simulation of electric field and current during the 11...

Simulation of electric field and current during the 11 June 1993disturbance dynamo event: Comparison with the observations

K. Z. Zaka,1 A. T. Kobea,1 V. Doumbia,1 A. D. Richmond,2 A. Maute,2 N. M. Mene,1

O. K. Obrou,1 P. Assamoi,1 K. Boka,1 J.‐P. Adohi,1 and C. Amory‐Mazaudier1

Received 1 March 2010; revised 6 July 2010; accepted 2 August 2010; published 9 November 2010.

[1] The ionospheric disturbance dynamo signature in geomagnetic variations isinvestigated using the National Center for Atmospheric Research Thermosphere‐Ionosphere‐Electrodynamics General Circulation Model. The model results are testedagainst reference magnetically quiet time observations on 21 June 1993, and disturbanceeffects were observed on 11 June 1993. The model qualitatively reproduces the observeddiurnal and latitude variations of the geomagnetic horizontal intensity and declination forthe reference quiet day in midlatitude and low‐latitude regions but underestimates theiramplitudes. The patterns of the disturbance dynamo signature and its source “anti‐Sq”current system are well reproduced in the Northern Hemisphere. However, the modelsignificantly underestimates the amplitude of disturbance dynamo effects when comparedwith observations. Furthermore, the largest simulated disturbances occur at different localtimes than the observations. The discrepancies suggest that the assumed high‐latitudestorm time energy inputs in the model were not quantitatively accurate for this storm.

Citation: Zaka, K. Z., et al. (2010), Simulation of electric field and current during the 11 June 1993 disturbance dynamo event:Comparison with the observations, J. Geophys. Res., 115, A11307, doi:10.1029/2010JA015417.

1. Introduction

[2] During high‐geomagnetic activity periods, disturbancewinds from high‐latitude regions may influence the low‐latitude ionosphere. The disturbance wind dynamo and itsinfluences in midlatitudes and low latitudes were firstinvestigated by Blanc and Richmond [1980] and Spiro et al.[1988]. These dynamo effects have been associated withmidlatitude winds driven by the high‐latitude Joule heatingand with fossil winds accelerated by strong ion convectionin the auroral regions. Blanc and Richmond [1980] proposedthe ionospheric disturbance dynamo mechanism to explainelectric field disturbances observed at the end of magneticstorms. These electric field disturbances are assumed to bedue to the disturbed thermospheric wind actions [Fejeret al., 1983; Fejer and Scherliess, 1995; Fejer, 2002;Sastri, 1988; Fambitakoye et al., 1990; Mazaudier andVenkateswaran, 1990]. Richmond et al. [2003] studied ion-ospheric electric field disturbances at midlatitudes and lowlatitudes using the Magnetosphere‐Thermosphere‐IonosphereElectrodynamics General Circulation Model (MTIEGCM)[Peymirat et al., 1998]. They pointed out that three compo-nents of comparable importance can be associated with low‐latitude ionospheric electric fields during disturbances,namely, dynamo effects of global tidal winds; direct pene-

tration of polar cap electric fields directly to the magneticequator; and dynamo effects of disturbance winds caused byhigh‐latitude Joule heating.[3] The influence of geomagnetic activity on midlatitude

and low‐latitude thermospheric winds and ionospheric elec-tric field has been investigated using the National Center forAtmospheric Research (NCAR) Thermosphere‐Ionosphere‐Electrodynamics General Circulation Model (TIE‐GCM) byHuang et al. [2005]. The model results showed that, when thegeomagnetic activity ceases, zonal disturbance winds can lastfor many days in the postrecovery period, while the meridi-onal disturbance winds vanish more rapidly. These results arein accordance with the changes in the thermospheric circu-lation [Testud and Vasseur, 1969; Richmond and Roble,1979; Mazaudier et al., 1985], which lead to disturbancesof electric fields and currents at midlatitudes and low latitudes[Volland and Mayr, 1971; Testud et al., 1975; Blanc, 1978].However, the ionospheric disturbance dynamo effects observedat midlatitudes and low latitudes through measurements ofelectric fields and currents [Fejer et al., 1983;Mazaudier andVenkateswaran, 1990; Fejer and Scherliess, 1995; Abduet al., 1997] have not yet been fully reproduced by models.[4] In the present study, we use the NCAR TIE‐GCM to

investigate the effects of disturbance winds on the electricfield and current in the midlatitude and low‐latitude iono-sphere during the disturbance dynamo (Ddyn) event on 11June 1993 [Zaka et al., 2009]. To that end, the observedmagnetic variations are compared with the model results. Infact, electric current simulations at midlatitudes and lowlatitudes with the TIE‐GCM have not yet been presented fordisturbed conditions. Using the TIE‐GCM to simulate the

1Laboratoire de Physique de l’Atmosphère, UFR SSMT, Université deCocody, Abidjan, Ivory Coast.

