SEARCH

SEARCH BY CITATION

Keywords:

  • ionospheric disturbance;
  • geomagnetic activity;
  • electrodynamics;
  • electric field penetration

Abstract

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

[1] The influence of geomagnetic activity on middle- and low-latitude thermospheric winds and ionospheric electric fields is investigated using model results from the National Center for Atmospheric Research Thermosphere-Ionosphere-Electrodynamics General Circulation Model. Model runs are made for different levels of geomagnetic activity. Model results show that the equatorward ionospheric currents produced by disturbance winds develop positive charge accumulation at low latitudes that maximizes in the premidnight sector. The local time of maximum electric potential perturbation depends significantly on universal time so that the local time of reversal of the equatorial zonal perturbation electric field varies with longitude by 2 to 3 hours, depending on the intensity of geomagnetic activity. The westward perturbation electric field in the postsunset period indicates that stronger geomagnetic activity will produce a lower driven height of the evening F region. After geomagnetic activity ceases, model results show that the zonal disturbance winds can last for many days in the postrecovery period, while the meridional disturbance winds decay more rapidly. The long-lasting zonal winds, through the Pedersen currents they drive, help maintain meridional disturbance potential drops that decay much more slowly than the zonal disturbance potential drops after the activity ceases.

1. Introduction

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

[2] Model studies of ionospheric electric field disturbances at middle and low latitudes associated with geomagnetic activity have been presented by Fuller-Rowell et al. [2002] using the Coupled Thermosphere Ionosphere Plasmasphere Electrodynamic Model (CTIPE) and by Richmond et al. [2003] using the Magnetosphere-Thermosphere-Ionosphere-Electrodynamics General Circulation Model (MTIEGM) of Peymirat et al. [1988]. Richmond et al. [2003] pointed out that three effects can be of comparable importance on the equatorial electric field, namely, global winds driven by solar heating, direct penetration of polar cap electric fields to the equator, and disturbance winds driven by high-latitude Joule heating and plasma convection. Furthermore, they found that the equatorial disturbance electric field produced by disturbance winds tends to be opposite to that produced by the other effects. In the poststorm period they also found that the low-latitude disturbance electric fields result more from the disturbance winds than from direct penetration of polar cap electric fields. The idea of a disturbance wind dynamo and its influence on middle and low latitudes came from the studies of Blanc and Richmond [1980] and Spiro et al. [1988], who examined the dynamo effects associated with midlatitude winds driven by high-latitude Joule heating and with fossil winds accelerated by strong ion convection in the auroral regions, respectively. As described by Richmond et al. [2003], a few hours are required to set up the disturbance winds, after which they can persist for several hours. Therefore we may expect that the low-latitude electric field disturbance associated with the disturbance winds should be more persistent than that associated with direct penetration from the polar cap, especially for the poststorm period. In this period, Richmond et al. [2003] found that the disturbance winds tend to charge the low-latitude ionosphere positively during the night, especially around midnight.

[3] In the present study, we use the Thermosphere-Ionosphere-Electrodynamics General Circulation Model to investigate the effect of disturbance winds on the midlatitude and low-latitude electric fields. During geomagnetic activity, the influence of disturbance winds can easily reach to the midlatitude and low-latitude regions [Fuller-Rowell et al., 2002]. In order to simulate and study the ionospheric effects associated with the disturbance winds, we perform model runs with a low Kp index, representing quiet conditions, and moderate and high Kp indices, representing moderate and strong storm conditions. We run each simulation to a steady state, and then the difference of these two simulations is used to study the electrodynamic influence of the disturbance winds on midlatitude and low-latitude electrodynamics. Because the boundary of the polar cap is fixed for this model, the effects of polar cap expansion and contraction [Fejer et al., 1990] at the beginning and the end of a storm cannot be simulated. However, as pointed out by Richmond et al. [2003], the influence of the disturbance winds on the low-latitude electric fields depends very little on polar cap contraction in the poststorm period. Thus the model runs made in the present study should be sufficient to study the residual effect associated with disturbance winds after the geomagnetic storm and to explore the characteristics of poststorm relaxation.

