SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Simulation Model
  5. 3. Results
  6. 4. Summary
  7. Acknowledgments
  8. References

[1] The first self-consistent study of the impact of storm-time penetration electric fields on the low- to mid-latitude ionosphere is presented. The inner magnetosphere is described by the Rice Convection Model (RCM) and the ionosphere is described by the Naval Research Laboratory (NRL) code SAMI3. The codes are coupled electrodynamically through the electrostatic potential equation, and the storm is modeled via changes in the polar cap potential. Neutral wind driven electric fields are estimated from the Fejer/Scherliess quiet time model. It is found that temporal changes in the polar cap potential produce electric fields that modify the F region equatorial E × B drift velocities: the velocities increase in the daytime and decrease in the nighttime by up to a factor of two. This causes the total electron content (TEC) in the daytime, mid-latitude ionosphere to increase by up to 35%. In addition, the ‘fountain effect’ is enhanced in the post-sunset period.

1. Introduction

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Simulation Model
  5. 3. Results
  6. 4. Summary
  7. Acknowledgments
  8. References

[2] It is well-established that electric fields of magnetospheric origin can penetrate to mid- and low-latitudes during magnetically active periods [Kelley et al., 1979; Spiro et al., 1988; Fejer and Scherliess, 1995; Basu et al., 2001; Huang et al., 2005] and that these fields can substantially modify low- to mid-latitude ionospheric plasma properties. However, determination of the impact of penetration electric fields on the ionosphere based on a self-consistent electrodynamic model has been lacking [Huba et al., 2003; Maruyama et al., 2005].

[3] In this Letter we report the first results of calculating storm-time penetration electric field effects on the low- to mid-latitude ionosphere using a self-consistently coupled inner magnetosphere/ionosphere model. The inner magnetosphere is modeled using the Rice Convection Model (RCM) and the ionosphere is modeled using the Naval Research Laboratory (NRL) code SAMI3. The codes are self-consistently coupled electrodynamically through the potential equation (see (1) below): SAMI3 provides the field-line integrated conductances to RCM and RCM provides the electrostatic potential to SAMI3. The storm is modeled using a time-dependent polar cap potential in RCM. Changes in the polar cap potential produce electric fields in the low- to mid-latitude that modify the F region equatorial E × B drift velocities: the velocities increase in the daytime and decrease in the nighttime by up to a factor of two. This causes the total electron content (TEC) in the daytime, mid-latitude ionosphere to increase by up to 35%. In addition, the ‘fountain effect’ is enhanced in the post-sunset period: the TEC in ionization crests increase and the crests move to higher latitudes.

2. Simulation Model

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Simulation Model
  5. 3. Results
  6. 4. Summary
  7. Acknowledgments
  8. References

[4] The Naval Research Laboratory has developed a three dimensional low- to mid-latitude ionosphere model – SAMI3. This model is based on the two dimensional model SAMI2 [Huba et al., 2000]. SAMI3 models the plasma and chemical evolution of seven ion species (H+, He+, N+, O+, N2+, NO2+ and O2+) in the altitude range 85 km–20,000 km. The complete ion temperature equation is solved for three ion species (H+, He+ and O+) as well as the electron temperature equation. Ion inertia is included in the ion momentum equation for motion along the geomagnetic field. The plasma is modeled from hemisphere to hemisphere in the low- to mid-latitude ionosphere up to ±55° magnetic latitude. The magnetic field is now modeled as dipole-fit to the IGRF at each longitude. However, in order to conform to RCM, the magnetic field is modeled as a dipole field aligned with the earth's spin axis in this study. The neutral species are specified using NRLMSISE00 [Picone et al., 2002]. The effect of neutral wind transport along the magnetic field is not included in the simulations to maintain hemispheric symmetry.

[5] The Rice Convection Model (RCM), developed at Rice University over 30 years, offers a self-consistent description of the electrodynamics of the coupled inner magnetosphere-ionosphere system in the slow-flow, closed field-line region [Wolf, 1983; Toffoletto et al., 2003]. The code solves the time-dependent coupled drift equations of magnetospheric plasma population represented by a number of ‘fluids’ with isotropic pitch-angle dependency, and the current-conservation equation for the ionospheric potential

  • equation image

where

  • equation image

Here the elements of the ionospheric conductance tensor, which include contributions from solar EUV and auroral particle ionizations, are field-line integrals Σθθ = ∫ σp/sin(I) ds, Σϕϕ = Σθθ sin2(I), and Σθϕ = ∫σHds, of Pedersen (σp) and Hall conductivities (σH), I is the inclination angle, J is the total density of the field-aligned current for both hemispheres, and jw represents contribution from neutral wind dynamo effects (see Wolf [1983] for details).