2High Altitude Observatory, National Center for AtmosphericResearch, Boulder, Colorado, USA.

Copyright 2010 by the American Geophysical Union.0148‐0227/10/2010JA015417

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 115, A11307, doi:10.1029/2010JA015417, 2010

A11307 1 of 11

http://dx.doi.org/10.1029/2010JA015417

equatorial electrojet (EEJ) magnetic perturbations, Doumbiaet al. [2007] focused their study on magnetically quietperiods only. Furthermore, the previous simulations per-formed by Richmond et al. [2003] did not assess the capa-bility of the model to reproduce the current circulationassociated with the ionospheric disturbance dynamo mech-anism at midlatitudes and low latitudes. By comparing thesimulated and observed geomagnetic perturbations and byexamining the TIE‐GCM winds in the dynamo region thathelped create these perturbations, we can gain informationabout how well the TIE‐GCM is able to simulate the dis-turbance winds.

2. Model Description and Input Parameters

[5] As described by Fang et al. [2008], the NCAR TIE‐GCM is a three‐dimensional, time‐dependent model whichsolves the full dynamical equations of the coupled thermo-sphere and ionosphere self‐consistently [Dickinson et al.,1984; Roble et al., 1988; Richmond et al., 1992]. It is de-signed to calculate the coupled dynamics, chemistry, ener-getic, and electrodynamics of the global thermosphere‐ionosphere system between about 97 and 500 km altitude. Inparticular, the TIE‐GCM calculates the ionospheric electricfields and currents and their associated magnetic perturba-tions. With a specified day number of the year, F10.7 solarflux, high‐latitude hemispheric power of precipitatingauroral particles, and cross‐polar‐cap electric potential, themodel calculates global electric fields and currents, ion andneutral densities, temperatures, compositions, and velocities.The solar extreme ultraviolet (EUV) radiation is specified bythe EUV flux model for aeronomic calculations (EUVAC)[Richards et al., 1994; Solomon, 2006]. At the lowerboundary, approximately 97 km, migrating tidal perturba-tions are driven by the Global Scale Wave Model (GSWM)from Hagan and Forbes [2002, 2003]. At the upperboundary, approximately 500 km, vertical O+ fluxesbetween the ionosphere and the plasmasphere are specified.The details of the TIE‐GCM calculations of electric fields

and currents and the associated magnetic perturbations aredescribed by Doumbia et al. [2007] and Fang et al. [2008].[6] In this study, the simulations are performed for the

specific days of 11 and 21 June 1993. The geophysicalindices of these days, especially the daily F10.7 number and3 hourly Kp index provided by the National GeophysicalData Center (NGDC), are used to drive the TIE‐GCM. TheTIE‐GCM uses a realistic geomagnetic field model (Inter-national Geomagnetic Reference Field) and calculates theelectrodynamics in magnetic apex coordinates [Richmond,1995]. The horizontal resolution is 5° by 5° in geographiclongitude and latitude, with two grid points per scale heightvertically. A higher resolution is used for calculating theelectric fields and currents on a geomagnetically orientedgrid, especially near the equator to resolve the EEJ. Thepolar cap electric potential distribution is prescribed by theHeelis et al. [1982] model. The cross‐polar potential dif-ference and the hemispheric power are estimated throughthe Kp index. At midlatitudes and low latitudes, the TIE‐GCM solves the equations of the ionospheric wind dynamo.At auroral latitudes, the dynamo equations are modified overa transition region 15° wide, with an increasingly strongimposition of the Heelis et al. [1982] model as magneticlatitude increases toward the polar cap. (Equations (4)–(6) ofPeymirat et al. [2002] detail how the electric potential issolved.) The imposed high‐latitude electric potential affectsthe low‐latitude solution in a manner qualitatively similar tothe effect of direct penetration electric fields, although theTIE‐GCM numerical procedure was not designed to simu-late these penetration fields physically. We make use of thisqualitative similarity in the present study to estimate roughlythe effects of true direct penetration electric fields on themidlatitude and low‐latitude magnetic perturbations.

3. Model Results

3.1. Model Simulation Context

[7] The primary goal of this study is emphasizing thedisturbance dynamo effects in midlatitude and low‐latituderegions on the basis of the TIE‐GCM simulations. Themodel simulations are performed in the geophysical con-texts of the quiet day 21 June 1993, for which the sum of theKp indices is 2+, and of the storm recovery day 11 June1993 (SKp = 17). Figure 1 shows variations of the Dstindex (Figure 1a) and the AE index (Figure 1b) for 20–21June (dashed lines) and 10–11 June (solid lines). On 21June, the Dst index varies very slowly with an amplitudelower than 14 nT. The amplitude of the AE index is rela-tively weak as well (less than 100 nT). On 10–11 June, theDst index (solid line) exhibits the different phases of amagnetic storm that started with an SC (sudden com-mencement) around 1600 UT on June 10. The SC is fol-lowed by the main phase of the storm, which started around2000 UT and ended around 0400 UT on 11 June. Therecovery phase of the storm started after 0400 UT. Duringthe main phase of the storm, the AE index (Figure 1, bot-tom) is strongly enhanced, attaining a peak value higherthan 1600 nT around 2300 UT on 10 June. During therecovery period, although having strongly decreased, thevalues of the AE index average around 200 nT, with peakvalues of about 300 nT at 1200 UT and 500 nT at 1800 UTon 11 June. The strong enhancement of the AE index

Figure 1. Variations of the (a) Dst and (b) AE indices on10–11 June (solid lines) and 20–21 June 1993 (dashedlines).