[4] In the following, the model description and input parameters will be presented in section 2. The results of model runs will be shown in section 3, followed by discussion and conclusions in the last section.

2. Model Description and Input Parameters

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

[5] The model used in this study is the NCAR TIEGCM, updated by Richmond et al. [1992] from the TIGCM [Roble and Ridley, 1987; Roble et al., 1988] by including wind dynamo effects such that the electric fields can be self-consistently solved. Densities of various species are calculated, including N2, O2, O, N(4S), N(4D), NO, N2+, O2+, O+, NO+, and N+, along with the ion, electron, and neutral temperatures and the neutral winds. The diurnal (1,1) and semidiurnal (2,2) through (2,6) tidal modes are included as perturbations to the lower boundary, as described by Fesen et al. [1991a). In the present study, the diurnal mode (1,1) is specified with an amplitude of 297 geopotential meters and a phase of 21 hours, while the semidiurnal (2,2), (2,3), (2,4), (2,5), and (2,6) modes have amplitudes of 437.2, 10, 280, 10, and 20 m and phases of 0.425, 5.8, 7.6, 8.3, and 8.4 hours, respectively. The spatial grid is 5° × 5°, ranging from −87.5° to 87.5° in latitude and from −180° to +180° in longitude, with 29 levels in the vertical from 97 km to about 600 km, depending on solar activity. (For more details, see Fesen et al. [1993].)

[6] Other parameters are required as inputs, including the solar flux (F10.7), the cross polar cap potential, and the hemispheric power. In the present study, we set F10.7 = 150 × 10−22 W m−2 Hz−1 for equinox conditions. The empirical high-latitude ion convection model of Heelis et al. [1982] is also used to account for the effects of ion drag and Joule heating. The empirical model developed by Foster et al. [1986], which describes the relations between precipitation activity and polar cap potential, is used to specify inputs for three different model runs. For cases 1, 2, and 3, the hemispheric power and cross polar cap potential are set to 3, 60, 120 GW, and to 20, 66, 78 kV, respectively, corresponding to Kp levels of 1, 4, and 6. Thus these three model runs represent quiet, moderate, and strong geomagnetic activity, respectively.

[7] For a given case, we hold the inputs constant and run the model for 5 days to reach essentially a steady state. We then take the results of case 3 minus case 1 and case 2 minus case 1 to represent effects produced by different levels of geomagnetic activity. When calculating electric fields for our figures, we remove effects of direct-penetration fields by instantaneously setting the polar cap potential to zero, while retaining the dynamo effects of the thermospheric winds. Hereafter, for convenience, we refer to case 3 minus case1 as case A, and to case 2 minus case 1 as case B.

3. Model Results

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

[8] Figure 1 shows the distribution of the disturbance winds for case A (strong geomagnetic activity) at 0000 universal time (UT) for daytime at 140 km altitude (Figure 1a) and for nighttime at 250 km altitude (Figure 1b), where only the region between geographic latitudes 62.5°S and 62.5°N is shown. The winds at altitudes of 140 km and 250 km are representative of disturbance winds that affect the daytime or nighttime ionospheric dynamo, respectively. At high latitudes, of course, one expects the wind to be organized more with respect with magnetic latitude than geographic latitude so that the local time variations of the high-latitude wind at other universal times may be rather different than those in Figure 1. At middle and low latitudes, however, the main features of the wind at other universal times are found to be quite similar to those in Figure 1. For the daytime sector at 140 km (Figure 1a), the disturbance winds are mostly westward, except for regions close to dawn at high latitudes. For the nighttime sector at 250 km (Figure 1b), disturbance winds are also mainly westward, except for regions at higher latitudes (northern hemisphere) after midnight where they are more equatorward, consistent with the WINDII observations of Emmert et al. [2004]. The magnitudes of the disturbance winds in Figure 1b are comparable with the observations of Emmert et al. [2004] and with the wind measurements during storm periods made at Millstone Hill [Fejer et al., 2002]. As discussed by Blanc and Richmond [1980], the westward component results from the Coriolis effect acting over time on the equatorward wind, more strongly at the higher latitudes. As expected, the strong geomagnetic activity case A produces larger disturbance winds than the moderate activity case B (not shown), especially in the nighttime period.

