Geoelectric power spectra over oceanic distances

GEOPHYSICAL RESEARCH LETTERS, VOL. 22, NO. 4, PAGES 421-424, FEBRUARY

Geoelectric

15, 1995

power spectra over oceanic distances

IkukoFujii•'•, L. J. Lanzerotti •, H. Utada• , H. Kinoshita1 J Kasahara1 L. V. Medford•, C. G. Maclennan • Abstract. We present initial results of simultaneous measurementsof the geoelectricpotential as measured

turbances

observed

on in-service

telecommunications

acrosstwo long cable routes in the Pacific (Guam to Philippinesand Californiato Hawaii) separatedby ~6h

Graves,1873; Axe, 1968; Runcorn, 1954, 1964; Medford et al., 1989; historicalreview by Lanzerotti and Gregori,

andtelegraphcables[e.g.,Barlow,1849;Prescott,1866;

in local time and ~30 ø in latitude. Over the fre1986],as well as on shorterlengthpowersystemcables quencyrange~10 -5 to 1.67x 10-3 Hz the powerlevels (see,e.g.,a recentreviewby ViljanenandPitjoia [1994]

and referencestherein). Recently, with the decommissioningof older, lowchannel capacity analog telecommunicationscables, it levelsup to 104 (mV/km)2/km were measuredon has becomepossibleto use some of these for geophysa geomagneticallydisturbed day. The power levels ical researchas well as to obtain new engineeringdata at the longest periods, ~17m to ~150m, are gen- for long conductingsystems.A discussionof the use of erally found to be largest along the route at lower ocean cables for researchpurposesis contained in Melgeomagneticlatitudes (Guam to Philippines) when oni et a/.[1983].Resultsin magnetospheric/ionospheric, tend to be enhanced near local noon along the two routes, especially during geomagneticquiet intervals. During the seven days analyzed, geoelectric power

comparisonsare made at equivalent local times. In con- oceanic, and solid Earth studies are included in sevtrast, the powerlevelsfor periodsin the hydromagnetic eral publications overthe last decade[e.g.,Medfordet

waveregion(periods•5m) tendto be largerat all local al., 1982; Lanzerotti et al., 1986, 1993; Thomson et al., times for the higher geomagneticlatitude route. The 1986;Chaveet al., 1992;and references therein]. This paper describes some initial results obtained nightsidegeoelectricpower levelsalong the higher latitude route tend to increasewith increasinggeomagnetic from simultaneousmeasurements of the geoelectricpoactivity. tential over large (1000's of km) distancesusingtwo separate Pacific cables. This is the first time that data from such cables are simultaneouslyavailable for reIntroduction search purposes. The nature of the spectralpowerof geoelectricfluctuations over large distancesis of intrinsic scientificinterest as well as of interest for more practical concernsrelated to the installation and maintenanceof technological systemscomposedof long conductorsgroundedto the Earth. While powerspectralstudiesof geomagnetic field fluctuations have been investigatedon a 'global' scale, using spacedmagnetometers,for more than two

Cable Systems and Measurements

The measurementsreported herein were made on two decommissionedand unpowered Pacific cables whose mid-points are separated by about 6h in local time and ~30 ø in geomagneticlatitude. The HAW-1 cable spansa distanceof 4050 km from Point Arena, CA, to decades[seeespeciallyCampbell,1976a,b,1977;Wertz Hanauma Bay, HI. The GP cable is a section of the and Campbell,1976], geoelectricmeasurements have decommissionedTPC-1 cable that spans a distance of generallybeenconfinedto more localand regionalstud- 2716 km from Guam to Baler, Luzon in the Philippines. ies, primarily associatedwith magnetotelluricstudiesof The locationsof the cablesare shownin geomagnetic coordinatesin Figure 1. The mid-point of the HAW-1 the Earth's conductivity structure. There have been relatively few scientific studies of cable has local time (LT) - UT- ~9h; the mid-point global scalegeoelectricpotentials. This is largely be- of the GP cable is at a LT - UT + ~9h