ZAKA ET AL.: DISTURBANCE DYNAMO CURRENT A11307A11307

2 of 11

indicates that the auroral activity was intensified during themain phase of the storm. Although it strongly decreases afterthe main phase of the storm, during the recovery the auroralactivity is still important.[8] This study focuses on the disturbance dynamo effect

by examining the diurnal variations of the simulated hori-zontal northward H and eastward D components of thegeomagnetic field. The TIE‐GCM model was run for thestorm recovery day 11 June and the quiet reference day 21June described above. The model includes the effects ofdisturbance dynamo and direct penetration of the imposedhigh latitude electric fields to low latitude. The diurnalvariations of H and D simulated by TIE‐GCM include thesetwo effects, and these runs are referred to as case 2. Toisolate the effect of the disturbance dynamo a second set ofH and D calculations was performed without penetrating

electric fields, which we refer to as case 1. This was done bysetting the imposed high‐latitude electric potential to a verysmall value but still using the neutral winds and conduc-tivities from the case 2 runs when solving for the low‐lati-tude and midlatitude electric fields and currents.[9] In the simulation, the strongest disturbance winds

occur at high magnetic latitudes, where forcing by ionconvection and Joule heating is largest. The disturbancewinds also extend to midlatitudes and low latitudes, asshown in Figure 2 at 130 km altitude, between ±60° latitude,near the end of the main phase of the storm (11 June, 0 UT)and many hours after the storm main phase (11 June,12 UT). The plots show the TIE‐GCM disturbance winds,which are the difference between TIE‐GCM winds on 11June and the reference quiet day 21 June. The height of130 km is in the region of large daytime Pedersen

Figure 2. TIE‐GCM disturbance winds at 130 km height. The disturbance winds are determined by thedifference between winds on the disturbed day 11 June and the reference quiet day 21 June (a) for 0 UTand (b) for 12 UT.


3 of 11

conductivity, where disturbance dynamo action is important.After the main phase of the storm, the forcing of the windssubsides and the winds relax back toward their quiet daypattern due to dissipative effects. However, between 0 and12 UT, the disturbance winds in Figure 2 do not simplydiminish in amplitude, they also change greatly in pattern, ascan be seen in the longitudinal change of the largest west-ward winds between 0 UT and 12 UT. The winds generallymaintain a westward direction during the recovery phase ofthe storm. Interestingly, at 12 UT a strong anticyclonicdisturbance wind vortex has expanded out of high latitudestoward western North America, and the westward windsaround 90°E have increased in strength in the NorthernHemisphere with respect to their values at 0 UT.[10] Emmert et al. [2002] presented an empirical model of

midlatitude daytime WINDII disturbance winds, represen-

tative of average disturbance winds during magneticallyactive periods, but not of poststorm winds. They showedthat the average disturbance winds are roughly constant inheight above about 130 km. The 0 UT winds in Figure 2exhibit some properties of the empirical model, such asthe tendency to generate a westward zonal disturbance windat midlatitudes and the tendency for the wind strength toincrease with latitude, but there are also significant differ-ences. Whereas the empirical model shows a daytimeequatorward disturbance wind component comparable withthe zonal component at midlatitudes, this feature is not soapparent in Figure 2. Emmert et al. [2002] noted that there isa large amount of variability from one storm to another, andso we do not expect to find good agreement in detail.Although direct measurements of disturbance winds for aparticular storm are always very sparse, we can test thesimulated winds indirectly by comparing the dynamo effectsthey produce with observed geomagnetic perturbations, aswe show in the following.[11] Figures 3a and 3b depict the TIE‐GCM equivalent

current function for 11 June, 1200 UT, with near‐zero polarcap potential (case 1) and with the disturbed polar cappotential (case 2), respectively. The coordinates are mag-netic local time (MLT) and magnetic latitude. Figure 3ashows two hemispheric vortexes, which represent storm‐modified Sq vortexes. The northern vortex, which is muchmore developed for these northern summer conditions, ex-pands to the Southern Hemisphere in the morning. InFigure 3b, the current function is strongly modified at highlatitudes owing to the large polar cap potential, and it isenhanced in the midlatitude and low‐latitude regions owingto direct penetration of the imposed polar electric fieldtoward lower latitudes. In section 3.2, we analyze and com-pare the simulated diurnal variations of the H and D com-ponents retrieved from the two cases in order to investigatethe ionospheric disturbance associated with electric fieldpenetrations as well as with disturbance winds.

