A long-distance coincidence radio-pulse experiment

A long-distance coincidence radio-pulse experiment1 University College, Dublin, I~elancl

Can. J. Phys. Downloaded from www.nrcresearchpress.com by Hunan Normal University on 06/03/13 For personal use only.

Received June 21, 1967 Three radio receiving stations, operating at 45, 70, and 70 h1Hz center frequencies, have been set up for air shower detection. Separations are 10, 12, and 20 km between stations. Pulses are selected by combinations of antennae pointing towards magnetic west at large zenith angles. In 1 0 0 0 hours of operation, about 150 events have been observed in excess of the random expectation over the 10-km distance. Over the 12-km separation, no excess has been observed in 244 hours, but operation is restricted in this case by intervening hills to zenith angles less than 84'. Over the 20-km separation, a small excess is observed, which may be due to chance. In a series of subsidiary experiments, radio pulses have been correlated with night-sky Cerenkov detectors and a scintillation counter, and with receivers at 12, 35, and 500 Mc/s. From these experiillents the rate of detection of cosmic-ray showers at a single station is believed to be at least 1 per hour, or about 5-10% of the radio pulses selected. Local radio coincidences at individual stations are in excess of randoin expectation, and the pulseheight spectrum for local events is steeper than would be expected for cosmic-ray events or interference pulses. The long-distance coincidences have not been established dhectly as cosmic-ray events, but are consistent with this interpretation. 1. INTRODUCTION

Most of the experimental work on radio pulses from extensive air showers has been carried out in conjunction with particle-detecting arrays (Jelley et al. 1965, 1966; Porter et al. 1966; Allan and Jones 1966; Borzhkhovskii et al. 1966; Barker et al. 1967; Vernov et al. 1967). In order to utilize the inherent advantages of the technique in directionality and large collecting area, it would be desirable to operate with pure radio-pulse selection, using particle detectors solely for correlation purposes. The most serious source of large interference pulses at frequencies below 100 MHz appears to b e vehicle ignition. This is limited in general to a range of 1-2 km, so can be eliminated by spacing receivers over distances of several kilometers. Since the lateral spread of coherent radio signals from showers incident at or close to the zenith is probably less than 1 km, it is necessary to work at large zenith angles, reducing IPresented at the Tenth International Conference on Cosmic Rays, held in Calgary, June 19-30, 1967, EAS-69. This research has been sponsored in part by the Air Force Office of Scientific Research through the European Office of Aerospace Research, OAR, United States Air Force under Grant AF EOAR 67-27. Support was also provided by A.E.R.E., Hanvell, under EMR Contract No. 1561. Canadian Journal of Physics. Volume 46. S S O (1968)

the effective separation between stations in the line of sight, but retaining a large effective spacing for ground-based interference. The use of large zenith angles has the further advantage that the lateral spread of the coherent signal is increased. Neglecting curvature and particle spread, a shower of about 10IS eV at a zenith angle of 80" will produce a fully coherent radiation pool at 75 MHz 10 k m q n area and distributed over an ellipse of area 60 kin? A partially coherent signal will be produced over an even larger area; the first five Fresnel zones at this frequency have a radiation pool 50 k m q n area and distributed over an ellipse of 300 km? An additional advantage of operating at large zenith angles is the enhancement of those radiation mechanisms which depend on geomagnetic charge separation (Kahn and Lerche 1966; Colgate 1966) since the showers develop higher in the atmosphere and Coulomb scattering is less important than for vertical incidence. On the other hand, the solid angle of collection will generally be reduced to less than 1 sr. Operation near the horizon implies the detection of vertically polarized radiation only, unless the antennae are mounted at considerable heights from the ground, since the ground image cancels the antenna pattern at the horizon for horizontal polarization.

FEGAX ET

AL.: LOXG-DISTANCE

C O I S C I D E N C E RADIO-PULSE EXPERIMENT

\

Can. J. Phys. Downloaded from www.nrcresearchpress.com by Hunan Normal University on 06/03/13 For personal use only.

magnetic

Belf ield

V

70

dV

15

FIG.1. A long-distance radio coincidence experiment.

The existence of radio pulses in association with showers at large zenith angles has been demonstrated by McBreen et al. (1966) at 75 MHz, and by Charman et al. (1966) at 44 MHz. Both groups used optical night-sky Cerenkov detectors to establisll the incidence of showers. The rate of detection of radio pulses in both experiments was rather higher than would be expected from an extrapolation of effects observed near the zenith. This may indicate an enhancement of the radio signal at large zenith angles, but the nature of the dominant radiation mechanism is not yet established, and these results cannot be interpreted unambiguously as evidence in favor of geomagnetic separation mechanisms over the original charge-excess mechanism of Askar'yan ( 1962, 1965). Irrespective of the detailed

radiation mechanism, however, the night-sky correlation encourages the belief that a longdistance coincidence experiment may be possible. The system described here is intended to investigate the method further, and if possible to detect large showers. 2. EXPERIMENTAL ARRANGEMENT

Three independent receiving stations are in operation at sites south of Dublin. The general interference level and thunderstorm rates in the area are low, and in the frequency bands 3045 MHz and 6-86 MHz only low-power communications systems are operating. The stations are located at the corners of a triangle 44 k m q n area (Fig. 1), and all antennae are pointing towards geomagnetic west, close to the horizon. The equipment at the three sites

Can. J. Phys. Downloaded from www.nrcresearchpress.com by Hunan Normal University on 06/03/13 For personal use only.

