Notice: Wiley Online Library will be unavailable on Saturday 30th July 2016 from 08:00-11:00 BST / 03:00-06:00 EST / 15:00-18:00 SGT for essential maintenance. Apologies for the inconvenience.
 The effect of a disturbance dynamo during geomagnetic activity on the equatorial ionospheric electric fields is investigated, using model results from the NACR/TIEGCM (National Center for Atmospheric Research Thermosphere Ionosphere Electrodynamics General Circulation Model). Model runs are made for different time-lengths of geomagnetic activity, for different seasons, and for different solar activities to investigate how and where the maximum electric potential forms. Model results show that the maximum electric potential is located at around 300 km altitude and at local time after the pre-reversal enhancement at equinox for high solar activity. For the case at solstice, without pre-reversal enhancement, the location moves to around 110 km altitude and to the local time close to midnight. Giving various parameters in the model runs to simulate different background conditions, many important output quantities are used to study the distributions of perturbed electric potential at the geomagnetic equator. Model investigation indicates that normal quiet time electrodynamics, at different seasons with different solar activities, significantly affect the distribution of perturbed current density associated with geomagnetic activity. Furthermore, model results also reveal that significant perturbed zonal electric fields tend to build up six hours after the onset of geomagnetic activity, except at regions close to sunset and sunrise, and the perturbed vertical electric fields increase with the time length of geomagnetic activity.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
 In the storm period, the ionospheric electric fields are mainly composed of three components that result from solar heating, direct penetration of polar cap electric fields, and disturbance winds. As shown by Richmond et al. , these influences on equatorial electric fields can be of comparable importance, and the magnetic local time (MLT) variations of the first two components tend to have a similar pattern in a steady state, while the third tends to have opposite variations. During geomagnetic activity, these components may simultaneously exist in the ionosphere, although the global disturbance dynamo requires a few hours to set up. The studies of Fuller-Rowell et al.  show that an equatorial response would appear within two hours of the storm onset. In the recovery phase, the perturbed electric fields associated with the disturbance wind effect tend to be more important than the direct penetration electric fields [Richmond et al., 2003].
 After the main phase of geomagnetic activity, the perturbed electric fields come primarily from the contributions of the disturbance dynamo. Equatorward currents produced by southward and westward disturbance winds build up a high electric potential at lower latitudes [Blanc and Richmond, 1980], maximizing at pre-midnight at equinox [Huang et al., 2005]. Some of the reasons for this effect can be attributed to larger F region disturbance winds in the nighttime than in the daytime [Huang et al., 2005]. The distribution pattern of electric potential associated with the disturbance dynamo indicates that eastward and westward perturbed electric fields will be formed in the post-midnight and pre-midnight, respectively, and the reversal time from westward to eastward is strongly dependent on the longitudes [Huang et al., 2005].
 It is clear that, whether or not in the storm period, the height variation of the vertical electric field in the ionosphere is generally more pronounced than that of the zonal electric field. In the storm period, the height variation of the vertical electric field should be strongly dependent with the accumulation height of the perturbed current density at the geomagnetic equator. To understand it we need to calculate the distribution of the perturbed electric potential associated with the disturbance dynamo. In this paper, we use the NACR/TIEGCM (Thermosphere Ionosphere Electrodynamics General Circulation Model) to investigate the effects of the disturbance winds on the equatorial electric fields. We make model runs with a low Kp index, representing the quiet condition, and a high Kp index, representing a strong storm condition; the difference of these two simulations is used to study the influence of the disturbance winds on the equatorial electrodynamics.
 In addition, another question for our study is whether different background conditions of the ionosphere respond to different electrodynamics at the geomagnetic equator during geomagnetic activity, especially for the distribution of perturbed electric fields. We also perform different model runs at different seasons, each with different solar conditions, and from the results, we find that a very pronounced difference exists between conditions at solar-maximum equinox and at solar-minimum solstices.
 In the following, the model description and parameter inputs will be briefly presented in section 2, and the results of model runs will be shown in section 3, followed by the discussion and conclusions in the last section.