3.2. Simulating the Ionospheric Disturbance DynamoEffects in the Midlatitude and Low‐Latitude Regions

[12] The diurnal variations of the geomagnetic field aresimulated at the locations of seven observatories and tem-porarily stations. These stations are distributed along ameridian chain in the Europe‐African longitude sector. Thegeographic and magnetic apex coordinates of the stationsare in Table 1.[13] Figures 4a and 4b show the simulated diurnal varia-

tions of the H and D components, respectively, of thegeomagnetic field in the Europe‐African longitude sector on21 June 1993 (the reference quiet day) and 11 June 1993(the disturbed day). The solid lines represent the simulationsin case 1, and the dashed lines represent simulations in case2. During the reference quiet day (Figure 4a), the amplitudesof the diurnal variations of H and D are similar in bothcases, except for the northernmost stations, which are Ler-wick (60.13°N) and Chambon‐La‐Foret (48.03°N), and forSikasso (11.34°N), close to the dip equator. Small differ-ences are observed between the diurnal variations of H andD in the two cases at Lerwick and Chambon‐La‐Foret; atSikasso the difference is observed only for the H component,owing to the enhanced east‐west Cowling conductivity nearthe magnetic equator. The difference in the H component

Figure 3. Equivalent current function (in kA) simulated bythe TIE‐GCM on 11 June 1993 at 1200 UT: (a) withoutdirect penetration of electric field (case 1) and (b) with directpenetration of electric field (case 2).


4 of 11

associated with the simulated electric field penetration pro-gressively decreases from high to low latitudes and slightlyincreases near the dip equator after vanishing in the mid-latitude stations. In contrast, during the disturbed day(Figure 4b), the differences between the diurnal variations ofH in the two cases are significant in all the stations, with thehighest amplitudes in the northernmost stations and at the dipequator. As regards the D component variations, they exhibitdecreasing difference in the amplitude from the northernmoststations to midlatitude which vanishes in the vicinity of themagnetic equator without any enhancement. In fact, theCowling conductivity drives east‐west currents that have noeffect in the D component variations at the magnetic equator.In the following analysis, we therefore focus on the H com-ponent variations only.[14] The trend of the latitudinal variation of the simulated

direct penetration disturbance effect in the H componentagrees with the latitudinal profile of DP2 disturbance es-tablished by Kikuchi et al. [1996] and Kobea et al. [2000].In fact, the DP2 disturbance associated with the fluctuationsin the north‐south component of the interplanetary magneticfield is characterized by a quasiperiodic variation with atime of half an hour to several hours which occurs simul-taneously at high and low latitudes; its equivalent currentsystem consists of two asymmetric vortices originating inthe polar cap current flow and of zonal flow in the equatorialregion [Nishida et al., 1966].[15] In Figure 5, we compare the case 1 simulations of the

diurnal variations of H on 11 June (solid lines) and on 21June 1993 (dashed lines). Since the polar cap potential wasset to nearly zero, the differences between the diurnal var-iations on the reference quiet day and on the disturbancedynamo day correspond primarily to the effects of the dis-turbance thermospheric winds, although there may also be arelatively small influence from conductivity differences onthe 2 days. We notice that, unlike the effects of electric fieldpenetrations which tend to enhance the strength of thediurnal variations of H at midlatitudes and low latitudes(Figures 4a and 4b), the effects of disturbance winds tend toreduce it. Figure 5 shows the amplitude of the diurnal var-iations of H is higher on 21 June (reference quiet day) thanon 11 June 1993 (disturbance day).

3.3. Comparison Between Model Simulations andObservations

[16] Figure 6 compares the simulated diurnal variations ofthe H component in case 1 (solid lines) and case 2 (dashedlines) with the observations (dotted lines) on 11 June (left)as well as on 21 June 1993 (right). Because the TIE‐GCMdoes not model the geomagnetic effects of distant magne-

tospheric currents like the ring current, the effects of asymmetric ring current, as derived from the Dst index, havebeen removed from the observations as of Zaka et al. [2009]and Le Huy and Amory‐Mazaudier [2008]. The magnitudeof the observations is larger than that of the simulations inthe two cases. During the disturbed day (11 June) the case 2simulations are closer to the observations at the low‐latitudestations than are the case 1 simulations. At high‐latitude andmidlatitude stations, the observed diurnal variation of Hexhibits a strong negative depression in the morning, at-taining a minimum around 1130MLT at Lerwick and 1000MLT at Chambon‐La‐Foret. Then it increases in the after-noon and gets closer to the case 2 simulation. On 21 Junethe morning depression is also observed at those stations,but is weaker. Note that the morning depression in theobserved diurnal variations of H is not reproduced by thesimulations. The amplitude of the observed diurnal variationof H at Sikasso near the dip equator is remarkably higherthan (more than double) the simulations. Doumbia et al.[2007] also remarked that the TIE‐GCM underestimatesthe amplitude of the diurnal variation of the equatorialelectrojet magnetic effects.[17] On 11 June 1993 (Figure 6, left), both simulation

