Apr 1, 2000 - from L= 1.3-2.0 was used to investigate the variation in Pc3-4 power with ... properties of low-latitude Pc3-4 (-7-100 mHz) geomagnetic pul-.
JOURNAL
OF GEOPHYSICAL
RESEARCH,
VOL. 105, NO. A4, PAGES 7747-7761, APRIL 1, 2000
Field line resonancesand waveguidemodesat low latitudes 1. Observations F. W. Menk, C. L. Waters, and B. J. Fraser Department of Physics andCooperative Research Centerfor SatelliteSystems, University of Newcastle, Callaghan New South Wales, Australia
Abstract. Field line resonances (FLRs) arean importantmechanism for the generationof Pc3-4 (-7-100 mHz) geomagneticpulsations.There is considerableobservationalevidencefor the existenceof FLRs at middle latitudes,bothin satelliteandgrounddata. However,the low-latitude regionsare lessaccessible for suchstudies,andconsequently manyaspectsof low-latitudeFLRs are not well understood.A temporary12-stationmagnet,meterarrayspanningeasternAustralia from L= 1.3-2.0 wasusedto investigatethe variationin Pc3-4 powerwith latitude,the natureand low-latitudelimit of FLRs, andpropertiesof spectralcomponents belowthe localresonant frequency.Examplesarepresentedfor representative days. Powerspectraareremarkablysimilar overthisrangeof latitudesandoftenexhibita multitudeof peaksseparated by -3-5 mHz. Using cross-phase techniques,we find thattheresonantfrequencyincreases with decreasing latitudeto L-1.6, thendecreases at lowerlatitudes.This is dueto theeffectof ionospheric heavyionsat low altitudes. The characteristicsize of the resonancesis L-0.15, the resonanceQ is -2 at L=2.0 and 1.3-1.4 at L= 1.3, and the normalizeddampingfactor¾/C0R-0.2-0.4.The low-latitudedetection limit of FLRs dependson a numberof factors,but on a day examinedin detail it wasL-1.4. For signalsbelow the local resonantfrequency,amplitudedecreasedwith latitudeat-3 dB/0.1 L. Interstationphasedelaysare not consistent with the time of flight of radiallypropagatingfastmodewavesin the equatorialplane,althougha peakoccursin the regionwherethe Alfv6n velocitypeaks. We concludethat theseresultsare consistenteither with modulationof the incomingfast-modewavesor the existenceof cavity or waveguidemodeswhichdrive discrete forcedoscillationsof low-latitudefield linesacrossa rangeof frequencies, andwhichcoupleto the local FLR wherethe frequenciesmatch.
1. Introduction
frequencymatchesthe field line eigenfrequency resonances are
The objectiveof this studyis to experimentallyinvestigatethe propertiesof low-latitudePc3-4 (-7-100 mHz) geomagneticpulsations.Specifictopicsinclude(1) the variationof the pulsation spectrumwith latitude; (2) the identificationof field line resonances (FLRs) at low latitudes, and their low-latitude limit; (3) the variationin resonantfrequencywith latitude;(4) propertiesof the resonance, such as the characteristic width, Q value and dampingfactor;(5) whetherthereis evidenceof discretefrequencies in the spectrum;and (6) the role of upstreamwave sources and cavity or waveguide modes in determiningthe low-latitude Pc3-4 spectrum. A companionpaper [Waters et al., this issue, hereafterdenoted paper 2] presentsmathematicalmodeling of waveguidemode waves in the magnet,spherewhich explains
boundary, theionosphere. However, atlowlatitudes a significant proportion of a field line's lengthlies withinthe ionosphere. Massloadingdueto ionospheric heavyionslowerstheeigenfre-
established. This relies on the existenceof a suitablelow altitude
features of these observations.
It is generallybelievedthat ULF wavespropagatethroughthe magnet,sphereand couple to and drive standingoscillationsof field lines [Orr, 1984; Allan and Poulter, 1992]. Satellite observationsdemonstratethat the inward propagatingwaves are most likely fast compressionalmode waves [e.g., Yumotoand Saito, 1983; Yumoto et al., 1985; Odera et al., 1991; Takahashi andAn-
derson,1992; Takahashiet al., 1994]. When the incomingwave
Copyright2000 by the AmericanGeophysical Union. Papernumber1999JA900268. 0148-0227/00/1999JA900268509.00
quency,so that a frequencymaximumoccursaround30ølatitude [Sutcliffeet al., 1987;Poulteret al., 1988]. The low-latitudelimit of FLRs hasnot yet beenestablished.
The low-altituderegionsare difficultto monitorwith spacecraft,whoserapidmotioncausesspectralbroadening andphase shear[Andersonet al., 1989]. Thereforeexperimental evidence for theexistence of FLRsat low latitudescomesprincipallyfrom ground observations. These have been basedon studiesof interstationpower,phaseandpolarization[Yumoto,1986;Ziesollecket
al., 1993], and cross-phase [Waterset al., 1991a, 1994; Greenet
al., 1993]. The amplitudeandfrequency of Pc3-4magnetic pulsationsat the groundmaybe correlated with solarwind parameters [Greenstadtet al., 1979; Yumotoet al., 1984; Vet6, 1986; Le
and Russell,1996; Vet6 et al., 1998], whichaccordingly is believed to be the sourceof the compressional wavesdrivingthe resonances.Undersuitableconditionsthesewavesmay alsoexecutestandingoscillationsin the magnet,sphericcavities[Kivelson and Southwood, 1985;Allan et al., 1985, 1986a,b], resultingin discretefrequencies in thespectrum.Discrete,latitude-dependent FLRs with stablefrequencies havebeenreportedin magnet,meter and radar data at high latitudesand interpretedas evidenceof magnet,sphericULF waveguide modes [Samsonet al., 1991, 1992]. Theseresultsare still controversial[e.g., Ziesolleckand McDiarmid, 1994;Rickardand Wright,1995].
