind pressure, which occurred at the same time as a northward turning of the IMF. Noticeable prompt penetration drifts were also seen over Arecibo and Jicamarca in the following evening and nighttime periods. The characteristics of these perturbation drifts are fully consistent with the corresponding undershielding and overshielding conditions suggested by changes in AE, and with predictions from global convection models, but the measured prompt penetration drifts were sometimes much larger than expected. These results illustrate the large variability of the low-latitude perturbation drifts possibly due to the importance of additional disturbance processes and the challenging task we face in the development of realistic ionospheric predictions during and shortly after highly geomagnetically active times.

Acknowledgments

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results
  5. 3. Discussion
  6. 4. Summary and Conclusions
  7. Acknowledgments
  8. References

[34] We thank one of the reviewers for pointing out the possible importance of Hall conduction effects. We also thank E. Kudeki and the Jicamarca and Arecibo staff for the radar measurements. Data from the WIND satellite were obtained from the NASA/GSFC CDAWeb site. The AE values were derived by G. Lu and obtained from the GEM Web site and the Dst indices were acquired on the web from the World Data Center, Kyoto, Japan. The Jicamarca Radio Observatory is operated by the Geophysical Institute of Peru, Ministry of Education, with support from the National Science Foundation as contracted through Cornell University. The Arecibo Observatory is operated by Cornell University under cooperative agreement with the National Science Foundation. This work was supported by the Aeronomy Program, Division of Atmospheric Sciences of the National Science Foundation through grant ATM-0004380.

[35] Arthur Richmond thanks Mangalathayil Ali Abdu and Timothy J. Fuller-Rowell for their assistance in evaluating this paper.

References

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results
  5. 3. Discussion
  6. 4. Summary and Conclusions
  7. Acknowledgments
  8. References