cases reproduce the observations reasonably well at the low‐latitude stations. The difference between the observationsand simulations is smaller for case 2 than for case 1 at thosestations, because the effects of the direct penetration electricfield increase the simulated amplitude. In contrast, on June21 (Figure 6, right) both simulations significantly underes-timate the observation amplitudes. On this quiet day, theamplitudes of the simulations are quite similar at low lati-tude, which means that the simulated effect of electric fieldpenetration is very weak there. This result suggests that themagnetospheric convection electric fields might not affectsignificantly the midlatitude and the low‐latitude currentsystems during quiet days in the TIE‐GCM. However, as wementioned earlier, the limitations of the TIE‐GCM for cal-culating direct penetration electric fields prevents a defini-tive conclusion on this question. Furthermore, regarding thebig difference between the amplitudes of the simulationsand observations at Sikasso, it could be possible that thepenetration eastward electric field contributes to increase theeastward current intensity near the dip equator even duringquiet days, as found in the MTIEGCM results shown byRichmond et al. [2003]. Another major contributor to theunderestimation of the magnetic perturbation at Sikasso inTIE‐GCM results is the underestimation of E region elec-trical conductivity.Fang et al. [2008] found that the TIE‐GCMunderestimates the E region electron density by approxi-mately 37%, as compared with the International Reference

Table 1. Geographic and Magnetic Apex Coordinates of the Magnetic Stations Used in the Europe‐African Longitude Sector

Code Name

Geographic Coordinates Magnetic Coordinates

LTLatitude (°N) Longitude (°W) Latitude (°N) Longitude (°W)

LER Lerwick 60.13 358.82 62.37 89.19 +0CLF Chambon‐La‐Forêt 48.03 2.26 49.84 85.06 +1TAM Tamanrasset 22.79 5.53 24.66 80.31 +1TOM Tombouctou 16.73 357.00 6.36 71.15 +0SIK Sikasso 11.34 354.30 0.37 67.96 +0TSU Tsmeb −19.20 17.58 −18.77 83.51 +2HER Hermanus −34.42 19.23 −33.98 81.35 +2


5 of 11

Ionosphere (IRI). The reasons for this are not well under-stood, but one possible reason could be an underestimation ofthe solar X‐ray fluxes used by the model. We did not attemptto correct this underestimation. In section 3.4, we compare thesimulated and observed disturbance dynamo effects in case 1as well as in case 2, in order to show how important are thedisturbance thermospheric wind dynamo effects at midlati-tudes and low latitudes.

3.4. Latitudinal Variations of Simulated and ObservedDisturbance Dynamo Effects

[18] The disturbance effects are estimated for the case 1and case 2 simulations and for observations on 11 June by

subtracting the corresponding diurnal variations of the Hcomponent on 21 June 1993. In Figure 7, the resultingdisturbance dynamo (Ddyn) effects alone (case 1) and withdirect penetration electric field effects (case 2) that areassociated with the 11 June disturbance simulations arecompared with observations from the ground‐based data.The observations represent a combination of Ddyn anddirect penetration electric field effects. The dotted lines arethe zero reference level. At Lerwick, the simulated Ddynand observed disturbance are positive and have the sameamplitude range, apart from the fluctuations in the observeddisturbance. The amplitude of the simulated Ddyn withdirect penetration electric field is higher than the two others.

Figure 4. (a) Diurnal variations of the horizontal (left) northward (H) component and (right) eastward(D) component of the geomagnetic field simulated by the TIE‐GCM on 21 June 1993 (the reference quietday). The solid lines represent the case 1 simulations, and dashed lines represent the case 2 simulations.(b) Same as in Figure 4a, on 11 June 1993 (the day of disturbance dynamo).


6 of 11

At Chambon‐La‐Foret, while the observed disturbance be-comes negative, the simulated disturbances in the two casesremain slightly positive with very weak amplitude. At low‐latitude stations (Tamanraset, Tombouctou, and Sikasso) theobservations and the simulated disturbance dynamo arenegative with increasing amplitudes as we get close to thedip equator, where the magnitude of the observed distur-bance is 3 times as large as the simulated Ddyn magnitudeof case 1 and more than 10 times as large as the simulatedDdyn with direct penetration electric field of case 2. Thesimulated amplitude of case 2 is very weak at the dipequator. Note that the amplitudes of both case 1 and case 2are far weaker than the observations and that the local timeof the minimum occurs earlier for the simulations (0900 LT)than for the observations (1200 LT). In the southern stations