[6] In the current version of the coupled model, SAMI3 provides the solar-EUV part of (1) to the RCM. RCM solves for Φ at each time step; this potential is used in SAMI3 to transport the ionospheric plasma via E × B drifts, and an updated Σ is evaluated every time step (e.g., 2 s). The solution to (1) is thus consistent between the inner magnetosphere and the underlying ionosphere. Since the RCM region extends from 10° latitude into the auroral zone, and SAMI3 extends to ∼±55° magnetic latitude, the SAMI3 conductances are continued in a smooth way to higher latitudes with RCM-computed auroral enhancements added where electron precipitation is present; the next version of SAMI3 will cover the entire ionosphere. We estimated contributions from jw to Φ for quiet-time conditions using the Scherliess and Fejer [1999] empirical model.

[7] The magnetospheric magnetic field model used in RCM was that of Hilmer and Voigt [1995] with the following parameters: dipole tilt = 0°, Dst = −20 nT, magnetopause standoff distance = 10 Re, and the auroral boundary index (ABI) = 65° (invariant). For simplicity the magnetic field was held constant in time. The plasma was represented by 3 species (electrons, H+, and O+). The initial condition on plasma fluxes was an empty magnetosphere. For the outer boundary (X = −18 Re at midnight) we used κ = 6 in the distribution functions with the plasma sheet number density n = 0.2 cm−3, and the temperatures Ti = 4 keV for the ions and Te = 512 eV for electrons.

[8] The SAMI3 code used a grid (nz, nf, nl) = (101, 100, 48) where nz is the number of points along a dipole field line, nf is the number of points in altitude along the magnetic apex (i.e., number of magnetic field lines), and nl is the number of points in longitude. We only consider the four majority ion species: H+, O+, NO2+ and O2+. The geophysical parameters used are the following: day 91, year 1999, F10.7 = 181, F10.7A = 181, and Ap = 21. A simulation was run for 24 hrs to establish a steady state system; the simulation results reported here start with this initial system and run for an additional 9 hrs.

3. Results

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Simulation Model
  5. 3. Results
  6. 4. Summary
  7. Acknowledgments
  8. References

[9] Three simulations were run using different time dependent polar cap potential functions as shown in Figure 1. Case 0 is the base run and uses a constant polar cap potential 40 kV from 0000 UT to 0900 UT. Case 1 uses a step function with a 120 kV potential between 0100 UT and 0500 UT and Case 2 uses a linear potential function that increases to 240 kV at 0300 UT starting at 0100 UT and then decreases to the base level at 0500 UT.

image

Figure 1. Time dependent polar cap potentials.

Download figure to PowerPoint

[10] In Figure 2 we plot the vertical E × B drift velocity at the magnetic equator as a function of universal time for the 3 cases (Case 0: solid, Case 1: diamond, Case 2: box) at an altitude of 598 km and two longitudes: 0° [top panel] and 180° [bottom panel]. The top panel shows the drift in the nighttime and early morning, while the bottom panel shows it for the afternoon and early evening. In the top panel, Case 1 shows a sharp decrease at onset (0100 UT) and a sharp increase at cessation (0500 UT) relative to the baseline Case 0 (undershielding/overshielding). The decay time following the sharp decrease/increase is ∼1 hr although it can take several hours before E × B drift completely returns to the non-disturbed case. In contrast, Case 2 shows a less abrupt departure from Case 0 but the overall change is greater than Case 2. In addition, the direction of change in the E × B drift relative to Case 0 is determined by the slope of the change in the polar cap potential. The maximum change in the drift velocity caused by the penetration electric field is roughly a factor of 2. The bottom panel (longitude 180°) shows the effect in the daytime: the time ranges from 1200–2100 LT. The behavior of the disturbed drifts at onset (100 UT) is similar to the top panel except that drifts are positive. However, at cessation (500 UT) there is no a very small effect of the penetration electric field; the pre-reversal enhancement of the equatorial drift dominates.

image

Figure 2. Vertical E × B drift as a function of universal time (UT).

Download figure to PowerPoint

[11] In Figure 3 we plot the vertical equatorial E × B drift as a function of magnetic local time (MLT) for Case 1 just after 0100 UT (top panel) and Case 2 at 0230 UT (bottom panel). In each panel we plot the drift for Case 0 for comparison. The qualitative behavior of the effect of the penetration electric field on the vertical drifts is the same for both cases: the drifts increase in the daytime and decrease in the nighttime. Moreover, the magnitude of the increase and decrease of the drifts are similar in both cases although Case 1 shows slightly larger variations consistent with Figure 2.

image

Figure 3. Vertical E × B drift as a function of magnetic local time (MLT).

