damage. Since the UGO normally (in the field) will op-erate from its own power source, voltage transients suchas this will not be a problem.

For programs that require a number of sensors or ahigh data rate, the UGO represents a significant advancein capabilities over existing automatic stations such ashave been used for ice and ocean buoys, since the UGOcan provide a continuous, highly reliable data link carry-ing in excess of 5,000 bits per second directly from theAntarctic to the United States. This amount of data isroughly equivalent to the amount of information carriedby 65 teletype machines operating simultaneously. In ad-dition, the data link is not affected by ionospheric dis-turbances, as shown by the uninterrupted flow of dataduring the intense geomagnetic storm of August 4-7,1972.

Because of its high data rate capability, the UGO willbe a key element in a proposed network for observingthe spatial and temporal distribution of geophysicalevents (National Academy of Sciences, in press), whichis expected to greatly increase scientific productivity. Al-though this network was conceived largely to meet theneeds in ionospheric and magnetospheric sciences, theUGO is sufficiently flexible to permit its use by other dis-ciplines for automated data collection either as part of orseparate from the network. Alternative designs for auto-matic stations that will not have the capability to usethe INTELSAT satellite system but will be somewhat lessexpensive are being investigated (Sites, in press).

Electron precipitation at middle latitudes

T. J . ROSENBERGInstitute for Fluid Dynamics and Applied Mathematics

University of Maryland

Particle precipitation from the magnetosphere is onemanifestation of a complex global phenomenon calledthe magnetosphere substorm. Most studies of substormprecipitation have been concentrated at high latitudes(;.e., in the auroral zones and polar caps) where dis-turbances are usually most pronounced. However, in re-cent years it has become apparent that particle precipita-tion at middle latitudes must be considered an importantaspect of the substorm phenomenon and the cause ofsignificant ionospheric effects.

Few locations in the world are so well-suited to thestudy of the mid-latitude magnetosphere as is Siple Sta-tion, Antarctica, and its geomagnetic conjugate nearRoberval, Quebec, Canada. These sites, located at 60°geomagnetic latitude, border on the nominal position ofthe plasmapause, an inner-magnetosphere plasma bound-ary that plays a major role in controlling the dynamicsof the magnetosphere. Several articles in this journal dur-ing the past 3 years have described the facilities for re-search at Siple and Roberval, and have given brief ac-counts of results obtained. In this note, some results ofthe University of Maryland program are summarizedand plans for continued studies are described.

References

National Academy of Sciences, Committee on Polar Research.In press. On the Use of Unmanned Geophysical Observatoriesin the Antarctic: A Proposed Program in Upper AtmosphericPhysics.

Sites, M. J . In press. Automatic stations for Antarctica: a reviewof the unmanned geophysical observatory project and somesuggestions for implementing a multidisciplinary program.Stanford University Report.

Sites, M. J . 1972. Geophysical data transmission from auto-matic stations in the Antarctic via earth synchronous satellites.1972 IEEE National Telecommunications Conference. Hous-ton, Texas. P. 31E-i.

July-August 1973

9000

700Cn

500I-300

0° 100

jLUIr

1 VLF SPECTRUM

kHzJZ

- -

0........-.--._----.....-- .....IIIIII89202122

SECONDSFigure 1. X-ray bursts and corresponding VLF data recorded at

Siple Station at about 0936 Universal Time on January 2, 1971.

179

JANUARY 2, 19710935 - 0937 UT 1.0

.5

-I

-0.5

r ROBERVAL

NORMALIZEDCROSSCOVARIANCE

2,3'4 5 6 7

LAG, SECONDS

Figure 2. Normalized cross-covariance of X-ray and VLFintensities as a function ofthe time lag of the VLF inten-sity with respect to the X-raycount rate. Maximum correla-tion corresponds to VLF emis-

10 II 12sions leading X-ray bursts by0.3-0.4 seconds.

ELECTRON S

NZLi"̂ WHIST- INTERACTION

REGION

EMISSION +WHISTLER ')

S

Figure 3. Schematic drawingof the mechanism postulatedto explain the association be-tween X-ray bursts and dis-

crete VLF emissions.

SIPLE STATIONThe steep plasma density gradient, characteristic of

the plasmapause, is expected to lead to important plasmaphysics phenomena involving wave-particle interactions.Balloon-borne X-ray detectors were flown from Siple Sta-tion during the 1970-1971 austral summer to examineelectron precipitation near the plasmapause and its rela-tionship to the generation of VLF and ULF wave dis-turbances. This was a coordinated experiment also in-volving groups from Stanford University and BellLaboratories.

