SUGAR

SUGAR - Tony Parkinson's contributions

First some context; C.B.A McCusker and Murray Winn were the architects of the SUGAR experiment, and many contributers followed over the years. The description of the experiment will be left to others. In brief, it is an EAS array of buried scintillators sensitive to the muon component of ultra-high energy cosmic rays. Michael Ryan was the first PhD candidate concerned mainly with the Pilot Array on the SydUni campus. He was followed by Leo Goorevich and Tony Parkinson, who both dealt with the preliminary data from the Narrabri array. Leo determined the muon shower core location and the shower size and estimated primary energy. I dealt with the curvature of the muon shower front and the arrival directions. I will only describe my contributions and leave it to Michael and Leo and those following, in order to build up a history of the experiment.

I spent the years 1964-1969 as a PhD student on SUGAR, plus a later reprise. On the hardware side, I designed, built and tested all the data handling electronics for the array. At each station an inexpensive tape recorder was purpose-built in the workshop. These consisted of a stepping motor capstan driving a standard quarter inch audio tape on a 7 inch reel. The recording used a pulse width modulation system and each event was recorded in triplicate to provide redundancy.

The data tapes, removed from the stations for processing, required much more complex equipment. The tapes were played back on standard audio tape decks. Eventually four pieces of hardware were constructed, and Derek whimsically called them "Tonoscopes" and the name stuck as it's convenient! In Mark I, the first stage was the strobing of bits to produce a result in ternary logic, consisting of zero, one or an error, and including a thorough check on the expected data format. The basis of this system was a unique phase-locked oscillator which could respond to changes over a period as short as a bit, in order to overcome the jitter produced by the very simple recording system, and to read tapes in which bits had dropped out. A voltage produced from this oscillator was used to control the strobing process to overcome the speed variations. Thanks to Murray Winn for the initial idea. The rest of the equipment consisted of a magnetic core memory, scalers and control logic, and the results were read out in various formats to a printer or paper tape punch. With the triplicate redundancy of the recordings, a majority decision could be made in the electronics. The Mark I version was eventually installed in the service truck at Narrabri.

In the Mark II version similar operations occurred, but at a much faster rate using a standard tape deck operating at 15 ips. This equipment was plugged directly into an asynchronous port on the KDF9 computer and run off-line, and was used continuously for all the SUGAR data up to ca 1974, at an average of ca 25 hours per week.

During 1973-1974 I returned to this experiment to help prepare for its transfer to using the then-new CDC Cyber 72 computer, which came into operation in 1974. I modified the Mark II equipment to operate at twice its old speed, with another standard tape deck operating at 30 ips, which is known as Mark III, in order to handle the increased number of SUGAR recording stations. I built some new interface equipment, called Mark IV, whose input was the Mark III equipment. This produced an industry-compatible seven track half-inch tape (IBM standard) on a Kennedy deck, which could be read directly by the new computer. This was used to handle all the data from ca mid-1973 until the termination of the SUGAR experiment in 1979. In addition for this later support work, I wrote the program for the Cyber computer which extracted the SUGAR array's coincident events and put them into a large data base in a form suitable for further analysis. This was a major task in large scale computing and sorting, and written in the FORTRAN code.

My scientific research with the SUGAR data involved the fast-timing analysis and interpretation up to 1969, when the array consisted of 34 stations covering an area of 40 sq km. I fitted plane shower fronts to find the cosmic ray arrival directions and I fitted spherical shower fronts to find the shower front curvature. The coincident events from the SUGAR array present a couple of problems as the array geometry varies for each event and the stations are non-coplanar. A generalised four-dimensional maximum likelihood iteration scheme provided the solutions. Since this nonlinear problem can be time-consuming (and possibly unstable), a good initial solution was essential, and I derived an excellent analytic approximation, which is not simply a linearised estimate (*). All estimates included the maximum likelihood errors, and the centre of the curvature was used to correct the arrival directions given the position of the EAS core on the array. The arrival directions were of course converted to celestial and galactic coordinates with error estimates and put onto equal area plots. Various related analyses using this data were presented in my PhD thesis.

The major question is the search for any anisotropy in the arrival directions for ultra-high energy cosmic rays. None was found given the statistical limitations of the number of events available. Now, nearly 50 years later, with many more ultra-high energy cosmic ray events from various arrays around the world, no such anisotropy has been established.

(*) On a number of later occasions, I have attempted to use this cunning "trick" without success!

--TonyP

SUGAR

Contents  [hide] 1 The Sydney University Giant Air Shower Recorder 1.1 Cosmic-ray Spectrum 1.2 Anisotropy 1.3 Neutral Cosmic-rays

The Sydney University Giant Air Shower Recorder

SUGAR, the Sydney University Giant Air-shower Recorder, operated close to the town of Narrabri in northern New South Wales, Australia, for more than a decade between 1968 and 1979. This array consisted of 54 stations with each station consisting of pairs of large liquid scintillation detectors (6 square metres viewed by a single photomultiplier tube) separated by 50 metres, covering a total area of 100 square kilometres. Each station operated in an autonomous manner with solar power units and receiving their timing information via radio receivers.

