Simultaneous observations of plasmaspheric and ionosphericvariations during magnetic storms in 2011: First resultfrom Chinese Meridian Project

Chi Wang,1,2 Qingmei Zhang,2,1 P. J. Chi,3 and Chuanqi Li4

Received 21 May 2012; revised 29 October 2012; accepted 2 November 2012; published 15 January 2013.

[1] The plasma transport between the plasmasphere and ionosphere during magneticstorms is a long-standing problem and is still not fully understood. Simultaneousobservations of the plasmasphere and ionosphere are vital to understand the couplingbetween the two regions. In this study, using the measurements from the newly developedChinese ground-based space weather monitoring network (Meridian Project), weinvestigate the plasmaspheric density at L’ 2 inferred from ground magnetometers and theionospheric electron density inferred by digital ionosondes and GPS signals duringmagnetic storms in 2011. Five moderate magnetic storms with minimum Dst indexbetween �47 and �103 nT during this period have been investigated. The observationsshow that the plasmaspheric density drops significantly by more than half of theprestorm value. The ionospheric F2 layer electron density NmF2 and the total electroncontent (TEC) show ~20–50% decreases, and the NmF2 and TEC reductions take placebefore the plasmaspheric density reaches its minimum. These findings suggest that theplasmaspheric depletion is very likely due to the reduced plasma supply from the ionospherefor the five moderate magnetic storms in 2011. Therefore, the plasmasphere dynamicsseems to be controlled by the ionosphere during magnetic storms.

Citation: Wang, C., Q. Zhang, P. J. Chi, and C. Li (2013), Simultaneous observations of plasmaspheric and ionosphericvariations during magnetic storms in 2011: First result from Chinese Meridian Project, J. Geophys. Res. Space Physics,118, 99–104, doi:10.1029/2012JA017967.

1. Introduction

[2] Geospace, defined by the extent of the terrestrial mag-netic field into space, includes the closely coupled regions ofthe middle-upper atmosphere (the thermosphere, ionosphere,and magnetosphere) and their interactions with the loweratmosphere. These regions are characterized by dynamicprocesses known as space weather which links solar activ-ity to the changes in the near-Earth space environment.[3] An important process in geospace that is still not fully

understood is the plasma transport between the plasmasphereand the ionosphere during magnetic storms. Specifically, pre-vious ground-based observations of ducted whistlers andfield line resonance (FLR) have shown the internal depletion

of the plasmasphere during magnetic storms [e.g., Park,1973;Chi et al., 2000]. Different from the erosion of the plas-masphere, this plasmaspheric depletion occurs within theeroded plasmasphere during magnetic storms, and the plasmadensity in the depleted region can drop by a factor of 4. Theplasmaspheric depletion was often found to be concurrentwith an ionospheric storm by which the ionosphere under-went significant changes. For example, Villante et al.[2006] used the FLR remote sensing technique to reveal asignificant reduction (by a factor of 1.6–2) in the plasma-sphere density at L = 1.7–1.8 during the recovery phase ofa geomagnetic storm. This decrease was accompanied by asignificant negative ionospheric storm phase in which thecritical frequency of the F2 layer (foF2) is depressed belowits median value and was confirmed by vertical totalelectron content (TEC) measurements. They pointed outthat each storm event has its own individual manifesta-tion because of the complex nonlinear interactions of the con-tributing processes and the large variety of the initial andboundary conditions. Even during magnetic quiet time, thereexists a connection between ionospheric and plasmasphericdensity variations. Clilverd et al. [1991], using the whistlermode group delays from very low frequency Doppler exper-iment, showed an annual variation in the equatorial electrondensity in the plasmasphere, which has a maximum inDecember and a minimum in June/July (maximum-minimumratio of about 1.7). This annual variation in equatorial

1State Key Laboratory of Space Weather, Center for Space Science andApplied Research, Chinese Academy of Sciences, Beijing, China.

2College of Math and Physics, Nanjing University of InformationScience and Technology, Nanjing, China.

3Institute of Geophysics and Planetary Physics, University of California,Los Angeles, CA, USA.