image

Figure 1. Distribution of the disturbance winds for case A (strong geomagnetic activity) at 0000 UT. (a) Daytime at 140 km altitude (maximum speed = 216 m/s); (b) nighttime at 250 km altitude (maximum speed = 235 m/s).

Download figure to PowerPoint

[9] As discussed by Blanc and Richmond [1980], Fuller-Rowell et al. [2002], and Richmond et al. [2003], the westward and equatorward winds create equatorward Pedersen and Hall currents, respectively. Positive charge accumulation at low latitudes results from the convergence of these electric currents. Figure 2 shows the current density distribution, integrated from 90 km to 500 km altitude, produced by the disturbance winds for case A. It is very clear that the integrated electric currents are mainly equatorward. Therefore we may expect that positive charge will accumulate at low latitudes. Figure 3 shows the corresponding electric potential distribution, averaged from 90 km to 500 km altitude, at 0000 UT for case A. (At midlatitudes, the height averaging has negligible effect, as the potential is nearly constant in altitude. At equatorial latitudes the averaged potential is similar to the potential at 300 km.) The geomagnetic equator is shown by a dashed line. As expected, high electric potentials appear in the equatorial regions. Note that the symmetrical potential distribution with respect to the geomagnetic equator in low and middle latitudes is due to the assumption of constant potential along geomagnetic field lines in the TIEGCM. From this figure we find that the maximum potential is located around midnight, as simulated and pointed out by Richmond et al. [2003]. This result indicates that perturbed eastward and westward electric fields will exist after and before midnight, respectively.

image

Figure 2. Current density distribution, integrated from 90 km to 500 km altitude, produced by the disturbance winds for case A.

Download figure to PowerPoint

image

Figure 3. Electric potential distribution, averaged from 90 km to 500 km, at 0000 UT for case A.

Download figure to PowerPoint

[10] The top part of Figure 4 shows the local time variations of the zonal electric field for case 1 (solid line), case 2 (dotted line), and case 3 (dashed line) at a specific grid point, 12.5°S, 75°W, which is very close to the Jicamarca radar station. The lower part shows the perturbation zonal electric fields for case A (dashed line) and case B (dotted line), corresponding to strong and moderate geomagnetic activity, respectively. Figure 4 can be used to describe the effect of disturbance winds produced by storms at one specific location for different local times. In contrast to the quiet nighttime behavior, the disturbed zonal electric field for both cases 2 and 3 (Figure 4, top) becomes eastward in the postmidnight period. As shown in the bottom part of Figure 4, the eastward perturbation in the zonal electric field begins at about 2130 local time (LT) for both case A (strong activity) and case B (moderate activity). It is also interesting to note that the disturbance field is westward during the prereversal time (1800–2000 LT), which implies that the height to which the equatorial ionosphere is driven due to the prereversal enhancement [Farley et al., 1986; Kelley, 1989] will be reduced by the effect of steady-state geomagnetic activity, more so for strong activity than for moderate activity. Although the local time for reversal of the electric field disturbance from westward to eastward in Figure 4, around 2130, at first sight may seem inconsistent with the electric potential distribution shown in Figure 3, remember that the result shown in Figure 3 is only representative of one specific universal time (0000 UT).

image

Figure 4. (top) Local time variations of the zonal electric field for case 1 (solid line), case 2 (dotted line), and case 3 (dashed line) at a specific grid point, 12.5°S, 75°W. (bottom) Electric field differences between cases 3 and 1 (dashed line; case A) and between cases 2 and 1 (dotted line; case B).