2. Model Description and Parameter Inputs
 The model used in this paper is the NCAR TIEGCM, advanced by Richmond et al.  from the TIGCM [Roble and Ridley, 1987; Roble et al., 1988] by including wind dynamo effects in order to self-consistently solve the electric fields. Densities of neutral and charged 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 neutral winds. The diurnal mode (1, 1) and semi-diurnal modes (2, 2) through (2, 6) are used as perturbations to the lower boundary, as described by Fesen et al. [1991a]. In the present study, two sets of tidal modes are used to represent two different background conditions. In the first set, representing condition at March equinox, the diurnal mode (1, 1) is specified with an amplitude of 200 geopotential meters and a phase of 21 h, while the semi-diurnal modes (2, 2), (2, 3), (2, 4), (2, 5) and (2, 6) have amplitudes of 437.2, 10, 280, 10, and 20 m, and phases of 1.2, 5.8, 7.6, 8.3, and 8.4 h, respectively. In the second set, representing condition at June solstices, the diurnal mode (1, 1) is specified with an amplitude of 300 geopotential meters and a phase of 16.1 h, and the semi-diurnal modes (2, 2), (2, 3), (2, 4), (2, 5) and (2, 6) have amplitudes of 380, 83.5, 121, 87.7, and 56.1 m and phases of 11.6, 11.7, 16.9, 9.6, and 13.7 h, respectively. The time resolution is three minutes and 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 the solar activity [see, Fesen et al., 1993, for more details].
 Other important parameters are required as input, including the cross polar cap potential drop, the hemispheric power, and the solar flux (F10.7). In the present study, we set F10.7 = 200 × 10−22 Wm−2 Hz−1, 150 × 10−22 Wm−2 Hz−1, and 85 × 10−22 Wm−2 Hz−1 to represent the high, mediate and low solar conditions, respectively. The empirical ion-convection model of Heelis et al.  is used in TIEGCM to account for the effects of ion drag and Joule heating, and the empirical model developed by Foster et al. , which describes the relations between precipitation activity and polar cap potential, is also used to specify inputs for the required model runs. In this study, the hemispheric power and cross polar cap potential drop are set to 3, 120 GW, and set to 20, 78 kV, respectively, corresponding to Kp levels of 1 and 6. Thus these two sets of parameters represent quiet and strong geomagnetic activity, respectively. In the previous study [Huang et al., 2005], we take the difference between model runs with different parameter inputs to represent effects produced by the geomagnetic activity. In the present study, we only consider the effects associated with the disturbance dynamo; therefore, in calculating the electric fields we remove the effects of direct-penetration fields by setting the polar-cap potential to zero, while retaining the dynamo effects of the thermospheric winds.
3. Model Results
 In order to build up the reliable background quiet time conditions, we perform many model runs at March equinox and June solstice, each with high (F10.7 = 200), mediate (F10.7 = 150), and low (F10.7 = 85) solar conditions. In these quiet time cases, the cross polar-cap potential drop and the hemispheric power are set as 20 kV and 3 GW, respectively, and the results of the vertical and ion drifts, averaged from 350 to 450 Km, at location (12.5°S, 75°W) very close to the Jicamarca Radio Observatory are shown in Figure 1. The solid, dot, and dashed lines are used to represent the low, mediate, and high solar activities, respectively. The modeled seasonal and solar cycle patterns are in good agreement with the Jicamarca observations and satellite data [Fejer et al., 1991; Scherliess and Fejer, 1999], except for the vertical ion drifts, at June solstice with low solar activity, in the early evening. Similar results have also been published by Fesen et al. , in which the PRE (pre-reversal enhancement) has been successfully simulated for the first time by the TIEGCM to exhibit the strong seasonal and solar cycle dependences.