Download figure to PowerPoint

[12] The results shown in Figure 3 for Case 1 are qualitatively consistent with previous RCM studies of penetration fields, where shielding by the Region-2 field-aligned currents breaks down because of a sudden strengthening of convection [Jaggi and Wolf, 1973], but there are quantitative differences. For example, the penetration field (difference between the two curves in the top panel of Figure 3) has a peak near 18.5 MLT because of Hall currents crossing the conductance change at the terminator [Wolf, 1970]. The dusk peak in Figure 3 is stronger than the corresponding feature in Figures 2 and 3 of Spiro et al. [1988] that used calculations based on a Chiu [1975] ionosphere; it is somewhat stronger than what the RCM predicts based on the IRI-90 empirical model, which is now typically used with the stand alone RCM, but is somewhat weaker than what the RCM calculates [Sazykin, 2000] from the SUPIM model.

[13] Finally, in Figure 4 we show the impact of the penetration electric field on the total electron content (TEC) at time 0501 UT when the changes in the polar cap potential end. We show color contour plots of the difference in TEC between Case 1 and Case 0 (top panel) and between Case 2 and Case 0 (bottom panel). The local time at the center of the plots is 2101 (longitude −120°). The qualitative behavior of the TEC variation is the same for both cases: the TEC increases in the afternoon, mid-latitude ionosphere at latitudes above the nominal ionization crests ±15° and decreases late in the evening until late morning. Interestingly, large TEC enhancements have been observed in the mid-latitude ionosphere during major storms [Foster et al., 2002]. There is also evidence of an enhanced fountain effect between 2000 LT and 2100 LT: the ionization crests are enhanced at higher latitudes and there is a reduction in TEC in the equatorial region. This occurs because of the increased daytime and pre-reversal enhancement of the upward E × B drift as shown in Figure 3. This result is consistent with the observations reported by Basu et al. [2001] and Foster et al. [2005]. In addition, we find that Case 2 has a much larger effect on the TEC than Case 1. The daytime, mid-latitude increase in TEC is ∼25 TECU or a ∼35% increase in TEC from Case 0. On the other hand, for Case 1 the increase in TEC is ∼15 TECU or a ∼25% increase in TEC from Case 0. The reason for this larger effect is that the E × B drift associated with Case 2 deviates from Case 0 over a much longer time than Case 1.

image

Figure 4. Differential TEC at 0501 UT.

Download figure to PowerPoint

4. Summary

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Simulation Model
  5. 3. Results
  6. 4. Summary
  7. Acknowledgments
  8. References

[14] We have presented the first self-consistent simulation study of the impact of storm-time penetration electric fields on the low- to mid-latitude ionosphere using an electrodynamically coupled ionosphere and inner magnetosphere model. We find that changes in the polar cap potential produce electric fields in the low- to mid-latitude that modify the F region equatorial E × B drift velocities: the velocities increase in the daytime and decrease in the nighttime by up to a factor of two. This causes the total electron content (TEC) in the daytime, mid-latitude ionosphere to increase by up to 35%. In addition, the ‘fountain effect’ is enhanced in the post-sunset period: the TEC in ionization crests increase and the crests move to higher latitudes.

[15] Some remarks are needed concerning the overall strength and duration of these penetration events. In Case 1, where the time change in the polar cap potential is sudden, the duration of the penetration electric field is shorter than typical observed penetration events. On the other hand, for Case 2, the duration increased when the polar cap potential changed more gradually in time. This may explain the long-lived storm-time eastward electric field reported by Huang et al. [2005]. It should also be noted that an increase in polar cap potential is usually caused by a southward turning of the IMF that also causes a change in the magnetospheric magnetic field (e.g., tail stretching, dayside erosion). The effects of the changing magnetic field, which have been neglected in our first runs with the coupled SAMI3/RCM model, would add to the undershielding associated with the change in potential drop [Fejer et al., 1990; Garner, 2003]; inclusion of these effects in the model are likely to result in stronger and longer lasting penetration electric fields, and thus, larger changes in the TEC.

[16] We emphasize that these results are preliminary and that more work is needed to fully understand and quantify the impact of penetration electric fields on the low- to mid-latitude ionosphere. For example, we intend to extend SAMI3 to the high latitude region. Subsequently, we intend to explore the ionospheric response to storm-time fields as a function of geophysical parameters (e.g., geomagnetic conditions, solar EUV, longitude, latitude, polar cap potential, etc.). These results will be presented at length in a future paper.

Acknowledgments

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Simulation Model
  5. 3. Results
  6. 4. Summary
  7. Acknowledgments
  8. References

[17] One of us (JDH) thanks C.-S. Huang, M. Swisdak, and J. Jasperse for helpful discussions. This research has been supported by the Office of Naval Research and by the National Aeronautics and Space Administration under grant NAG5-13524 issued through the Living with a Star program.

References

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Simulation Model
  5. 3. Results
  6. 4. Summary
  7. Acknowledgments
  8. References