S252

CtIShDI.4S JOURNAL O F PI-IYSICS. VOL. 46. 1968

is suinnlarized in Table I. The field of view between Glenculleil and Kilbride is limited by intervening hills to a maximum zenith angle of 84". The collection area for the first Fresnel zone given a shower at Z = 80" is about 10 knP'at 75 MHz; for 10 zones, corresponding to a pulse duration of about 60 ns, it would be about 100 km? The collection solid angle is determined largely by the impact parameter and the number of detectable zones, and would lie between 0.1 and 0.5 sr. The establishment of simultaneity between events at different stations presents problems, since cable links are unfeasible, and line-ofsight paths are not available. Initially, clocks were photographed at each event, and by using standard time signals for correction, an accuracy of about 2 s could be obtained. More recently, the audio wave form from a commercial radio station has been photographed on a separate oscilloscope at each station. By comparison of the three wave forms, spurious clock-coincidences can be eliminated. This method gives an effective resolving time of 0.05 s, but is not free from ambiguity in all cases. 3. PRELIbIINARY RESULTS

Data have been analyzed for night hours only, because of interference during the daytime. Parts of the system have been operated since July 1966.

(a) Lotzg-Distance Coincidences 10-km separation. Operating hours 1014. Observed clock coincidences 755; expected 24. Fine timing from random events 598 (with a restricted sample) : probable or definite events 23; expected number 1.8. 12-km separation. Operating hours 244. Observed coincidences 48; expected 52 * 7. Fine timing: observed 0; random expectation 0.05. 20-knz separation. Operating hours 160. Observed coincidences 111; expected 85 * 9. Fine timing: observed 0; random expected 0.13. A significant number of coincidences is observed over the 10-km separation, but none over the 12-km distance, which is restricted in zenith angle and also has two 70-MHz systems, whi& may reduce sensitivity. A possible effect may be present over the 20-km separation. The excess is statistical only, and individual genuine coincidences cannot be

isolated, but the pulse type which appears to be most noticeably in excess over random is a structured, relatively small, clean pulse with bandwidth-limited components and total duration up to 300 ns.

( b ) Sz~bsidiaryExperiments at Belfield The 45-MHz coincident system has been used to display a number of subsidiary detectors. ( i ) Optical night-sky Cerenkov system. In 73 hours of clear-sky operation, 80 coincident Cerenkov events were observed out of 1555 radio triggers. This implies a shower detection rate for the radio system of at least 1 shower/ hour, out of 10-20 events/hour. (ii) 70-MHz system. About two-thirds of the 45-MHz trigger pulses also show a 70-MHz component; many of these are presumably ignition interference events, but bandwidthlimited pulses are also observed. (iii) 12-MHz system. A bandwidth of only 50 KHz could be used with this receiver. In 37 hours of observation with 904 master events, 225 had pulses at 12 MHz. In 18 hours of operation with the night-sky Cerenkov system also, 305 master events gave 14 optical 45-MHz coincidences, 171 coincidences at 12 and 45 MHz, and 1 event with 45 MHz, 12 MHz, and optical events in coincidence. Since selection was in all cases on the basis of a 45-MHz signal, these results do not imply a stronger field at the higher frequency, but may imply a relatively slow falloff in pulse size with frequency. (iv) Scintillation detector. A vertical detector of 1.5 m5ensitive area has just been put into operation. In 17 hours, five coincidences were observed with the 45-MHz system, and also showed a 70-MHz signal. A U.H.F. receiver operating in this period showed no signals for these events. ( v ) U.H.F. system. Out of 547 inaster 45-MHz events, 64 coincident U.H.F. signals were obtained. In 40 hours with the scintillation detector, 2 bandwidth-limited coincident signals were observed, in coincidence with the 45-MHz system also. ( c ) Sz~bsicliaryExpe~imentsat Glencullen The optical night sky - radio correlation at 70 MHz has been described by McBreen et al. (1966). An attempt has been made to carry

Can. J. Phys. Downloaded from www.nrcresearchpress.com by Hunan Normal University on 06/03/13 For personal use only.