4Guangxi Normal University, College of Electronic Engineering,Guilin, China.

Corresponding author: C. Wang, State Key Laboratory of SpaceWeather, Center for Space Science and Applied Research, ChineseAcademy of Sciences, Beijing 100080, China. (cw@spaceweather.ac.cn)

©2012. American Geophysical Union. All Rights Reserved.2169-9380/13/2012JA017967

99

JOURNAL OF GEOPHYSICAL RESEARCH: SPACE PHYSICS, VOL. 118, 99–104, doi:10.1029/2012JA017967, 2013

plasmaspheric electron density (Neq) can be modeled fromthe combined foF2 medians at each end of the field line by as-suming that diffusive equilibrium is maintained from the F2layer to the equator over long (>1 month) time scales. How-ever, the cause of internal plasmaspheric depletion duringmagnetic storms is still uncertain. It could be due to thedownward flux to the ionosphere [Park, 1973] or the reducedplasma supply from the ionosphere during negative iono-spheric storms [Chi et al., 2000]. The negative ionosphericstorms are likely caused by the composition changes of theneutral atmosphere, especially by the atomic oxygen to mo-lecular nitrogen concentration ratio ([O/N2]) that alters thebalance of the production and loss processes of the ionizedplasma. Additionally, the vertical air motion can also changethe [O/N2] ratio and further affect the negative ionosphericstorm [Rishbeth, 1998]. To help answer this question, moresimultaneous observations of the plasmasphere and the iono-sphere are needed to understand the coupling between thetwo regions.[4] In order to monitor the geospace environment, the

Meridian Space Weather Monitoring Project (or ChineseMeridian Project) [Wang, 2011] operates a chain of 15ground-based observatories located roughly along the120�E longitude meridian and the 30�N latitude. Asidefrom the station in Antarctica, the stations are locatedroughly 4–5� of latitude or about 500 km apart. Each ob-servatory is equipped with multiple instruments to measurekey parameters such as the baseline and time-varying geo-magnetic field, as well as the middle and upper atmo-sphere and ionosphere from about 20 to 1000 km. Thisproject started test observations in 2011 and began normaloperation after the middle of 2012. In this study, we willtake advantage of the simultaneous observations from theground magnetometer, ionosonde, and ionospheric TECmonitor in Mohe station (MHT) of this project to investi-gate the plasmaspheric and ionospheric variations duringmagnetic storms in 2011.[5] In this paper, we examine the variations of the plasma-

spheric density using the gradient technique [Chi et al.,2000]. The phase-difference method was first introducedby Kurchashov et al. [1987] and then was first used byWaters et al. [1991]. The phase difference of field line reso-nances observed by two nearby stations along the same lon-gitude maximizes at the eigenfrequencies of the field linemidway between them. Stations from Chinese MeridianProject are located in the middle to low latitudes with thegeographic latitudes ranging from 18.3� to 53.5�. It is foundthat the first harmonic of field line resonances is in the Pc 3–4 frequency range at mid- and low-latitude, and the gradienttechnique can produce clear phase-difference patterns ateigenfrequencies [Waters et al., 1991, 1994]. If the dipolemagnetic field and plasma density models (r� r�m, wherer is the geocentric distance) are assumed, we can infer the

plasma mass density from the observed eigenfrequency.Similarly, the plasma mass density has been studied byground-based ULF magnetic field measurements [e.g. Menket al., 1999; Clilverd et al., 2003; Kawano et al., 2002;Berube et al., 2003].[6] In the same time, the digisonde and TEC monitor give

information about the ionosphere. The foF2 observed by thedigisonde gives the peak of the ionospheric electron densityin F2 layer (NmF2). The profile of the F2 layer is needed toconvert the foF2 value into the TEC value, which is not easyto know exactly in most cases. Note that the F2 layer maynot be described by a simple Chapman profile in the daytimeclose to solar maximum [Rishbeth and Garriott, 1969].Therefore, real-time ionospheric TEC and scintillation moni-tors deployed in the meridian chain are common instrumentsto infer TEC by GPS signals.[7] The remainder of the paper proceeds as follows. We

describe the data used in the study and present the observa-tions in section 2, and the discussion and summary is givenin section 3.