The HAW-1 cable is groundedin Hawaii using the standardtelecommunicationsgroundbed locatedunder large-scale(hundredsto thousandsof km) geoelectric sea water in Hanauma Bay. At Point Arena, a separate measurementshavebeenreportedfollowingvoltagedis- telecommunicationsgroundrod is installedfor the measurements.An AT2•T computer-baseddata acquisition systemrecordsthe geoelectricpotential betweenHI and CA at 2s intervals, together with local measurements EarthquakeResearchInstitute, Universityof Tokyo,Japan

causeof the costthat wouldbe involvedin establishing

such investigations. Therefore, most of the reported

of the threecomponents of the geomagnetic field [see, e.g., Lanzerottiet al., 1993]. The GP cablehas been

2AT&T Bell Laboratories, Murray Hill, NJ

Copyright1995 by the AmericanGeophysicalUnion.

groundedto the old TPC-1 ground bed in the Philippines;the ground bed at Guam is that presentlyused

Paper number 94GL03282

for the active cables at that telecommunications

0094-8534/95/94GL-03282503.00

The geoelectricpotential is measuredat 3s intervals. 421

station.

422

FUJII

ET AL.: GEOELECTRIC

POWER

.,j

OVER

OCEANIC

DISTANCES

responses.The signalson both cablesare more variable during 19 - 20 January,when the geomagneticactivity waslarger. The Sq daily minima in the potentialsoccur earlier in UT, as they should, along the HAW-1 route than along the GP route on days suchas January 22, 23, 25 when they are reasonablyevident visually.

60 ø

LUZONqc-•--•

SPECTRA

•

----_. 0 o

Power Spectral Results

;Figure 1, ;Routesof the two cab]esshow• •n geomagnetic coordinates.

The original data are first treated with a low passfilter with a 90s cutoff and then each time series is resam-

pled at 30s intervals. The upper two panelsof Figure 2 show the resampled geopotentialsmeasuredon the two cablesfor a 7d interval in January 1994. The lowest panel containsthe 3h planetary Kp indices(solid) and the 3h K indicesfrom College(dashed),an auroral zone station at approximatelythe same local time as the HAW-1 cable (to provide a measureof higher latitude geomagneticactivity). The most striking differencesin the two recordsare that the near-equatorial GP potentials exhibit much clearer diurnal variations than do the HAW-1 potentials. The GP diurnal variations during this week in austral winter are also generally larger in amplitude than the HAW-1

diurnal

variations.

In addition

to

Power spectra were computed for both cable routes for daily time seriesas well as for six hour intervals throughout the 7d interval analyzed. Four prolate data windows were applied to the individual time series prior to usinga fast Fourier transform(FFT) al-

gorithm[Thomson,1982].The time-bandwidth for the ld and 6h spectraare 9.259x 10-5Hz and 3.7037x 10-a Hz, respectively. Shown in Figure 3 are power spectra of the geopo-

tentialoverthe frequency range10-5 - 1.6x10-3Hzfor GP (solidlines) and HAW-1 (dashedlines)for four individual days, one geomagneticallydisturbed(January 19) and three relatively quiet. The spectra are red, but are not strict power laws. The spectra have quite similar shapes.They tend to roll over to lesssteepfre-

quencydependencies at frequencies •10 -3 Hz. At the longer periods the spectral power levels can be larger acrossthe near-equatorialcablethan they are alongthe HAW-1

route.

The computed 6h power spectra provide a measure lower amplitude Sq ionospherevariations in northern of the daily variation of the geopotential power levwinter conditions, the lessclear diurnal variation across els while at the same time preservinginformation on the HAW-1 route is principallybecauseof a very large the power levelsof the longer-periodfluctuations. Plotocean tide effect there. This is evident as a strong ted in Figure 4 are the time seriesof the power levels beating signal (period ..•15d) betweenthe ocean tidal

signal (M2; 12.4h period) and the Sq signal (12h period) [Lanzerottiet al., 1994].A visualinspection of the contoursof M2 ocean tidal amplitudes in the model of

for five differentperiods(frequencies).Top to bottom, they correspondto periodsof 170m, 45m, 17m, 5m, and