Download figure to PowerPoint

[11] The results shown above reveal that the local time of maximum electric potential is probably different for different universal times. As seen in Figure 5, the maximum electric potential for case A at 0600 UT is located at about 2100 LT, which is different from that shown in Figure 3. We may also expect that the variations of electric field disturbance with local time should be different for different locations. To further investigate this behavior, we present in Figure 6 the results for another location, 7.5°N and 60°E, which is also close to the magnetic equator. The top part shows the same general pattern as shown in Figure 4 for the quiet-time case 1 (solid line), but obvious differences from Figure 4 exist at night for both case 2 (dotted line) and case 3 (dashed line). In Figure 6, the zonal disturbance electric field in the midnight-to-dawn period is still eastward, though with smaller magnitude than in Figure 4. Other differences can be seen between the bottom parts of Figures 4 and 6 for the two locations. First, the local time for electric field reversal from westward to eastward is closer to midnight in Figure 6 than in Figure 4. Second, the daytime westward disturbance electric field is larger in Figure 6 than in Figure 4. These comparisons indicate that the influence of geomagnetic activity on the low-latitude and equatorial ionospheric regions should be different for different locations along the geomagnetic equator. Other variations of equatorial electrodynamics also appear to be longitudinally dependent. For example, observational results obtained by the satellite CHAMP have revealed a longitudinal dependence of the average noontime equatorial electrojet (EEJ) peak current density [Lühr et al., 2004]. In that paper, many factors which could affect the intensity of the EEJ were suggested, for example, the angle between the ambient magnetic field and the EEJ current, the angle between the EEJ and geographic east, and the differences in great circle distance of the EEJ to the corrected geomagnetic north and south poles. The last one was suggested by Lühr et al. [2004] as the preferred factor. It describes the penetration of the cross polar cap electric potential to the geomagnetic equator, which, in turn, affects the intensity of EEJ current. However, the effect of direct electric field penetration [Peymirat et al., 2000] is removed in our simulations, as explained earlier.

image

Figure 5. Electric potential distribution, averaged from 90 km to 500 km, at 0600 UT for case A.

Download figure to PowerPoint

image

Figure 6. (top) Local time variations of the zonal electric field for case 1 (solid line), case 2 (dotted line), and case 3 (dashed line) at a specific grid point, 7.5°N, 60°E. (bottom) Electric field differences between cases 3 and 1 (dashed line; case A) and between cases 2 and 1 (dotted line; case B).

Download figure to PowerPoint

[12] In order to further understand the variations, Figure 7 shows the local time of reversal of the disturbance electric field as a function of geographic longitude along the magnetic equator (dashed line for case A; dotted line for case B). As shown in this figure, the reversal times can vary with longitude by up to 3 hours and 2.5 hours for case A and case B, respectively. It should be noted that the nonsmooth curves in this figure result from the spatial and temporal discretization used in the model runs.

image

Figure 7. Local time of reversal of the disturbance electric field as a function of geographic longitude along the magnetic equator (dashed line for case A; dotted line for case B).

Download figure to PowerPoint

[13] An important process we will now examine is how the winds and electric fields recover after a geomagnetic storm. As mentioned above, the perturbed electric fields result from the disturbance dynamo driven by the disturbance winds. We may expect that the magnitudes of disturbance winds decrease after the end of geomagnetic activity, which, in turn, would cause the perturbed electric fields to decrease. To examine the poststorm effects, we linearly relax the geomagnetic activity over 2 hours (0000 UT to 0200 UT), starting from case 3 and ending at the quiet-time case-1 activity level, i.e., Kp = 1, in order to examine the poststorm recovery behavior.