2. Observations

2.1. Data

[8] The data of ground magnetometer stations MHT andManZhouLi (MZL) are available from the data center ofthe Chinese Meridian Project. Fluxgate magnetometers datasampled at 1 Hz are used in this study. The noise level of thesystems is about 0.1 nT. The magnetic latitudes of MHT andMZL are 48.6� and 44.9�, respectively, with L’ 2 (Table 1).The fluxgate magnetometers at these stations record the geo-magnetic H, D, and Z components. The digisonde at MHTgives 15 min time resolution of the foF2 values in a standardformat, and the TEC monitor provides 30 s time-resolutiondata. We search the Dst index in 2011 available from theWorld Data Center for Geomagnetism (http://wdc.kugi.kyoto-u.ac.jp/index.html) for magnetic storms with Dst <�50 nT [Gonzalez et al., 1994]. However, the Chinese Me-ridian Project did not provide full data coverage for all thesemagnetic storms, since it was in its test phase in 2011. In theend, five magnetic storms with relatively good data coveragehave been selected, which took place on 6 April, 12 April,10 September, 17 September, and 26 September 2011, re-spectively. For each event, we infer the plasmaspheric den-sity from the ground magnetometer observations by usingthe gradient technique and compare them with the iono-spheric electron density variations.

2.2. A Typical Example

[9] On 26 September 2011, an interplanetary shock ar-rived at the Earth’s magnetosphere at 1144 UT and causedthe sudden commencement (SC) monitored by ground mag-netometer stations. A moderate magnetic storm took placeafterward and reached its maximum (the Dst index reachedits minimum of �103 nT) near the end of the day. We takethis event as an example to give details of the data analysisprocess. The phase differences between H component ofthe magnetic fields at MHT and MZL were used to distin-guish FLR events and identify the eigenfrequencies of themagnetospheric field line midway between their latitudes.Four days of phase-difference spectrograms are given inFigure 1. All spectrograms present the time interval

Table 1. Locations of Stationsa

StationsMagneticLatitude

MagneticLongitude

GeographicLatitude

GeographicLongitude L Value

MHT 48.6 195.9 53.5 122.4 2.29MZL 44.9 191.2 49.6 117.5 1.99

aThe resonance point halfway between MHT and MZL has an L valueequal to 2.14. MHT, Mohe station; MZL, ManZhouLi station.

WANG ET AL.: PLASMASPHERE AND IONOSPHERE

100

between 0000 and 1200 UT, which corresponds to the lo-cal time interval 0800–2000, since the phase-differencepatterns are usually seen on the dayside. The bright colorcorresponds to larger phase differences. FLR signals canbe seen in a short time of the period on 0000–1000 UT(0800–1800 LT) every day.[10] On September 26, the FLR signatures appear obvi-

ously, and the fundamental mode frequency at about30 mHz can be clearly identified over the interval 0000–1000 UT in Figure 1. Additionally, it is slightly higher inthe morning than in the afternoon from 0000 to 1000 UT.This kind of diurnal variation of the FLR frequency is typi-cal at low-middle latitudes [Waters et al., 1994]. On 27September, the eigenfrequency at 32 mHz at 0700 UT(1500 LT) is a little bit higher than 30 mHz at the samehour on the previous day. From 0600 to 0800 UT,phase-difference signatures are observed at the high fre-quencies 60–90 mHz, which corresponds to higher FLRharmonics. On 28 September, the phase-difference patternis identified during 0500–0800 UT and is not as clear asthat on the previous 2 days, indicating much weaker fieldline resonance. The corresponding fundamental mode fre-quency is about 35–50 mHz, and it is 42 mHz at 0700 UT(1500 LT). On September 29, the eigenfrequency is higherthan that on the first 2 days and is identified at about47 mHz at 0700 UT (1500 LT). On the following daysof 30 September to 2 October (not shown here), the fun-damental mode frequency is identified clearly, and the

Figure 1. Phase-difference spectrograms of BH for the station pair Mohe and ManZhouLi stations(MHT-MZL) during the time period of the 26 September 2011 magnetic storm event. Each diagram showsthe spectrogram for the local time interval 0800–2000.