(Tsmeb and Hermanus), the observed disturbance is nega-tive and weak, while the simulated disturbance in the twocases is nearly zero so that their traces matched exactly thezero reference level at HER. The latitudinal trend of Ddyn isconsistent with the observations, in that the daytime hori-zontal Ddyn variations are southward at the low‐latitudestations and northward at high‐latitude stations. The struc-ture of the horizontal Ddyn variations is associated with anequivalent current system, which flows westward at lowlatitudes and eastward at high latitudes. Such a currentsystem corresponds well to the structure of an “anti‐Sq”current system that had been set forth by diverse workersthrough observations [Fambitakoye et al., 1990]. However,the TIE‐GCM model seems to underestimate the magneticeffects of this current. Indeed, the simulated disturbance in

Figure 4. (continued)


7 of 11

the two cases is far weaker than the observations. It ispossible that the simulated disturbance winds are too weakand may not have the correct longitudinal variation to pro-duce the observed Ddyn variations. We noted earlier that theTIE‐GCM winds at northern midlatitudes around middayhave greatly diminished by 12 UT. The relatively weakdisturbance winds over much of the summer hemispherehelp explain the relatively weak disturbance dynamo effectspredicted by the TIE‐GCM. Since the disturbance windsdepend on the intensity and distribution of high‐latitudeenergy inputs during the preceding storm, it is entirelypossible that the assumed distribution of the energy inputs

does not adequately represent the true inputs for this stormand that the simulated winds therefore underestimate thetrue winds in the dynamo region. Such information can beuseful to improve future modeling of thermospheric storms.

4. Conclusion

[19] In the present study, we have examined the magneticfield variations associated with the ionospheric disturbancedynamo event on 11 June 1993. This event has been sim-ulated by the TIE‐GCM and compared with observations inthe Europe‐Africa longitude sector. The TIE‐GCM simu-

Figure 5. Comparison of the diurnal variations of the case 1 simulated H component on 11 June 1993(solid lines) with those of the reference quiet day 21 June 1993 (dashed lines).


8 of 11

lations were performed first by ignoring the penetration ofthe magnetospheric convection electric field to midlatitudesand low latitudes and then by including it. In addition thegeophysical conditions of 21 June 1993, selected as thereference magnetically quiet day, were used to simulate thequiet time diurnal variations of the geomagnetic field.[20] The analysis of the simulated diurnal variations of the

H and D components of the geomagnetic field showed thatthe TIEGCM qualitatively reproduces the features of quietday magnetic variations associated with the midlatitude andlow‐latitude ionospheric current systems, although it un-derestimates the amplitudes. At the dip equator, our resultsare in accordance with the equatorial electrojet magneticsignature analyzed by Doumbia et al. [2007]. The analysisof the geomagnetic field variations during the disturbanceperiod showed that the contributions of the regular iono-spheric wind dynamo, of the disturbance dynamo, and of themagnetospheric convection electric field penetrations over-

lap at midlatitudes and low latitudes. On the one hand, theeffects of the disturbance dynamo tends to reduce the mag-netic effects of the regular ionospheric wind dynamo, whileon the other hand the effects of magnetospheric convectioneastward electric field penetrations tend to increase the effectsof the regular ionospheric wind dynamo currents at low lati-tudes. The patterns of the disturbance dynamo signature andits source “anti‐Sq” current system are well reproduced in theNorthern Hemisphere. However, the model significantlyunderestimates the amplitude of disturbance dynamo effectswhen compared with observations. Furthermore, the H dis-turbance minima occur at different local times than the ob-servations. The discrepancies suggest that the assumed high‐latitude storm time energy inputs in the model do not ade-quately represent the true inputs for this storm.[21] The magnitudes of the penetration electric field and

disturbance dynamo effects are strongly enhanced at the dipequator. The mechanism of this enhancement is related to

Figure 6. Comparison of the diurnal variations of the simulated H component in case 1 simulations(solid lines) and in case 2 simulations (dashed lines) with the observations (dotted lines) on (left) 11 June1993 and (right) 21 June 1993.


9 of 11

the enhanced Cowling east‐west electrical conductivity as-sociated with the equatorial electrojet along the magnetic dipequator. During magnetically quiet days, the thermosphericwind disturbance effects do not exist or are very weak, butthe effects of the convection electric field can be significantaccording to the level of the cross polar electric potentialdrop. Thus, the day to day, seasonal, and solar activitychanges might have nonnegligible influences on the vari-ability of the equatorial electrojet. At the equator, the elec-tric field penetrations should be realistically taken intoaccount in evaluating the disturbance as well as in analyzingthe magnetic signature of the equatorial electrojet.

[22] Acknowledgments. We thank T.‐W. Fang for helpful commentson an earlier draft of the manuscript. A. Richmond and A. Maute were sup-ported in part by the NASA LivingWith a Star program and NASA C/NOFSGuest Investigator grant NNX09AN57G. The National Center for Atmo-spheric Research is sponsored by the National Science Foundation.

[23] Robert Lysak thanks the reviewers for their assistance in evaluat-ing this paper.