 In this study we perform model runs with different time lengths, as specified in Figure 2. In each model run, the cross polar-cap potential and the hemispheric power are linearly increased from 20 kV and 3 GW (quiet time condition) to 78 kV and 120 GW (storm time condition), respectively, in one hour. The purpose of this specification is that it tends to reach the same geographic condition (at UT = 00), where the effects of geomagnetic activity on the equatorial electric fields in response to different time lengths can easily be seen. We perform model runs with energy inputs of 3 h, 6 h, 12 h, 18 h, and 24 h, respectively, and subtract from the quiet time result. Before going to the detail model results, we also check the magnitudes of the disturbance winds, which are comparable with the observations made by Emmert et al.  and with the wind measurements in the storm periods made at Millstone Hill [Fejer et al., 2002]. The first case is made at March equinox with mediate solar flux. Figure 3 shows the variations of perturbed zonal (left panels) and vertical (right panels) electric fields with geographic longitudes and local time (along the geomagnetic equator) at 00 UT for heights of 100 km to 500 km (bottom panel to top panel), for 3 h (solid line), 6 h (dot line), 12 h (dash line), 18 h (dot-dash line), and 24 h (3-dot-dash line) after the onset of geomagnetic activity, respectively. As mentioned in the previous section, the direct-penetration field has been removed; therefore only electric fields associated with the neutral disturbance dynamo are present. In the left panels, we see that the westward electric field in the daytime and eastward electric field in the nighttime, beginning at about LT 23 (−15° in geographic longitude), is consistent with previous model studies [Richmond et al., 2003; Huang et al., 2005], and the magnitudes are almost unchanged after 6 h, except at locations close to sunrise and sunset. The maximum electric field, over 1 mv/m, appears at around sunrise after 18 h and almost maintains the same magnitude after 24 h, although this is not shown in the figure. One more important feature of the perturbed zonal electric fields is that the largest magnitude occurs at around LT 5, which is in good agreement with the empirical model published by Scherliess and Fejer . In contrast with the zonal electric fields, the vertical electric fields show a clear change with heights, as expected by the Faraday's law [Kelley, 1989]. At height of 200 km, the upward electric field mainly exists in the nighttime, beginning at about LT 23 and maximizing around sunrise with a magnitude of 2 mv/m, and the downward electric field mainly occurs in the daytime, maximizing around sunset with a magnitude of 1.5 mv/m. At 300 km and higher altitudes, the upward electric field is very different from that at a height of 200 km, and is especially clear in the nighttime period. The greatest difference exists at a location after sunset, downward at 200 km and upward at 400 km. This difference implies that a relatively high electric potential exists at a location between the altitudes of 200 km and 400 km. Also shown in Figure 3, the magnitudes of the vertical electric field in most of the daytime are very small as compared with that in the nighttime.
 The second case shown in Figure 4 is made at June solstice with mediate solar flux. The magnitude of the perturbed zonal electric field is a little smaller than that in the first case, but with a larger perturbed vertical electric field. In addition, the local time variation of the zonal electric field in the post-midnight is in excellent agreement with the empirical model [Scherliess and Fejer, 1997]. Another difference between Figure 3 and Figure 4 is the height variations of the perturbed vertical electric field. As shown in Figure 4, a relatively high electric potential exists at a location after sunset and between the altitudes of 100 km and 300 km.
 It is well known that the perturbed current density produced by the westward and southward neutral disturbances through the disturbance dynamo [Blanc and Richmond, 1980; Richmond et al., 2003] tend to charge the low-latitude ionosphere positively during the storm-time period [Huang et al., 2005]. From Figure 3 and Figure 4, we may find that high electric potential may exist in the nighttime period, and the maximum electric potential would locate between 100 and 400 km heights and occur in the pre-midnight. The importance of the location of maximum electric potential is obvious, because it determines the directions of the perturbed electric fields associated with the disturbance dynamo in the storm-time period, even in the recovery phase.