Lj

TABLE I Equipment in use a t the three receiving stations

S

Site Belfield (Suburban ca~rlpus; . 4uiet a t night; high building)

Glencullen (Shielded valley; galaxy limited)

Icilbride (Quiet, open site; galaxy limited)

Receiver freque~lcy Bandwidth (MHz) (MHz) Two a t 45

4

70

20

510 12

2 50 kHz

70 70 35

70

15 15 10

15

2 Resolving time (M)

A~lten~la

AGC

0.13

Rate-meterco~ltrolled

Master pulse selection

Hor. long wire

-

None

Display only

Nor. Yagi Long wire

-

Total power None

Display only Display only

None

Master pulse selection Independent correlation experiment

}

0'02

None

-

Vert. half rho~nbic

t; -

Ilor. Yagi

Vert. half rhornbic Twin helical I-lor. rhombic south ~ o i n t i n r

Particle detectors

Com~nents

-

Total power

Optical Cerenkov Scintillatio~l counter Gas Cerenkov under construction

c3

t2

C! 1.1

n

8 2 m tl

Optical Cere~lkov

Biased above None noise, single receiver selection

2

n

m U

?

$

V)

m

m X

't

m

S254

CANADIAX JOURNAL OF PHYSICS. VOL. 46, 1968

Can. J. Phys. Downloaded from www.nrcresearchpress.com by Hunan Normal University on 06/03/13 For personal use only.

out a similar experiment with a receiver at 35 2 5 MHz. These runs were inconclusive, partly because of large unidentified pulse interference, but it tvould appear, at least, that the correlation at 35 MHz is not an order of magnitude stronger than at 70 MHz.

cess of 100 MHz. The effects of the earth's electrostatic field (Charman 1967) may also be significant. We concludc that in a widely spaced system of receivers at quiet sites, operation with pure radio selection is feasible.

(d) Local Coincidences ACKNOWLEDGMENTS At both Belfield and Glencullen the rate of We would like to thank Miss M. Bolger for local coincidences between the two radio the analysis of a very large anlount ot data, receivers is far greater than random exnectaand the Irish Glass Bottle Company for the tion. The rate depends rather strongly on the gift of five hundred bottles which have been single's rate in each receiver but may be used in the construction of scintillation 10-20/hour, and the pulse height spectrum counters. for these local events appears steeper than wonld be expected from eithcr cosmic-ray REFERENCES events or locd interference. It is comnarablk ALLAN, H. R. 1967. This Conference, paper EAS-63. L with or steeper than galactic noise. Pulse ALLAN,H. R. and JONES,J. K. 1966. Nature, 212, 129 shapes are generally bandwidth-limited if not ASKAR'YAN, G. A. 1962. Sovict Phys. JETP, English clearly of interference type. The origin of Transl. 14, 441. these events remains unexplained. Their rate --- 1965. Zh. Eksperim. Teor. Fiz. 48, 988. BARKER,P. R., HAZEN,W. E., and HENDEL,A. Z. of occurrence is random in time. 1967. Phys. Rev. Letters, 18, 51. No coincident particle or night-sky events I. A., VOLOVIK,V. D., KOUIZSKOI, which are also long-distance coincidences have BOR~SHKHOVSKII, V. I., and SHMATKO,E. 1. 1966. JETP Letters, been observed, but statistically these would English Transl. 3, 118. not as yet be expected. The long-distance CEIAR~LAN, W. N. 1967. Private communication. W. N., FRUIN,J. H., JELLEY,J. V., PORTER, coincidences have not, therefore, been directly CI-IARMAN, R. A., and SMITH, F. G. 1966. A.E.R.E. Harwell established as due to air showers, but their Rept. R5322. characteristics are consistent with this inter- COLGATE,S. A. 1966. Private communication. pretation. The tl~resholdenergy, based on rate, JELLEY,J. V., CHARMAN, W. N., FRUIN,J. H., S X I I ~ , F. G., Ponmn, R. A., P o n ~ ~ nN., A., WEEKES, estimated area, and solid angle, wonld be in T.C., and MCBREEN,B. 1966. Nuovo Ciinento, the vicinity of 101g eV. No marked sidereal 46, 649. correlation has been observed, apart from that JELLEY, J. V., F R ~ I NJ., H., PORTER,N. A., WEEKES, caused by galactic variation in sensitivity. The T. C., SMITH, F. G., and PORTER,R. A. 1965. presence of significant pulse amplitudes at Nature, 205, 327. 70 MHz would seem to imply fast fronts in KAHN, F. D. and LERCHE,I. 1966. Proc. Roy. Soc. (London), Ser. A, 289,206. showers at large zenith angles, since theoretiMCBREEN, B., O'MONGAIN,E. P., PORTER, N. A., cal models (Colgate 1966; Allan 1967) predict and SLEW, P. J. 1966. Phys. Letters, 23, 677. a rapid falloff at high frequencies for the main PORTER, N. A., LONG,C. D., MCBREEN,B., ?dURelectron component. At these zenith angles NAGHAN,D. J. B., and WEEKES,T. C. 1966. Proc. Intern. Conf. Cosmic Rays, London, 2, the /.L mesons may be separated geomagneti706. cally into two parallel showers, which would VERNOV, S. N., ABROSIMOV, A. T., VOLOVIK,V. D., radiate independently by the Askar'yan i k ~ ~ u n o v s V. ~ ~ I., r , and KHNSTIANSEN,G. B. mechanism, possibly up to frequencies in ex1967. JETP Letters, English Transl. 5, 126. 0

I