7747
7748
MENK
ET AL.: OBSERVATIONS
OF LOW-LATITUDE
ULF RESONANCES
Table 1. NamesandCoordinates of StationsUsedin This Study Geographic
Geomagnetic
Longitude
L Shell
Station
Code
Latitude
Longitude
Latitude
Ingham
ING
-18.7
146.0
-28.0
219.6
1.28
Charters Towers
CHA
-20.1
146.3
-29.5
220.1
1.32
Safina
SAR
-21.4
149.1
-30.6
223.5
1.35
Blackwater
BLA
-23.5
149.1
-32.9
223.8
1.42
Munduberra
MUN
-25.6
151.6
-34.8
227.0
1.49
Dalby
DAL
-27.2
151.2
-36.6
226.9
1.56
Glen Innes
GLE
-29.7
151.7
-39.3
228.0
1.68
Armidale
ARM
-30.5
151.6
-40.1
228.1
1.72
Muswellbrook
MUS
-32.3
150.5
-42.2
227.2
1.83
Kulnura
KUL
-33.4
151.1
-43.3
228.2
1.90
Canberra
CAN
-35.3
149.4
-45.6
226.7
2.04
Cooma
COO
-36.2
149.3
-46.6
226.8
2.12
Whereasmostpreviousstudieshave focusedon FLRs, at low latitudessignificantpoweroccursat frequencies belowresonance [Kurchashovet al., 1987; Waterset al., 1991a,b; Ostwaldet al., 1993]. Detailed investigationof thesesignalsusing multistation groundarrayscan provideinformationon propagationcharacteristicsof the incomingwaves[e.g., Matsuokaet al., 1997], suchas whether they are traveling fast-modewaves or whether cavity resonances or waveguidemodesare excited.
2.2. FLR
Detection
The first stepin data analysisinvolvedexaminingwhole-day time seriesplots, power spectra, cross-powerspectraand crossphasespectrafrom the stationpairs to determinethe presenceof FLRs. IntervalsexhibitingPc3-4 activity (not necessarilyat the same frequency)jointly over the entire station array were then selectedfor detailedanalysis. This involvedinspectionof the H componentcross-power,coherence,power difference,power ratio, and cross-phase spectrabetweenstationpairs. In orderto in2. Data Recording and Analysis crease the number of measurementpoints, combinationsof nonadjacentstation pairs were also used. For comparisonwith 2.1. Magnetometer Array anothermethodof detectingFLRs, the ratio of the H and D comIn orderto investigatethe aboveissuesa temporary12-station ponentspectrawas alsoexaminedat selectedstations. magnetometerarray was establishedand operatedat low latitudes The variationin field line eigenfrequencywith latitudemay be in easternAustralia. The arrayspannedthe 226ø magneticme- exploited to determine the resonant frequency approximately ridian (i.e., a meridional array) from L=1.28 to L=2.12. Station midwaybetweentwo closelyspacedrecordingstations.A numlocationsare shownin Table 1 and Figure 1. The magnetometers ber of techniqueshave been described[Baranksyet al., 1985, were deployedin closelyspacedpairsto facilitateresonancede- 1989; Waters et al., 1991a, 1994, 1995; Menk et al., 1994] and tection using cross-phaseand allied techniques[Waters et al., werereviewedby Pilipenkoand Federov[1994]. Briefly, the si1991a, 1994]. multaneoussignaturesof a field line resonancebetweena poleInstrumentationat each site compriseddigital inductionmag- ward (P) andequatorward(E) stationare (1) a peakin the H comnetometersfeaturing spectral correction over the 10-100 mHz ponentcross-phase spectrum,A4)E_p; (2) a zerocrossing in the H •angc anu c•o•c•y matched ampntud t•cqucucy,and pl,a•c componentamplitude(or power)subtraction spectrum,AHp_E; (3) sponse. The error in phasematchingbetweeninstrumentswas a unity crossingin the H componentamplitude(or power) ratio •1
•1
•
•1
....
.•1..
....
1;...
A
1• .....
typicallyof order+12ø. The magneticnorth-south (H) component spectrum,Hp/HE;and (4) a dip in the coherencespectrum.Expe-
wasrecorded ateachsiteandtheeast-west (D) component atone rience hasshown thatnosingle criterion unambiguously identifies station intwo.Dataweresampled onsiteat0.5Hzwitha resolu-theresonance, andthatdifferent methods mayindicate slightly tionof 0.03-0.1 nT. Details oftheinstrumentation appear in the different resonant frequencies. Previous cross-phase studies used work
of Fraser et al. [ 1991] and also at stationseparations of order0.6ø at low latitudes[Waterset al., http://plasma.newcastle.edu.au/spwg/. At eachsite,timingwas 1991a]andup to 2.5ø at high latitudes[Waterset al., 1995]. In referenced via shortwaveradioto standard time andtYequencythis study,stationswere separatedby -1 øin latitude,but it was
transmitters, with an errortypicallyof