Confirmation of the occurrence of one type of wave-particle interaction (cyclotron resonance) was obtainedduring a balloon flight on January 2, 1971. An exampleof the data is shown in fig. 1. Bursts of X-rays with en-ergies greater than 30 keY at the balloon were correlatedwith bandlimited VLF emissions in the frequency range1-to-4 kHz measured on the ground. Evidence that thecorrelated bursts were triggered by whistlers from light-fling discharges in the northern hemisphere was also ob-tained. The high degree of correlation between X-raysand emissions was typical of the majority of bursts ob-

served in an event lasting more than one-half hour. Thisis illustrated in fig. 2 by the normalized crosscovarianceperformed on two minutes of data. The time delay atmaximum correlation, 0.3-0.4 sec, indicated that thearrival of the VLF emissions at Siple preceded that of theenergetic electrons producing the X-rays. Informationfrom these records has led to the interpretation of theobservations sketched in fig. 3. A whistler initiated inthe northern hemisphere interacts with energetic elec-trons near the equatorial plane producing a burst of VLFemissions. The cyclotron resonance condition for this in-teraction requires that the resonant electrons and VLFwaves be oppositely directed along the magnetic fieldline. Consequently, the VLF emission reaches Siple afterhaving traveled half the length of the field line, whereasthe resonant electrons reach Siple after mirroring abovethe atmosphere in the north, having traveled one andone half times the length of the field line. Theoreticalestimates of the time delay between the arrival of wavesand electrons at Siple agree well with the experimentaldata. Further description of this event can be found in

180 ANTARCTIC JOURNAL

Rosenberg et al. (1971), and a final report is in prog-ress.

The wave-particle event described above is part of acomplex substorm involving many other ionosphere andmagnetosphere phenomena. Efforts are being made todevelop a unified interpretation of the many diverse dis-turbances that were recorded (see, for example, Carpen-ter et al., 1973).

Also being studied for the possible relationship toelectron precipitation at middle latitudes are perturba-tions to the phase and amplitude of long-distance sub-ionosphere VLF transmissions. Perturbations on night-time transmission paths well within the plasmaspherewere observed during the electron precipitation and VLF

emission event on January 2, 1971, as shown in fig. 4.The onset of the VLF phase disturbance on the NLK-APL

path at 0820 Universal Time occurred simultaneouslywith the enhancement of VLF emissions recorded at Sipleand the expansion phase of the associated magnetic sub-storm. Enhanced X-rays were not detected until 25minutes later, but further analysis of this event as de-scribed in Potemra and Rosenberg (1973) indicated thatweak fluxes of greater than 40 keV electrons, below thesensitivity threshold for detection by the X-ray method,could still account for the phase perturbations. The rela-tionship of phase perturbations to polar magnetic activity

LU0708>ci20_____

<wJ0—J U)

cc

44I 0 7

T

and the use of VLF transmissions as a highly sensitiveindicator of magnetospheric disturbance are being stud-ied as an outgrowth of Siple Station research.

Results from the work at Siple Station on wave-par-ticle interactions make it clear that particle precipitationmeasurements must also be made in the vicinity of theSiple conjugate point. From the asymmetry in conjugatemagnetic field strengths above the ionosphere at Sipleand Roberval, one is led to expect asymmetries in theflux of precipitated electrons at the conjugate points. Bycoordinating balloon flights with the operation of theSiple VLF transmitter, it should be possible to studythese effects under controlled conditions. The first co-ordinated effort, in which balloons will be launchedfrom the Roberval area, is planned for July 1973. Dur-ing the 1973-1974 austral summer, balloons will beflown from Siple and, as it is presently planned, simul-taneously from Roberval as well.

This research is being supported by National ScienceFoundation grant GV-28841x2.

References

Carpenter, D. L., L. J . Lanzerotti, and T. J. Rosenberg. 1973.A middle-latitude view of the substorm; conjugate wavephenomena, convection E fields, and particle precipitationeffects within and beyond the plasmasphere. EOS, 54(4):113.

09101112

N LK—APL

GROWTH EXP.2ND EXP.PHASE PHASE PHASE

ww—D

Cr

Figure 4. Relative nighttimephase of the 18.6 KHz NLK(Jim Creek, Washington) toAPI (Applied Physics Labora-tory, Maryland) transmissionand the Siple Station record-ings of relative VLF emissionsintensity (0.1-10 KHz) and

X-ray count rate.

VLFSIPLE STATION

X-RAYS,400— E> 30 keySIPLE STATION..

200 =0_111-0809101112

UT, JANUARY 2, 1971Johns Hopkins University

0wU)

U)

z

00

July-August 1973 181

Potemra, T. A., and T. J. Rosenberg. 1973. VLF propagationdisturbances and electron precipitation at mid-latitudes.Journal of Geophysical Research, 78: 1572.