Figure 5 shows the distribution of the field-line integrated wind dynamo divided by field-line integrated Pedersen conductivity produced by the neutral disturbances for the first case (i.e., at March equinox with mediate solar flux), six hours after the onset of geomagnetic activity and along the geomagnetic equator, ranging from 100 to 500 km in height. Downward electric currents prevail except at regions close to sunrise and sunset. High electric potential essentially results from the convergence of electric current. From this figure, we may expect that higher electric potential would be built up in the nighttime than in the daytime; the disturbance dynamo currents are almost shorting out by the high conductivity at the off-equatorial E region in the daytime. One spectacular point in this figure is the regions adjacent to the pre-reversal enhancement, an electrodynamics process at the geomagnetic equator [Rishbeth, 1971; Farley et al., 1986; Eccles, 1998]. Investigating this figure, we may expect that higher electric potential exists between -60° and 0° in geographic longitudes (or LT20 and LT00) and at height between 200 km and 400 km due to the large convergence of electric current density. To understand the details further, the associated neutral disturbances, which produce the perturbed electric current density, are presented in Figure 6. Only the zonal component is shown in this figure; the effects of vertical and meridional component in producing perturbed electric current at the geomagnetic equator can be neglected. At a first glance, it is suspicious because the neutral disturbances are westward at regions between −60° and 0°. In fact, the upward electric currents result from the combined effects of neutral disturbances and conductivities in the storm period. As shown by Huang et al. , the perturbed zonal electric field can effectively depress the uplift height of the ionosphere in the pre-reversal period, and the higher conductivity results at the lower F region, between 200 and 400 km height, as compared with the quiet time condition. On the other hand, the zonal neutral wind is eastward at the equatorial ionospheric F region in the nighttime, and the magnitude is reduced by the geomagnetic activity, as shown in Figure 6, but the direction is still eastward. Now, we may understand that the upward electric currents, which appeared between sunset and mid-night, mainly resulted from the increased conductivity.
Figure 7 shows the variations of electric potential, 6 h after the onset of geomagnetic activity, in the geographic longitudes (along the geomagnetic equator) at 00UT, ranging from 100 to 500 km in altitude. Larger electric potential is accumulated in the nighttime, maximizing between 20 and 00 LT at about an altitude of 240 km. In this figure, the smaller contour interval has been used in area close to the maximum electric potential. This location, clearly correlated with the region with a large convergence of perturbed electric current density shown in Figure 5, is very close to the occurrence region of the pre-reversal enhancement. Therefore the model results make us strongly believe that the electrodynamics at normal quiet time equatorial ionosphere, i.e., the pre-reversal enhancement, play an important role in creating perturbed electric fields under the influence of geomagnetic activity.
 For the second case (i.e., at June solstice with mediate solar flux), we also do the same calculation as the first case. Figure 8 shows the distribution of the neutral disturbances dynamo with the same format as Figure 4. In Figure 8, the convergence of the disturbance dynamo after sunset is very small and locates at a lower altitude as compared with those in Figure 4. The corresponding distribution of the electric potential associated with the neutral disturbance dynamo is shown in Figure 9. We find that the location of the maximum electric potential has moved down to an altitude of about 180 km. This change of the position of maximum electric potential also reflects the change of perturbed electric fields (from westward to eastward and downward to upward).
 Direct comparison made from these two cases indicates that the effect of the geomagnetic activity on the equatorial ionosphere is season dependent and strongly related to the normal quiet time equatorial electrodynamics, i.e., the pre-reversal enhancement. From Figure 1 or the previous important works [Fejer et al., 1991; Fesen et al., 2000], the magnitudes of the pre-reversal enhancement is season and solar cycle dependent. Under the different conditions, the corresponding locations of the maximum electric potential should be different. To point out this fact, additional model runs associated with the geomagnetic activity are made with conditions same as those shown in Figure 1, i.e., at March equinox with high and low solar activities and at June solstice with high and low solar activities. The results show that the positions (altitude, local time) of the maximum electric potential locate at (280 km, 21.6 LT), (240 km, 22.3 LT), and (180 km, 23.3 LT) for March equinox with high, mediate, and low solar conditions, respectively; for the cases at June solstice, the positions locate at (200 km, 22.0 LT), (180 km, 22.3 LT), and (120 km, 23.3 LT) for high, mediate, and low solar conditions, respectively. It is very clear the magnitude of solar activity or pre-reversal enhancement unambiguously relates with the location of the maximum electric potential associated with the geomagnetic activity. Larger magnitude makes the location closer to the occurrence region, at height and local time, of the pre-reversal enhancement.