−100−50

050

Dst

(nT

)

2030405060

Fre

quen

cy(m

Hz)

10002000300040005000

Den

sity

(am

u/cm

3 )

0

5

10

15

Nm

F2

(105 /c

m3 )

9/26 9/27 9/28 9/29 9/30 10/1 10/2 10/30

10

20

30

Days in 2011

TE

C(1

012el

/cm

2 )

Figure 2. From the top to bottom panels, shown are Dst in-dex, fundamental mode frequencies of field line resonancesat L’ 2 for the local time 1500, equatorial plasma mass den-sities, the ionospheric F2 layer electron density valuesNmF2, and the total electron content (TEC) values duringthe time period of the 26 September 2011 magnetic stormevent. The solid and dashed lines in the last panel indicatethe prestorm and the minimum TEC daily peak values,respectively.

WANG ET AL.: PLASMASPHERE AND IONOSPHERE

101

eigenfrequency declines slowly and is close to the pre-storm value at 0700 UT (1500 LT).[11] Several more days of phase-difference spectrograms

have been examined, and the results of eigenfrequency mea-surements at L ’ 2 for the local time 1500 LT (0700 UT) aregiven in the second panel of Figure 2 as the same timeframes as the Dst index (top panel). The error bar indicatesthe standard deviation of all the eigenfrequency values iden-tified during the day. The eigenfrequency increases withtime during the SC, and at the beginning of the recoveryphase the eigenfrequency is slightly larger than the prestormvalue, reaches a maximum on September 29, and thendecreases closely to the prestorm value during the recoveryphase of the magnetic storm dates from 29 September. Sincethe equatorial plasma mass density was derived from theeigenfrequency values, the third panel of Figure 2 plots theequatorial plasma mass density at L ’ 2 for the local time1500 LT (0700 UT). The plasma mass densities were calcu-lated by using the formula given by Schulz [1996], in whicha dipole magnetic field and the plasma density r� r�m areassumed, where r is the geocentric distance. In particular,for this range of latitudes (1.8<L< 2.2), an index m in therange 0–2 may be appropriate [Chi et al., 2005;Vellanteand Förster, 2006], and the inferred equatorial r value is

not very sensitive to the particular choice of m within thisrange, so we choose the index m = 1 in our calculation.Figure 2 shows a reduction of the plasma density whenthe main phase of the magnetic storm was over, whichis about 12% less than prestorm value and even furtherdecreases by 58% to the lowest value later in the recoveryphase, then the density gradually increases and recovers,and it takes about 2 days to return to the prestorm value.[12] The decreases of the NmF2 and TEC peak values of

the ionosphere are significant during the main phase of thestorm. The bottom two panels show the NmF2 value fromthe digisonde and the vertical TEC value from GPS mea-surements above MHT during this magnetic storm event.Both parameters can be used as a proxy for the electron den-sity in the ionosphere. In addition to the typical diurnal var-iation of NmF2 and TEC, which present the highest valuesin the early afternoon and the lowest values just before sun-rise, a significant drop of the daily NmF2 and TEC peaks byapproximately 56% and 36%, respectively, which occurredon 27 September 2011 during the magnetic storm. However,unlike the plasmaspheric density at L ’ 2, the daily NmF2and TEC maxima rose to slightly higher than 98% and 93%of the prestorm value on the next day. After 28 September,the NmF2 and TEC values recover to the prestorm values.One of the most striking features is that the NmF2 and TECreductions take place before the plasmaspheric densityreaches its minimum. These results are consistent with thefindings by Chi et al. [2000].