2.5m. Error bars are shown for each power level point. The two shortest periods fall into the hydromagnetic

$chwiderski [1980]suggests that the two cablesshould wave regime.

have at least qualitative differencesin their ocean tidal

le+05

Jan.22, Ap

2

7

N

•'

1e+00

-2 2

'-

le-05 le+05

0

'•""-..•..• Jan.25, Ap =4 .....,,.,,,,,,•J Ap =5

-2

• 1e+00 0 1/19

1/20

1/21

1/22

1/23

1/24

1/25

days

Figure 2. Resampledgeopotentialdata from the two cable routes. The bottom panel containsthe 3h plan-

etary Kp indices(diamonds;solidlines)and the 3h K indicesfrom College(stars;dashedlines}.

1 e-05

le-05

le-04

le-03

le-02

Frequency(Hz)

1 e-05 1 e-04 1 e-03 le-02

Frequency(Hz)

Figure 3. One day power spectra from the HAW-1 (dashedlines) and the GP (solid lines) cableroutesfor four days in January 1994. Daily geomagneticactivity indexesAp are shown.

FUJII

ET AL'

GEOELECTP•IC

POWER

SPECTRA

OVER

OCEANIC

DISTANCES

oftenquitesimilaroverthe frequencyrange,,,10-5 to ,,,1.67x10-3 Hz. The powerlevelsvary with localtime

le+12

and geomagneticactivity, with a local noon enhancement often observed,particularly during geomagnetic quiet times. At the lowest frequenciesthe power levels during geomagneticallydisturbed times can reach

N le+08

E le+o4

valuesOfthe orderof 104(mV/km)2/Hz.

E

i e-04

1/19

1/20

1/21

1/22

1/23

1/24

1/25

days

Figure 4. Geomagnetic power levels from 6h power spectra for five different periodsfor the two cable routes.

Comparisonsof the long distancegeopotentialpower levels at several different frequenciesindicate that the levels are 'controlled' by different physical processesat different frequencies.The power levelsat the longer periods of 17m, 45m, and 150m tend to be higher for the cable route G P, closerto the geomagneticequator. At periods of 2.5m and 5.0m, in the hydromagneticwave

region,the powerlevelstend to be higheralong HAW-1 at the highergeomagneticlatitudes. Theseobservations routes. The dual setsof time series,top to bottom, cortend to be independent of local time. Thus, the lowest respond to periodsof 170m(valuesx10S), 45m(x106), frequencypower levelsare controlledby fluctuationsin 17m(x104), 5m (x102), and2.5m. the dayside ionosphericcurrent system, including, particularly in the case of the GP route, the equatorial The geopotential variations in power level from the electrojet. The shorter period, hydromagnetic wave GP route(solidcurves)showa distinctdiurnalvariation region power is controlled by dayside magnetosphere for all five frequencies,with the largest power tending sources(i.e., the power is higher on the geomagnetic to occurduring the 6h interval00 - 05 UT (09 - 14 LT, field linesthat extend further into the magnetosphere). at the centerof the cable), especiallyfor the five days Such sourcesare associatedwith HM wave activity profollowingthe relatively intensegeomagneticactivity on duced by solar wind flow around the dayside magneto19- 20 January. sphereand the penetration of the producedenergydeep The powerlevelsfrom the HAW-1 route at the longest into the magnetosphere [e.g, Wolfeet al., 1989,1994]. period tend to be enhanced in the 6h interval 18-23 UT Plotted in Figure 5 are the power levelsfor the 45m (09- 16LT). TheseHAW-1 powerenhancements leadin (open diamonds) and 2.5m (starred points) periods UT those measuredon the GP cable at this frequency, from the GP and HAW-1 routes as a function of the and showthat the largest power level is associatedwith College geomagneticactivity K index. The K values local daytime ionosphericand magnetosphericcurrent shownwere constructedfrom simpleaveragesof the two systemsduringgeomagnetically quiettimes. 3h index values that correspondto the same UT interWhile the geopotential power levelsfor the 45m and vals as for the cable data. The College data used in the 17m periods on the GP cable appear to have their the comparisonwith the GP cable correspondsto local maximum levels during the 09 - 14 LT interval (es- morning activity at College;these indexeswere used as pecially during the five relatively geomagneticallyquiet similar data from Russia are not yet available. days),the localdaytimeenhancements in HAW-1 power at these two periods tend to be broader for this more GP •t le+02 northerly cable, although there are instanceswhen the power as measuredon this cable is largest around local