Rosenberg, T. J . , R. A. Helliwell, and J. P. Katsufrakis. 1971.Electron precipitation associated with discrete very-low-fre-quency emissions. Journal of Geophysical Research, 76: 8445.

Atmospheric particulates atSouth Pole Station

JOHN M. ONDOV, ERNEST S. GLADNEY andWILLIAM H. ZOLLER

Department of ChemistryUniversity of Maryland

ROBERT A. DUCE

Graduate School of OceanographyUniversity of Rhode Island

ALUN G. JONES

Shields Warren Radiation LaboratoryHarvard Medical School

Samples of atmospheric particulate matter have beencollected at Pole Station over the past 3 years. Duringthe first year, 1970-1971, samples were collected at bothMcMurdo and Pole Stations for several months. Pre-liminary results of these measurements of the gaseousand particulate halogens and of the metals vanadium,aluminum, and manganese were reported by Duce et al.(1971) and Gladney et al. (1972). The final analysisof these samples has been completed by non-destructiveneutron activation analysis and atomic absorption toyield the atmospheric concentrations of more than 20elements at Pole Station. The interpretation of the par-ticulate and gaseous halogen portion of the data hasbeen reported elsewhere (Duce et al., 1973). Aspreviously stated (Gladney et al., 1972), most of theMcMurdo samples were contaminated by wind-blowndust and, consequently, are of only local interest. ThePole samples originally were collected by 8- by 10-inchDelbag polystyrene filters and both 47-mm and 90-mmMillipore EA filters. The sample-to-blank ratio of theDelbag filters was considerably better than that of theMillipore filters for all elements except zinc and chlo-rine. Owing to the extremely low concentrations en-countered in the polar atmosphere, the blank usually ac-counted for 10 to 90 percent of the total elemental con-centration in the samples when up to 10,000 standardcubic meters of air were filtered by the high-volumepumps.

Interpretation of earlier data was aided by the compu-tation of the ratios of the various atmospheric elemental

concentrations to that of atmospheric aluminum. Theseatmospheric ratios were then compared with similar ra-tios in the earth's crust, as described by Gladney et al.(1972) and Zoller et al. (1973), to determine the crus-tal contribution to the aerosol for the various elements.In this work we have facilitated this comparison by com-puting the enrichment factor, E.F., relative to averagecrustal rock defined as follows:

E.F. - 'V--/Alati,,

crust/Alerustwhere Xatni and Xerugt refer to the concentrations of anelement in the atmosphere and in the crust, respectively.This E.F. is approximately I for elements in atmosphericparticles largely derived from crustal material and isgreater or less than one if the element is enriched or de-pletei with respect to aluminum, relative to the crust.

Based on the mean trace element distribution in theearth's crust as reported by Taylor (1964), our resultsfrom Pole Station fall into three groups as shown in thetable. Except for such clearly marine related elements asNa, Mg, Ca, and K, the sources for the enriched ele-ments, i.e., groups II and III, are unknown. Possiblesources for these elements include volcanism, the ocean,and pollution, particularly sources producing small par-ticles such as oil combustion and motor vehicle operation.

Severe problems with pumps, filter holders, and thelevel of impurities in the filters hampered the first year'swork and caused large analytical uncertainties in themeasurement of certain elements. During the past 2years new pumps, filter holders, and filter materials havebeen evaluated at Pole Station for the intense samplingeffort in 1973-1974. The work has indicated that 4-inchfilters prbvide the optimum sample to blank ratio, whenused with an improved carbon vacuum pump.

Since very low levels of contamination can com-pletely overwhelm the natural level of many trace ele-ments in the samples, clean areas for the collection andhandling of the samples are essential. An 8- by 8- by 12-foot building has been built for sampling at Pole Sta-tion, which is planned to begin again in November 1973.The building is equipped with a clean bench so that fil-ters may be handled without contamination. Air isbrought into the building through an 18-foot-high, 12-inch diameter rvc pipe that connects into a manifoldcapable of handling nine experiments simultaneously. In-line filter holders will be used to collect particles with thepumps exhausting the filtered air directly outside. Otherprograms such as the collection of gaseous mercury,halogens, and chlorinated and non-chlorinated hydro-carbons will be undertaken, partly in cooperation withother research groups. The station will contain a wind di-rectional control system to allow automatic sample col-lection only when the wind is blowing from a prede-termined favorable sector relative to local camp con-tamination sources. A recording condensation nucleus

182 ANTARCTIC JOURNAL