2.3. Statistical Results

[13] We apply the same approach to the other four mag-netic storms in 2011. First of all, it is noted that the iono-spheric electron density reduction indicated by the NmF2and TEC reductions happened before plasmaspheric densityreached its minimum as well for three other magnetic storms(i.e., the 6 April, 12 April, and 10 September events) and al-most on the same day for only one case (the 17 Septemberevent). The former three cases followed the similar variationpattern as the above 26 September storm event. The latterone is plotted in Figure 3, which has the same format asFigure 2.[14] The main results about the plasmasphere and iono-

sphere variation (i.e., density decrease) are summarized inTable 2. The columns give the date, the hour of the magneticstorm Dst reached its minimum value, the largest plasma-spheric density reduction percentage from the prestormvalue, the delay time (in days), the largest drop of the dailyNmF2 and TEC peak values from the prestorm peak values,and the delay time (in days) as well. For moderate magneticstorms with Dst > �100 nT investigated here, it seems thatthe decreases of the plasmaspheric density at L’ 2, the daily

−50

0

50

2030405060

10002000300040005000

0

5

10

15

9/16 9/17 9/18 9/19 9/20 9/21 9/22 9/230

10

20

30

Days in 2011

Dst

(nT

)F

requ

ency

(mH

z)D

ensi

ty(a

mu/

cm3 )

Nm

F2

(105 /c

m3 )

TE

C(1

012el

/cm

2 )

Figure 3. From the top to bottom panels, shown are Dst in-dex, fundamental mode frequencies of field line resonancesat L’ 2 for the local time 1500, equatorial plasma mass den-sities, the ionospheric F2 layer electron density valuesNmF2, and the total electron content (TEC) values duringthe time period of the 17 September 2011 magnetic stormevent. The solid and dashed lines in the last panel indicatethe prestorm and the minimum TEC daily peak values,respectively.

Table 2. Summary of the Plasmaspheric and Ionospheric Variations During Magnetic Stormsa

No. Date Time (UT) (Dst)min Density Reduction D1 NmF2 Reduction TEC Reduction D2

1 26 Sep 2400 �103 nT 58% 3 56% 36% 12 17 Sep 1600 �63 nT 58% 2 41% 22% 23 10 Sep 0500 �64 nT 45% 3 42% 17% 24 12 Apr 1000 �47 nT 36% 2 55% 46% 15 6 Apr 2000 �61 nT 49% 2 19% 21% 1

aD1 and D2 represent the delay time in days when the plasmaspheric density, NmF2, and TEC got the largest reduction after theDst reached the minimum value, respectively.

WANG ET AL.: PLASMASPHERE AND IONOSPHERE

102

NmF2 peak value, and TEC peak value do not depend on thestrength of magnetic storms, which is indicated by the min-imum value of the Dst index. However, we need to investi-gate more magnetic storm events with different intensity tomake reliable conclusions in the future.

3. Discussion and Summary

[15] In this paper, we estimated the plasmaspheric densityby applying the gradient technique to the H component mag-netometer data from two ground stations of the ChineseMeridian Project, in which the difference in geomagneticlatitude is 3.7�. However, this is slightly larger than whatis recommended (1�2�) for an accurate inference of theplasmaspheric density. The large difference in geomag-netic latitude of the two stations may be a source of erroreven though the exact amount of uncertainty is hard to es-timate. Kawano et al. [2002] found that the resonancewidth is nearly constant in the geomagnetic latitude rangeof 28� to 32�, and its magnitude �4� � 400 km. For theevent of 26 September 2011, it was estimated that the res-onance width is about 3� for the frequency of 32 mHz,which is close to the difference in geomagnetic latitudeof the two stations. Nevertheless, the clear FLR signaturesduring magnetic storms indicate that the method is stillvalid.[16] As mentioned above, the equatorial plasmaspheric