The solid(dashed)linescorrespond to the GP (HAW-l)

,

,

N

noon(e.g.,23, 24, 25 Januaryat 17mperiod). The daily variations in the HAW-1 power at 45m are lessstructured than in the longestand shortestperiod bands. For the 17m period, the local daytime power levels

on the GP route are larger on two of the days (22 and 23 January UT) than they were during the two daysof highergeomagneticactivity (19 and 20 January UT). This observationdemonstrates clearlythe importance of local daytime plasma processes,probably the ionosphericequatorialelectrojet,in producingthe nearequatorial geopotentialvariations over large distances. The geoelectricpower levelsat the two periods that correspondto the hydromagneticwave regime (2.5m and 5.0m) tend to be larger over the higher latitude cable route, independent of local time. Discussion

E

1e+00

• le-02

le-04 o

HAW-1

le+02 N

?1e+00 E

• le-02

le-04 o

I

2

3

4

5

6

averagedK indices

Figure 5. Local night geomagneticpower levelsfor two The power spectra of the geoelectricpotential mea- periods (45m: diamonds;2.5m: stars) for each cable suredsimultaneouslyacrosstwo long cable routesshow route as a function of geomagneticactivity level at the that the overall spectralshapesat the two locationsare Collegemagneticobservatory(the CollegeK index).

424

FUJII ET AL.: GEOELECTRIC

POWER SPECTRA OVER OCEANIC DISTANCES

The local night power for both periods along the HAW-1 route increaseswith increasingauroral zoneactivity, with the largest effect occuringfor the longer period. An increasein powerwith increasingactivity is also seenfor the longer period power on the GP route. However,the HM powerlevels(2.5mperiod)on the GP route do not showa relationshipto the geomagneticactivity levels.

Lanzerotti, L. J., A.D. Chave, C. H. Sayres,L. V. Medford, and C. G. Maclennan,Large-scaleelectricfield measurements on the Earth's surface: a review, J. Geophys.Res., 98, 23525, 1993.

Lanzerotti, L. J., C. H. Sayres,L. V. Medford,J. S. Kraus, C. G. Maclennan, and D. J. Thomson, Statistical examination of inducedvoltagesacrosslongoceanictelecommunicationscables,in proceedingsof 1992 Solar-Terrestrial Predictions Workshop, 1, 224, 1994.

Medford,L. V., A. Meloni, L. J. Lanzerotti, and G. P. Gregori, Geomagneticinductionon a transatlantictelecom-

Acknowledgments. We thank AT&T cable station munications cable, Nature, 290, 329, 1982. personnelin Point Arena and in Guam and AT&T InterMedford, L. V., L. J. Lanzerotti, J. S. Kraus, and C. G. national personnel in Morristown, NJ, for their assistance Maclennan,Transatlantic earth potential variationsdurin making this researchpossible.We also thank P. Hattori, ing the March 1989 magneticstorm, Geophys.Res. Left., Guam GeomagneticObservatory,for assistance.The geo16, 1145, 1989. magnetic activity information was obtained from the PreMeloni,A., L. J. Lanzerotti,and G. P. Gregori,Inductionof liminary Report and Forecast Data reports of the NOAA currentsin longsubmarinecablesby natural phenomena, Space Environment ServicesCenter. Rev., Geophys.,SpacePhys., 21, 795, 1983. Prescott,G. B., History, Theory, and Practiseof the Electric References Telegraph,IVed., Tichnor and Fields, Boston,1866. Runcorn,S. K., The earth's core, Eos 7•rans.A GU, 35, 49,