4. Discussion and Conclusions
 The influence of geomagnetic activity on the equatorial ionospheric electric fields mainly comes from the disturbance winds, set up by the Joule heating and ion drag in the auroral and polar cap regions, and the direct penetration field, through the interaction of the solar wind and magnetosphere. To study and explore the effects, we have made TIEGCM runs to simulate many required conditions. However, due to the limitation of TIEGCM [see Huang et al., 2005, for details], only the effect associated with the disturbance dynamo is considered in the present study.
 A significant feature result from the disturbance dynamo is the equatorward electric current, which tends to positively charge the low-latitude ionosphere and determine the distribution of the perturbed electric potential. The perturbed electric potential, in turn, determines the perturbed electric fields and the reversal time. These quantities, a departure from the quiet time electric fields, are often used to discuss and investigate the effects of geomagnetic activity on the ionosphere.
 In the present study, we find that the location of maximum electric potential is strongly affected by the normal quiet time electrodynamics in the equatorial ionosphere. During the occurrence of pre-reversal enhancement, the uplift height of the ionosphere is depressed by the westward perturbed electric field, which, in turn, enhances the conductivity in altitudes between 200 km and 400 km (for example, for the case at March equinox with mediate solar activity), and then produces the upward electric currents, relative to the quiet time condition. The upward electric currents coupled with the surrounding downward electric currents produce a large current convergence at locations after the pre-reversal enhancement. The associated distribution of electric potential, then, determines the perturbed electric fields and the reversal time of the zonal perturbed electric field. In the case without pre-reversal enhancement (for example, at June solstice with low solar activity), we find that the large current convergence disappears and the location of the maximum electric potential moves down to the region with very low conductivity, i.e., the E region.
 As shown in the model results, a large convergence of the perturbed electric current density produced by the disturbance dynamo occurs at regions close to the pre-reversal enhancement, and the intensity and the position of the associated positive charge accumulation is dependent with the magnitude of the pre-reversal enhancement. Larger magnitudes push the location of the maximum electric potential closer to the pre-reversal enhancement, at altitude and local time. The location of the maximum electric potential determines the reversal time of the perturbed zonal electric field from westward to eastward, and the altitude at which the perturbed vertical electric field changes from downward to upward. It is also clear that the magnitude of the pre-reversal enhancement is seasonal and solar cycle dependent [Fejer et al., 1991; Fesen et al., 2000], which, in turn, implies that the local time and height variations of the perturbed zonal and vertical electric fields in the storm period are seasonal and solar cycle dependent. Although the pre-reversal enhancement plays an important role in the perturbed electric fields produced by the disturbance dynamo in the storm period, the detailed examination of the deriving mechanism of the pre-reversal enhancement is beyond the scope of this paper.
 Other significant features revealed by the TIEGCM runs are: (1) a significant perturbed zonal electric field could be built up in six hours after the onset of geomagnetic activity, except at regions close to sunset and sunrise; (2) the magnitude of the perturbed vertical electric field is increased with the time length of energy input; and (3) above an altitude of 400 km, the perturbed vertical electric field is only significant in the nighttime period and very small in the daytime period.
 Although the results in this paper are only associated with the neutral disturbances, the effect also comes from the direct penetration electric field. The model studies of interaction between the direct penetration and the disturbance dynamo electric fields in the storm time equatorial ionosphere shown by Maruyama et al.  reveals that the mid- and low-latitude ionospheric conductivity and neutral wind changes originated from the direct penetration electric field are sufficient to influence the subsequent development of the disturbance dynamo at night. In TIEGCM, geomagnetic activities are simulated with assumed hemispheric power and cross polar cap potential drop at high latitudes. In this study, therefore, the role played by the direct penetration from high latitudes also exists and interacts with the disturbance dynamo.
 In this paper we only investigate the effect of the disturbance dynamo on the storm time equatorial electric fields. As for the other component, i.e., the direct-penetration electric fields, the readers may make a reference to an empirical model developed by Fejer and Scherliess .
 We thank A. D. Richmond for useful discussion concerning this paper. This work was supported by National Science Council of Taiwan through grants NSC-95-2111-M-008-013-MY3 and 95-NSPO(B)-RS3-FA07-02A.
 Amitava Bhattacharjee thanks Harish Chandra and another reviewer for their assistance in evaluating this paper.