[14] After the geomagnetic activity ends, the disturbance winds decay. In Figure 8, the time variations of the zonal winds at 300 km, averaged over 180° longitude on either the dayside or nightside, are shown at 24 hour intervals at different geographic latitudes (as shown at top of each part). All the data points are for 0000 UT, and point “0” represents the beginning of the ramp-down of the geomagnetic activity. As can be seen, the initial decay rate is faster at high latitudes than at low latitudes. In addition, the initial decay rate is faster for the nighttime region than for the daytime region. Figure 9 shows the corresponding variations of the meridional winds. The initial decay rates of the meridional winds are also more significant at higher latitudes during the nighttime. However, in contrast to the zonal component, the decay rates are not so different for different latitudes in the daytime. Furthermore, the decay rates are not monotonic at low latitudes in the nighttime part. In Figures 10 and 11, we show the variations of disturbance winds, again averaged over 180° of longitude on the dayside and nightside, at 120 km altitude. Comparing with Figures 8 and 9, we see that during the active period (time 0), the wind magnitudes are smaller at 120 km than at 300 km. This results from the smaller heating per unit mass at lower altitudes. From Figure 11, we also see that the meridional component at 120 km decays very fast as compared with that at 300 km. In fact, the time variations of disturbance winds at heights above 150 km are more or less similar to Figures 8 and 9. However, at heights below 150 km, more complicated variations appear, though these are not shown in this paper.

image

Figure 8. Time variations of the zonal winds at 300 km, averaged over 180° longitude on the (left) dayside or (right) nightside, with 24 hour intervals at different geographic latitudes (as shown at top of each part). Positive number in the y-axis indicates eastward wind.

Download figure to PowerPoint

image

Figure 9. Time variations of the meridional winds at 300 km, averaged over 180° longitude on the (left) dayside or (right) nightside, with 24 hour intervals at different geographic latitudes (as shown at top of each part). Positive number in the y-axis indicates northward wind.

Download figure to PowerPoint

image

Figure 10. Time variations of the zonal winds at 120 km, averaged over 180° longitude on the (left) dayside or (right) nightside, with 24 hour intervals at different geographic latitudes (as shown at top of each part). Positive number in the y-axis indicates eastward wind.

Download figure to PowerPoint

image

Figure 11. Time variations of the meridional winds at 120 km, averaged over 180° longitude on the (left) dayside or (right) nightside, with 24 hour intervals at different geographic latitudes (as shown at top of each part). Positive number in the y-axis indicates northward wind.

Download figure to PowerPoint

[15] Figure 12 shows the height profiles of the poststorm disturbance winds (differences from case 1) for specific locations (noon, left side; midnight, right side) at 0000 UT. The solid lines denote disturbance winds just before magnetic activity decreased, followed by various dashed lines for each of the first 5 days after storm and then followed by solid lines that denote each of the next 5 days. During the active period, we find that at middle and low latitudes, the maximum disturbance winds built up by geomagnetic activity are mainly located at altitudes of about 200 km at midnight and about 150 km at noon. The fact that the winds increase with altitude below the maximum is readily understandable in terms of the increasing heating per unit mass with altitude, which drives increasing winds. The cause of the amplitude decrease at higher altitudes is less obvious. Viscosity should tend to eliminate height variations of the wind at high altitudes, but it appears that ion drag in the F region ionosphere (above 150 km at day and above 200 km at night) may retard the disturbance winds there. It is worth noting that the magnitudes of the disturbance winds are larger at midnight than at noon, which can help explain why the disturbance electric potential maximizes at night, as seen in Figure 3. Another interesting feature shown in Figure 12 is that the maximum wind at noon is at 300 km at high latitudes, but is at about 145 km at lower latitudes. To understand it, a detailed forcing analysis will be necessary, and such an analysis would go beyond the scope of the present study.

image

Figure 12. Time variations of height profiles of the poststorm disturbance winds (differences from case 1) at specific locations (noon, left; midnight, right) at 0000 UT, with 24 hour intervals at different geographic latitudes (as shown at top of each part). See text for discussion of the various lines in this figure.