ReferencesAbdu, M. A., J. H. Sastri, J. MacDougall, S. I. Batista, and J. H. A. Sobral(1997), Equatorial disturbance dynamo electric fields longitudinal struc-ture and Spread‐F: A case study from GUARA/EITS campaigns, Geo-phys. Res. Lett., 24(13), 1707–1710, doi:10.1029/97GL01465.

Blanc, M. (1978), Midlatitude convection electric fields and their relation toring current development, Geophys. Res. Lett., 5(3), 203–206,doi:10.1029/GL005i003p00203.

Blanc, M., and A. D. Richmond (1980), The ionospheric disturbancedynamo, J. Geophys. Res. , 85(A4), 1669–1686, doi:10.1029/JA085iA04p01669.

Dickinson, R. E., E. C. Ridley, and R. G. Roble (1984), Thermosphericgeneral circulation with coupled dynamics and composition, J. Atmos.Sci., 41, 205–219.

Doumbia, V., A. Maute, and A. D. Richmond (2007), Simulation of equa-torial electrojet magnetic effects with the Thermosphere‐Ionosphere‐Electrodynamics General Circulation Model, J. Geophys. Res., 112,A09309, doi:10.1029/2007JA012308.

Emmert, J. T., B. G. Fejer, G. G. Shepherd, and B. H. Solheim (2002), Alti-tude dependence of middle and low‐latitude daytime thermospheric dis-turbance winds measured by WINDII, J. Geophys. Res., 107(A12), 1483,doi:10.1029/2002JA009646.

Fambitakoye, O., M. Menvielle, and C. Mazaudier (1990), Global distur-bance of the transient magnetic field associated with thermospheric stormwinds on March 23, 1979, J. Geophys. Res., 95(A9), 209–218,doi:10.1029/JA095iA09p15209.

Fang, T. W., A. D. Richmond, J. Y. Liu, A. Maute, C. H. Lin, C. H. Chen,and B. Harper (2008), Model simulation of the equatorial electrojet in thePeruvian and Philippine sectors, J. Atmos. Sol. Terr. Phys., 70, 2203–2211.

Fejer, B. G. (2002), Low latitude storm time ionospheric electrodynamics,J. Atmos. Sol. Terr. Phys., 64, 1401–1408.

Fejer, B. G., and L. Scherliess (1995), Time‐dependent response of equa-torial ionospheric electric fields to magnetospheric disturbances, Geo-phys. Res. Lett., 22(7), 851–854, doi:10.1029/95GL00390.

Fejer, B. G., M. F. Larsen, and D. T. Farley (1983), Equatorial disturbancedynamo electric fields, Geophys. Res. Lett., 10(7), 537–540, doi:10.1029/GL010i007p00537.

Hagan, M. E., and J. M. Forbes (2002), Migrating and nonmigrating diur-nal tides in the middle and upper atmosphere excited by troposphericlatent heat release, J. Geophys. Res., 107(D24), 4754, doi:10.1029/2001JD001236.

Hagan, M. E., and J. M. Forbes (2003), Migrating and nonmigrating semi-diurnal tides in the upper atmosphere excited by tropospheric latent heatrelease, J. Geophys. Res., 108(A2), 1062, doi:10.1029/2002JA009466.

Heelis, R. A., J. K. Lowell, and R. W. Spiro (1982), A model of the high‐latitude ionospheric convection pattern, J. Geophys. Res., 87(A8), 6339–6345, doi:10.1029/JA087iA08p06339.

Huang, C. M., A. D. Richmond, and M.‐Q. Chen (2005), Theoretical ef-fects of geomagnetic activity in low‐latitude ionospheric electric fields,J. Geophys. Res., 110, A05312, doi:10.1029/2004JA010994.

Kikuchi, T., H. Luehr, T. Kitamura, O. Saka, and K. Schlegel (1996),Direct penetration of the polar electric field to the equator during aDP2 event as detected by the auroral and equatorial magnetometer chainsand the EISCAT radar, J. Geophys. Res., 101(A8), 17,161–17,173,doi:10.1029/96JA01299.

Kobea, A. T., A. D. Richmond, B. A. Emery, C. Peymirat, H. Luehr, T.Moretto,M. Hairston, and C. Amory‐Mazaudier (2000), Electrodynamic couplingof high and low latitudes observations on May 27, 1993, J. Geophys.Res., 105(A10), 22,979–22,989, doi:10.1029/2000JA000058.

Le Huy, M., and C. Amory‐Mazaudier (2008), Planetary magnetic signa-ture of the storm wind disturbance dynamo currents: Ddyn, J. Geophys.Res., 113, A02312, doi:10.1029/2007JA012686.

Mazaudier, C., and S. V. Venkateswaran (1990), Delayed ionosphericeffects of the geomagnetic storms on March 22, 1979, studied by thesixth coordinated data analysis workshop (CDAW‐6), Ann. Geophys.,8, 511–518.