Axe, G.A., The effectsof Earth's magnetismon submarine cables,Electr. Eng. J., 61, 37, 1968. Barlow, W. H., On the spontaneouselectrical currents observedin the wires of the electric telegraph, Phil. 7•rans. R. Soc., 61A, 61, 1849. Campbell, W.H., Spatial distribution of the geomagnetic spectral compositionfor disturbed days, J. Geomag. Geoelectr., 28, 481, 1976a. Campbell, W.H., An analysisof the spectraof geomagnetic variations having periods from 5 min to 4 hours, J. Geophys. Res., 81, 1369, 1976b. Campbell, W.H., Spectral characteristicsof geomagnetic field variations at low and equatorial latitudes, J. A tmos. Terr. Phys., 39, 1217, 1977. Chave, A.D., D. S. Luther, L. J. Lanzerotti, and L. V. Medford, Geoelectric field measurementson a planetary scale: oceanographicand geophysicalapplications, Geophys. Res. Lett., 19, 1411, 1992. Graves, J., On Earth currents, J. Telegr. Engr., 2, 102, 1873.

Lanzerotti, L. J., L. V. Medford, C. G. Maclennan, D. J. Thomson,A. Meloni, and G. P. Gregori, Measurementsof the large-scaledirect-current earth potential and possible implications for the geomagneticdynamo, Science, 229, 47, 1985. Lanzerotti, L. J. and G. P. Gregori, Telluric currents:The natural environment and interactions with man-made systems, in The Earth's Electrical Environment, National Academy Press,Washington,D.C., p.232, 1986. Lanzerotti, L. J., D. J. Thomson, A. Meloni, L. V. Medford, and C. G. Maclennan, Electromagneticstudy of the Atlantic continental margin using a sectionof transatlantic cable, J. Geophys. Res., 91, 7417, 1986. Lanzerotti, L. J., C. H. Sayres, L. V. Medford, J. S. Kraus and C. G. Maclennan, Earth potential over 4000km between Hawaii and California, Geophys.Res. Left., 19, 1177, 1992.

1954.

Runcorn, S. K., Measurementsof planetary electriccurrents, Nature, 202, 10, 1964.

Schwiderski,E. W., On charting global ocean tides, Rev., Geophys.,SpacePhys., 18, 243, 1980.

Thomson,D. J., SpectrumEstimation and HarmonicAnalysis, Proc., IEEE, 70, 1055, 1982.

Thomson, D. J., L. J. Lanzerotti, L. V. Medford, C. G. Maclennan,A. Meloni, and G. P. Gregori, Study of tidal periodicitiesusinga transatlantictelecommunications cable, Geophys.Res. Left., 13, 525, 1986. Viljanen, A., and R. Pirjola, Geomagneticallyinducedcurrents in the Finnish high-voltagepower system: A geophysicalreview Surv. Geophys.,15, 383, 1994. Wertz, R., and W. H. Campbell, Integrated power spectra of geomagneticfield variationswith periodsof 0.3 - 300s, J. Geophys.Res., 81, 5131, 1976. Wolfe, A., C. Uberoi, C. T. Russell,L. J. Lanzerotti, C. G. Maclennan, and L. V. Medford, Planet. SpaceSci., 37, 1317, 1989. Wolfe, A., R. D. Kelman, S. E. Warren, C. G. Maclennan, and L. J. Lanzerotti, Hydromagnetic frequency spectra in the high latitude quiet magnetosphere,in Solar Wind Sourcesof Magnetospheric Ultra-Low-Frequency Waves,, Am. Geophys. Union, Washington, 375, 1994.

I. Fujii, J. Kasahara, H. Kinoshita, and H. Utada, Earthquake Reseas-ch Institute, The University of Tokyo,

Tokyo113, Japan (fujii•utada-sun.eri.u-tokyo. ac.jp). L. J. Lanzerotti, C. G. Maclennan, and L. V. Medford, AT&T Bell Laboratories, Murray Hill, New Jersey 07974

( ljl•physics.art.corn). (receivedJune 6, 1994; revisedAugust 22, 1994; acceptedNovember17, 1994.)