Download figure to PowerPoint

[16] After the end of geomagnetic activity, we see that the initial decay rate is larger at midnight than at noon, most obviously so at higher latitudes, as was already shown in Figure 8. Just 1 day after the end of magnetic activity, we may also see that differences in disturbance wind between noon and midnight are small as compared with those during the storm. As the disturbance wind becomes more nearly zonally symmetric, it might be expected that the electric potential differences between daytime and nighttime will be reduced accordingly.

[17] Because the changes of the disturbance winds depend on many factors, including changes of pressure gradients at higher latitudes, viscosity, ion drag, and the redistribution of global mass density and temperature with time, the detailed variations of the disturbance winds cannot be easily described in any simple way. On the other hand, the electric potential tends to have a simpler behavior, since the potential at any given location responds to the combined dynamo effects of winds at different locations and thus tends to average out the effects of smaller-scale features of the winds. In Figures 13 and 14, the poststorm variations of zonal and meridional potential drops are shown daily for 0000 UT. As in Figures 3 and 5, the electric potential is averaged from 90 km to 500 km altitude. Three local time periods are considered in Figure 13: the postmidnight period (top, 0000–0600 LT), the daytime period (middle, 0600–1800 LT), and the period after sunset (bottom, 1800–2100 LT). In Figures 13 and 14, the quiet-time condition from case 1 is shown as a dashed line in each part. In Figure 13 we see that about 1 day after the end of geomagnetic activity the zonal potential drops have relaxed to the quiet-time condition. Figure 14 shows the potential differences between the geomagnetic equator and 50° geomagnetic latitude in the northern hemisphere at midnight (top, 0000 LT), noon (middle, 1200 LT), and sunset (bottom, 1800 LT). In contrast to the zonal potential drops, we see that the meridional potential drops relax more slowly. In order to see more details of this asymmetrical relaxation in electric potential, we present in Figure 15 the distribution of the difference potential (the actual potential minus the potential for case 1) 1 day after the end of geomagnetic activity. As compared with Figure 3, we see that the region of high electric potential extends along the geomagnetic equator, associated with the faster decay of potential drops in the zonal direction than those in the meridional direction seen in Figures 13 and 14.

image

Figure 13. Time variations of zonal potential drops, averaged from 90 km to 500 km altitude, for 0000 UT with 24 hour intervals. Three local time periods are shown: the postmidnight period (top, 0000–0600 LT), the daytime period (middle, 0600–1800 LT), and the period after sunset (bottom, 1800–2100 LT).

Download figure to PowerPoint

image

Figure 14. Time variations of the potential differences, averaged from 90 km to 500 km altitude, between the geomagnetic equator and 50° geomagnetic latitude with 24 hour intervals in the northern hemisphere at midnight (top, 0000 LT), noon (middle, 1200 LT), and sunset (bottom, 1800 LT).

Download figure to PowerPoint

image

Figure 15. Distribution of the difference potential (the actual potential minus the potential for case 1) 1 day after the end of geomagnetic activity, averaged between 90 km and 500 km altitude.

Download figure to PowerPoint

4. Discussion and Conclusions

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

[18] The influence of geomagnetic activity on the low-latitude and equatorial regions partly comes from the disturbance winds set up by Joule heating and ion drag forcing in the auroral and polar cap regions. To investigate the effects associated with disturbance winds on the electric field at low latitudes, we have made TIEGCM runs to simulate situations representing strong and moderate geomagnetic activities. We have run the disturbed simulations to a steady state, which strictly speaking would be appropriate only for a very extended period of activity but which in practice should produce results fairly representative of typical storm conditions as well. In the present study, directly penetrating electric fields have been removed from the total fields so that our results represent only the wind-dynamo effects.