densities increases during the recovery phase to the pre-stormvalue after decreasing with the main phase for moderate mag-netic storms in 2011 investigated here. The results show theplasma depletion and refilling of the plasmasphere duringmagnetic storms, as pointed out by Chi et al. [2000]. Toroughly estimate the location of the plasmasphere, we em-ploy the empirical formula of the plasmapause location Lppfrom spacecraft and whistler observations at higher L shells[Carpenter and Anderson, 1992]. The plasmapause locationLpp was reduced from 5.1 to 2.8 on the 26 September 2011event in response to an increase of Kp up to 6+ at 18–21 UTon that day (Kp from 1+ to 6+). This implies that the plasma-pause never came inside of L ’ 2 for this moderate magneticstorm event. In this case, the plasmasphere depletion is un-likely due to the enhanced convection which drives plasmasunward. Instead, it is a result of the coupling with the ion-osphere. In the recovery phase, the plasmasphere is filledat a similar rate, which is closely related with the plasmaion of the ionosphere outflow into the newly formed plas-masphere. For the case of the 26 September event, it wasestimated that the mean value of the depletion rate ofthe plasmasphere was about 31 amu cm�3 hr�1 duringthe interval between 26 and 29 September, and the plas-masphere was refilled at a rate of 47 amu cm�3 hr�1 at theequator after it was drained during the interval between 29September and 1 October. However, there was also a diurnalvariation of the plasmaspheric density associated with ionoutflow from the ionosphere on the dayside and loss of parti-cles from the flux tubes in the nightside [Chi et al., 2000]. Forexample, it can be estimated that the refilling rate was ~210amu cm�3 hr�1 during the interval of 0400–0800 UT on 27September after the TEC value decreased rapidly on 26–27September. The NmF2 and TEC reductions take place 1 or2 days before the plasmaspheric density reaches its minimumin four of the five magnetic storms, and in the case of

exception the density reductions in the ionosphere and inthe plasmasphere occurred on the same day. The resultsstrongly imply that the plasmaspheric depletion is due tothe reduced plasma supply from the ionosphere.[17] The comparison of the results from the five magnetic

storm events in 2011 as shown in Table 2 implies thatdecreases of the plasmaspheric density at L ’ 2, the dailyNmF2 peak value, and TEC peak value seem to be not af-fected by the strength of the moderate magnetic storms.The percentage decreases of the NmF2 and TEC peak valuesare generally lower than that of the plasmaspheric density inthe main phase during magnetic storms. These results sug-gest that a minor change in the ionospheric content may re-sult in a significant impact on the density in the plasma-sphere. However, the results from these five moderatemagnetic storms in 2011 do not lead to a general conclusionin a statistical sense. More investigations with magneticstorms of different intensities are needed to establish a reli-able quantitative relationship between the magnetic stormstrength, the plasmaspheric density variation, and the iono-spheric density variation.[18] In summary, the plasmaspheric density at L ’ 2 and

ionospheric electron density variations during five moderatemagnetic storms in 2011 are investigated using the recentobservations of Chinese Meridian Project during its first yearof test operation. The observations show the plasmasphericdensity drops significantly by about half of the prestormvalues. At the same time, the ionospheric F2 layer electrondensity NmF2 and the TEC also show sizeable decreases.There is no clear relationship between the magnitude of iono-spheric depletion and plasmaspheric depletion. For the fivemoderate magnetic storms in 2011, it is suggested that theplasmaspheric depletion is very likely due to the reducedplasma supply from the ionosphere. The accurate statementof the plasmaspheric and ionospheric densities can be usedto further explore the dynamics of the plasmasphere andionosphere and their coupling. The multiple parametersobserved by the Chinese Meridian Project give us a uniqueopportunity to simultaneously examine the changes amongdifferent coupled regions in geospace.

[19] Acknowledgments. We acknowledge the use of data fromChinese Meridian Project and the Dst index data provided by the WorldData Center for Geomagnetism. Q.M. Zhang would like to thank H. Shen forhelping with TEC calculations. This work was supported by grants 973 pro-gram (2012CB825600), NNSFC (40921063, 41231067) and in part by theSpecialized Research Fund for State Key Laboratories of China.

ReferencesBerube, D., M. B. Moldwin, and J. M. Weygand (2003), An automatedmethod for the detection of field line resonance frequencies using groundmagnetometer techniques, J. Geophys. Res., 108(A9), 1348, doi:10.1029/2002JA009737.

Carpenter, D. L., and R. R. Anderson (1992), An ISEE/whistler model ofequatorial electron density in the magnetosphere, J. Geophys. Res., 97,1097–1108.