The large area of the array was designed to observe the most energetic extensive air-showers (above 10E17 eV). Another novel feature of this array compared to other ones was that each of the scintillation detectors was buried under a few metres of soil, thereby reducing the "soft" (electromagnetic) component of the air-showers and only being sensitive to the hard component (muons and hadrons). This feature of the SUGAR array meant that, while obtaining important information about the number of muons in cosmic showers, it was difficult to perform an absolute energy calibration of the showers to compare to other experiments. The energy was calculated by measuring the number of muons recorded and translating this into an energy scale via different models. The systematic error of this technique is quite large and could explain some of the differences observed between the energy spectra observed by SUGAR and other extensive air-shower arrays that operated above ground.

Cosmic-ray Spectrum

The energy spectrum of the most energetic cosmic rays recorded by SUGAR was reported in the following publication:

M.M. Winn, J. Ulrichs, L.S. Peak, C.B.A. McCusker and L. Horton, The cosmic-ray energy spectrum above 10E17 eV, J. Phys. G: Nucl. Phys. 12 (1986) 653-674.

This paper summarises the results obtained during the 11 year period that SUGAR was operational. The differential energy spectrum shows a typical power-law distribution, with a slope of 3.19+-0.01 between 10E17 and 10E19 eV, and a flattening of the slope to 2.99+-0.13 above 10E19 eV. This feature has been observed in other experiments and is known as the "ankle" of the cosmic ray spectrum. This change of slope might indicate a change in the origin of the cosmic rays (extra-galactic or galactic) or of its composition. The origin of this feature is still subject to speculation and debate.

Another interesting feature of the cosmic-ray spectrum above 10E19 eV is that there is no evidence for a cut-off at an energy of around 10E20 eV. 80 showers were observed with an energy above 4x10E19 eV and eight showers with energy larger than 10E20 eV (the most energetic shower had an energy of 2x10E20 eV). It is suggested that any interactions of these highly energetic cosmic-rays with the microwave background would produce an attenuation of the cosmic-ray spectrum at around 10E20 eV with an attenuation length of 100 Megaparsec (about 300 million light years) at 10E20 eV and 30 Megaparsec at 2x10E20 eV. If the cosmic rays at these energies are of extra-galactic origin a significant attenuation of the spectrum would be observed (the attenuation depends also on the type of particle), so the lack of a cut-off might indicate that the showers are either due to nearby galaxies or of galactic origin.

Anisotropy

The arrival directions of the cosmic-rays were also studied to search for any anisotropies that might indicate that these events originate from active galactic nuclei. This was made possible by the angular resolution of 4 degrees that could be achieved by measuring the relative timing of the signals in each of the stations. An analysis of the arrival directions was performed in the following paper:

M.M. Winn, J. Ulrichs, L.S. Peak, C.B.A. McCusker and L. Horton, The arrival directions of cosmic rays above 10E17 eV, J. Phys. G: Nucl. Phys. 12 (1986) 675-686.

No evidence for anisotropies was found in this analysis except for maybe a small excess from the two places on the galactic equator which pass over the array site for showers in which the number of muons exceeded 3.16 x 10E8 (plot c). Nearby galaxies as the source of the highest energy cosmic rays can then be ruled out, since there is no cut-off and no anisotropies due to point sources or clusters. However, a galactic origin for these showers could not be statistically confirmed even though there might be some indication of this from the small excess coming from the galactic plane.

Neutral Cosmic-rays

Another topic of interest that was studied with the SUGAR data was the possibility of correlating an excess of activity in the direction of certain X-ray binary sources in the southern hemisphere. With colleagues from the University of Adelaide, evidence for such a correlation of cosmic-rays with the period from the X-ray binaries was observed. The source of these cosmic-rays is likely to be neutral hadron emission since charged particles would be deflected by the galactic magnetic field. Photons are unlikely primaries because of the unique feature of SUGAR that only recorded muon and hadron activity and not the electromagnetic component of cosmic showers. Hence it is likely that these showers might be initiated by primary neutrons. This would be a remarkable example of Einstein's theory of relativity since one expects neutrons to decay with a lifetime of 887 seconds, but due to the boost in energy (average energies around 10E18 eV) the neutrons have an 80% probability of surviving the 2.5 kiloparsec (about 7500 light years) distance separating the 2A 1822-37.1 binary source and the earth (for the LMC X-4 source, the distance is 50 kiloparsecs or 150,000 light years, and the survival probability is only 2%, but this signal is less significant than the former). These interesting results have been published in the following papers:

• R. Meyhandad, B.R. Dawson, R.W. Clay, L. Horton, J. Ulrichs, M.M. Winn, Comparisons of some Apparent EHE Point Sources, 22nd Int. Cosmic Ray Conf., Dublin, 1, (1991) 384.
• B.R. Dawson, R. Meyhandad, R.W. Clay, L. Horton, J. Ulrichs, M.M. Winn, Search for EeV Neutral Particle Sources in the Southern Hemisphere, 22nd Int. Cosmic Ray Conf., Dublin, 1, (1991) 452.
• R.W. Clay, R. Meyhandad, L. Horton, J. Ulrichs, M.M. Winn, Neutral Particle Emission from the X-ray Binary 2A 1822-37 at Energies above 10E17 eV, Astron. and Astrophysics, 255, (1992), 236-241.
• R. Meyhandad, B.R. Dawson, R.W. Clay, L. Horton, J. Ulrichs, M.M. Winn, Evidence for Neutral emission above 10E17 from LMC X-4, The Astrophysical Journal, 391 (1992) 236-241.