[19] Some significant features are revealed by the TIEGCM runs: (1) the local time for maximum disturbance electric potential associated with the disturbance winds mainly appears in the premidnight period and depends on universal time; (2) this longitudinal variation increases with geomagnetic activity; (3) the wind-generated electric fields associated with geomagnetic activity should depress the driven height in the equatorial evening ionospheric F region, more so for strong activity than for moderate activity; (4) the disturbance electric potential relaxes considerably more slowly in the zonal than in the meridional direction in the poststorm period.

[20] As explained in the previous section, equatorward electric current driven by disturbance winds converges at the magnetic equator, creating poleward electric fields with magnitudes mainly determined by the direction and intensity of the disturbance winds. This phenomenon can be readily understood by considering the Pedersen currents driven by the winds represented in Figures 1 and 2. The equatorward electric current, and therefore the charge accumulation, is affected by the configuration of geomagnetic field lines, resulting in the differences between Figure 3 (0000 UT) and Figure 5 (0600 UT) and the longitudinal variations seen in Figure 7.

[21] An important influence of the electrodynamics on the equatorial ionosphere is the height to which the F region ionosphere is driven after sunset. During this period the perturbation electric field is essentially westward, as shown in Figures 4 and 6. The reduction of the postsunset upward drift in the active period with respect to quiet conditions due to disturbance-dynamo effects implies that the occurrence of equatorial spread F (ESF) should be reduced, due to the greater ion-neutral collision frequency at lower altitudes, causing a reduced growth rate for the Rayleigh-Taylor instability.

[22] As shown by Fuller-Rowell et al. [2002], the global neutral wind disturbance can be quickly built up by the influence of geomagnetic activity. It is also interesting to know the relaxation characteristics of the winds and electric fields after the geomagnetic activity ceases. As revealed by the TIEGCM runs, the poststorm variations of disturbance winds cannot be easily represented by simple relations, since many important factors are involved, as mentioned in the previous section. However, some general behaviors can be described as follows. (1) For the zonal and meridional wind components above 150 km, the initial decay rates are faster at higher latitudes than at lower latitudes and are faster at night than at day. (2) In contrast to the zonal component, the poststorm variations of meridional winds are not monotonic, for example, as shown in Figure 9. (3) For heights below 150 km, the decay of the meridional component at all latitudes is much faster than that of the zonal component, as shown in Figures 10 and 11. (4) A sharp decrease occurs within 1 day after the end of geomagnetic activity, after which the wind decays gradually for heights above 150 km.

[23] The poststorm electric potential varies in a simpler manner than the wind, due to the fact that it depends on integrated wind effects. As shown in Figures 13 and 14, the meridional potential drops decrease monotonically in both the daytime and nighttime periods, while the zonal potential drops decay rapidly to the quiet time condition in about 1 day. Comparison of Figures 3 and 15 reveals that 1 day after the end of geomagnetic activity, the electric potential has become more uniform with longitude. The longer persistence of the meridional than the zonal disturbance electric field is associated with the longer persistence of the zonal than the meridional wind, since it is primarily the wind-driven Pedersen current that maintains the disturbance potential in the poststorm period.

[24] The theoretical effects associated with geomagnetic activity on the electric field at low latitudes have been investigated by using TIEGCM runs for equinox conditions. The results indicate that neutral wind disturbances strongly affect the electrodynamics of the ionosphere. In order to see the corresponding variations for other seasons, TIEGCM runs will be made for the solstices in the near future.

Acknowledgments

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

[25] The authors thank Ben Foster of NCAR/HAO for his assistance in getting the TIEGCM to run at the ISS/NCU (Institute of Space Science, National Central University), Taiwan. This work was supported by National Science Council of Taiwan through grants NSC93-2111-M-008-026-AP3, NSC93-2111-M-008-021-AP5, NCS93-NSPO(B)-RS3-FA07-01-A, and by the NASA Sun-Earth Connection Theory Program.

[26] Shadia Rifai Habbal thanks Hermann Lühr and another referee for their assistance in evaluating this paper.

References

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