Chi, P. J., et al. (2000), Plasmaspheric depletion and refilling associatedwith the September 25, 1998 magnetic storm observed by groundmagnetometers at L = 2, Geophys. Res. Lett., 27, 633–636.

Chi, P. J., C. T. Russell, J. C. Foster, M. B. Moldwin, M. J. Engebretson,and I. R. Mann (2005), Density enhancement in plasmasphere-ionosphereplasma during the 2003 Halloween Superstorm: Observations along the330th magnetic meridian in North America, Geophys. Res. Lett., 32,L03S07, doi:10.1029/2004GL021722.

Clilverd, M. A., F. W. Menk, and G. Milinevski (2003), In situ andground-based intercalibration measurements of plasma density at L =2.5, J. Geophys. Res., 108, 1365, doi:10.1029/2003JA009866.

WANG ET AL.: PLASMASPHERE AND IONOSPHERE

103

Clilverd, M. A., A. J. Smith, and N. R. Thomson (1991), The annual varia-tion in quiet time plasmaspheric electron density, determined from whis-tler model group delays, Planet Space Sci. 39, 7, 1059–1067.

Gonzalez, W. D., J. A. Joselyn, Y. Kamide, H. W. Kroehl, G. Rostoker,B. T. Tsurutani, and V. M. Vasyliunas (1994), What is a geomagneticstorm? J. Geophys. Res., 99, 5771–5792, doi:10.1029/93JA02867.

Menk, F. W., D. Orr, M. A. Clilverd, A. J. Smith, C. L. Waters, D. K.Milling, and B. J. Fraser (1999), Monitoring spatial and temporal var-iations in the dayside plasmasphere using geomagnetic field line reso-nances, J. Geophys. Res., 104, 19,955–19,969.

Kawano, H., K. Yumoto, V. A. Pilipenko, Y. M. Tanaka, S. Takasaki, M.Iizima, and M. Seto (2002), Using two ground stations to identify magne-tospheric field line eigenfrequency as a continuous function of ground lat-itude, J. Geophys. Res., 107, doi: 10.1029/2001JA000274.

Kurchashov, I. P., I. S. Nikomarov, V. Pilipenko, and A. Best (1987), Fieldline resonance effects in local meridional structure of mid-latitudegeomagnetic pulsations, Ann. Geophys., 5A, 147–154.Park, C. G. (1973), Whistler observations of the depletion of theplasmasphere during a magnetospheric substorm, J. Geophys. Res., 78,672–683.

Rishbeth, H., and O. K. Garriott (1969), Introduction to IonosphericPhysics, New York and London: Academic Press.

Rishbeth, H. (1998), How the thermospheric circulation affects theionospheric F2-layer, J. Atmos. Solar-Terr. Phys., 60, 1385–1402.

Schulz, M. (1996), Eigenfrequencies of geomagnetic field lines and implica-tions for plasma-density modeling, J. Geophys. Res., 101, 17,385–17,397.

Vellante, M., andM. Förster (2006), Inference of the magnetospheric plasmamass density from field line resonances: A test using a plasmaspheremodel, J. Geophys. Res., 111, A11204, doi:10.1029/2005JA011588

Villante, U., et al. (2006), ULF fluctuations of the geomagnetic field andionospheric sounding measurements at low latitudes during the firstCAWSES campaign, Ann. Geophys., 24, 1455–1468.

Waters, C. L., F. W. Menk, and B. J. Fraser (1991), The resonance structureof low latitude Pc3 geomagnetic pulsations, Geophys. Res. Lett., 18,2293–2296, doi:10.1029/91GL02550.

Waters, C. L., F. W. Menk, and B. J. Fraser (1994), Low latitude geomag-netic field line resonance: Experiment and modeling, J. Geophys. Res.,99, 17,547–17,558.

Wang, C. (2011), New chains of space weather monitoring stations inChina, Space Weather, 8, S08001, doi:10.1029/2010SW000603.

WANG ET AL.: PLASMASPHERE AND IONOSPHERE

104