Mazaudier, C., R. Bernard, and S. V. Venkateswaran (1985), Saint‐Santinobservation of lower thermospheric storms, J. Geophys. Res., 90(A3),2885–2895, doi:10.1029/JA090iA03p02885.

Nishida, A., N. Iwasaki, and T. Nagata (1966), The origin of fluctuations inthe equatorial electrojet: A new type of geomagnetic variation, Ann. Geo-phys., 22, 478–484.

Peymirat, C., A. D. Richmond, B. A. Emery, and R. G. Roble (1998), Amagnetosphere‐thermosphere‐ionosphere‐electrodynamics general‐

Figure 7. Comparison with observations (dash‐dottedlines) of the latitudinal variations of the simulated H distur-bances associated with the disturbance wind dynamo alone(Ddyn, case 1; solid lines) and of the disturbance dynamowith penetration electric field effects (case 2; dashed lines).All values represent the difference between the disturbancedynamo day (11 June 1993) and the reference quiet day (21June 1993). The dotted line is the zero reference level.


10 of 11

circulation‐model,J. Geophys. Res. , 103(A8), 17,467–17,477,doi:10.1029/98JA01235.

Peymirat, C., A. D. Richmond, and R. G. Roble (2002), Neutral wind influ-ence on the electrodynamic coupling between the ionosphere and themagnetosphere, J. Geophys. Res., 107(A1), 1006, doi:10.1029/2001JA900106.

Richards, P. G., J. A. Fennely, and D. G. Torr (1994), EUVAC: A solarEUV flux model for aeronomic calculations, J. Geophys. Phys., 99,8981–8992.

Richmond, A. D. (1995), Ionospheric electrodynamics using magnetic apexcoordinates, J. Geomag. Geoelectr., 47, 191–212.

Richmond, A. D., and R. G. Roble (1979), Dynamic effects of aurora gen-erated gravity waves on the midlatitude ionosphere, J. Atmos. Terr.Phys., 41, 841–852.

Richmond, A. D., E. C. Ridley, and R. G. Roble (1992), A thermosphereionosphere general circulation model with coupled electrodynamics,Geophys. Res. Lett., 19(6), 601–604, doi:10.1029/92GL00401.

Richmond, A. D., C. Peymirat, and R. G. Roble (2003), Long‐lasting dis-turbances in the equatorial ionosphere electric field simulated with a cou-pled magnetosphere‐ionosphere‐thermosphere model, J. Geophys. Res.,108(A3), 1118, doi:10.1029/2002JA009758.

Roble, R. G., E. C. Ridley, A. D. Richmond, and R. E. Dickinson (1988), Acoupled thermosphere/ionosphere general circulation model, Geophys.Res. Lett., 15(12), 1325–1328, doi:10.1029/GL015i012p01325.

Sastri, H. (1988), Equatorial electric field of ionospheric disturbancedynamo origin, Ann. Geophys., 6, 635–642.

Solomon, S. W. (2006), Numerical models of the E‐region ionosphere,Adv. Space Res., 37, 1031–1037, doi:10.1016/j.asr.2005.09.040.1.

Spiro, R. W., R. A. Wolf, and B. G. Fejer (1988), Penetration of high‐latitude‐electric‐field effects to low latitudes during SUNDIAL 1984, Ann.Geophys., 6, 39–50.

Testud, J., and G. Vasseur (1969), Ondes de gravité dans la thermosphère,Ann. Geophys., 25, 525–546.

Testud, J., P. Amayenc, and M. Blanc (1975), Middle and low latitudeeffects of auroral disturbances, J. Atmos. Terr. Phys., 37, 989–1009.

Volland, H., and H. G. Mayr (1971), Response of the thermospheric den-sity to auroral heating during geomagnetic disturbances, J. Geophys.Res., 76(16), 3764–3776, doi:10.1029/JA076i016p03764.

Zaka, K. Z., A. T. Kobea, P. Assamoi, K. O. Obrou, V. Doumbia, K. Boka,B. J.‐P Adohi, and N. M. Mene (2009), Latitudinal profile of the iono-spheric disturbance dynamo magnetic signature: Comparison with theDP2 magnetic disturbance, Ann. Geophys., 27, 3523–3536.

J.‐P. Adohi, C. Amory‐Mazaudier, P. Assamoi, K. Boka, V. Doumbia,A. T. Kobea, N. M. Mene, O. K. Obrou, and K. Z. Zaka, Laboratoire dePhysique de l’Atmosphère, UFR SSMT, Université de Cocody, 22 BP582 Abidjan 22, Ivory Coast. ([email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; obrou.olivier@univ‐cocody.ci; [email protected])A. Maute and A. D. Richmond, High Altitude Observatory, National

Center for Atmospheric Research, Boulder, CO 80307, USA. ([email protected]; [email protected])


11 of 11

simulation of electric field and current during the 11...

Documents