Figure 1: A plot, in galactic coordinates of sources in the 3rd EGRET catalog. The energy range is from 100 MeV to less than 10 GeV where the upper limit depends on source luminosity and viewing time.
A new observatory, designed to measure gamma rays in an unexplored part of the electromagnetic spectrum, has obtained its first results. The STACEE collaboration has announced that they have detected the Crab Nebula at energies in the neighborhood of 100 GeV with their partially completed detector. This article will outline the motivation for and realization of this remarkable instrument. Gamma rays, the highest energy photons detected by astronomers, are born in some of the most exotic and violent places in the universe. Objects such as supernova remnants (SNR), blazar-class active galactic nuclei (AGNs) and whatever is behind the enigmatic gamma-ray bursts (GRBs) have been shown, over the past decade, to be bright sources of GeV gamma rays.
| Most of the information in this field of inquiry has come from the EGRET detector aboard the Compton Gamma Ray Observatory (CGRO), in orbit since 1991. EGRET has detected 6 gamma ray pulsars and more than 60 AGNs as well as a host of sources not yet uniquely identified with objects known at other wavelengths [1]. A plot of EGRET sources is shown in figure 1. |
|
Figure 1: A plot, in galactic coordinates of sources in the 3rd EGRET catalog. The energy range is from 100 MeV to less than 10 GeV where the upper limit depends on source luminosity and viewing time. |
| At energies above 250 GeV, ground based detectors such as the Whipple telescope in Arizona have used the air-Cherenkov technique (ACT) to observe astrophysical gamma rays using the light from the electromagnetic cascades intitiated by their collision with the atmosphere. The ACT detectors have overcome low duty factors (they can only operate on clear, moonless nights) and ferocious backgrounds from similar signals due to charged cosmic rays. The technique, perfected in the late 1980's and copied around the world during the 1990's [2], has resulted in the solid detection of 2 AGNs and 4 galactic sources associated with supernova remnants. A few more sources, detected by one instrument, await confirmation. The high energy situation is summarized in figure 2 where sources detected above 250 GeV are plotted. Clearly the sky is less alive at TeV energies than at the GeV scale. |
|
Figure 2: As in figure 1 but for sources detected by ground based instruments at energies in excess of 250 GeV. |
On the one hand, this is a major step forward. Detection of gamma rays with large terrestrial instruments is a triumph and will result in progress at the highest energy frontier where fluxes are far too low to be registered by anything that can be put into orbit. On the other hand a mystery has arisen: where are all the EGRET sources when one looks for them at energies above 250 GeV? Although most of the AGNs in the EGRET catalog have been observed at high energies by the ACT instruments, only 2, Mrk421 and Mrk501, have been detected.
For some sources the answer is simply that it is not possible to produce such energetic photons in the first place. Or, even if they are produced close to the centre of some powerful engine, they cannot emerge without being degraded in energy due to scattering or pair production processes. However we know that since two AGNs have already been seen by the high energy telescopes, some sources can produce photons with energies in excess of 250 GeV.
It is thought that the reason most sources are not seen at high energy is because the universe is opaque at those wavelengths due to the effects of intergalactic radiation fields [3]. The cosmic microwave background radiation is the most familiar example of such a field but there are also photons at shorter wavelengths. Indeed, one expects the presence of infra-red photons as cosmologically red-shifted relics of starlight from the earliest galaxies. High energy gamma rays can collide with these photons to make electron-positron pairs and thereby be absorbed. Lower energy gammas are below threshold for this process and thus survive to reach detectors on or near the earth. This explanation is given credence by the fact that the AGNs which have been detected at energies above 250 GeV are the closest of those in the EGRET catalog, closer than the effective absorption length. Thus the interval between 10 GeV and 250 GeV is scientifically very promising. It is in this range that one expects to see this cutoff mechanism at work.
The STACEE experiment was conceived to lower the energy threshold for ground-based detection of gamma ray sources. The aim is to achieve a threshold of approximately 20-30 GeV, about the level at which the spectra from the EGRET instrument run out of statistics. (Typical energy spectra from astrophysical sources fall with the square of the photon energy.) To get to such a level requires using an enormous collector for the Cherenkov photons since, as the energy of the primary gamma ray drops, so does the number density of photons at ground level. A convenient, cost effective way to get a large, steerable, mirror is to use the heliostats of a solar power station to synthesize the collector. This is the approach taken by the STACEE (Solar Tower Atmospheric Cherenkov Effect Experiment) collaboration.
Figure 3: The National Solar Thermal Test Facility in Albuquerque, New Mexico. STACEE uses a subset of the facility's 212 heliostats to collect Cherenkov light produced by electrons in air showers generated by high energy gamma rays.
The STACEE team comprises 9 faculty members and 6 graduate students from McGill and Alberta in Canada and Chicago, Barnard/Columbia, UC Santa Cruz, UC Riverside and Calstate/LA in the US. It is in the process of installing and commissioning the STACEE detector at the National Solar Thermal Test Facility (NSTTF) located at Sandia National Laboratories in Albuquerque, New Mexico. As shown in figure 3, the NSTTF is a field of 212 heliostats, each with an area of 37 square metres, and a tower approximately 60 m in height.
| The experiment works by using a subset of the heliostats (32 during the 1998-99 season but eventually 64) to collect and focus Cherenkov light onto secondary optics located near the top of the tower. The secondary optics, shown in figure 4, consist of 2 m diameter mirrors which direct the light onto cameras of photomultiplier tubes. Importantly, each heliostat is mapped onto a single phototube. |
|
Figure 4: Secondary mirrors located near the top of the tower direct Cherenkov light onto phototubes. A third mirror is being installed bewteen the first two and a fourth will be added later at a lower level. |
Figure 5: A member of the STACEE team near one of the secondary-mirror/camera combinations. Each black cylinder contains a light concentrator and phototube and views a different heliostat.
The pulses from the phototubes are digitized and recorded for off-line analysis but the data recording is not continuous as in the case of radio astronomy. Rather, it is similar to a particle physics experiment; one records the state of the experiment only if a trigger condition is satisfied. This trigger is based on the simultaneous arrival of pulses from several of the phototubes, typically within 10 nanoseconds of one another. This trigger selects air-showers since the Cherenkov photons in these arrive at the ground as a thin disk 300 m in diameter and a metre or so thick. The normal to this disk is the shower axis and is collinear with the direction of the incident gamma ray. Relative times of firing of each channel in the detector can be used in a fit to the shower front to obtain an estimate of the gamma ray direction with a precision of order 0.1 degree.
Even though individual phototubes in the detector may be producing 10 million pulses per second due to being hit by visible light from stars (night-sky background), STACEE records data at a rate of less than 5 Hz thanks to this multiple coincidence requirement. But most of these are not due to gamma rays. In addition to starlight, charged cosmic rays are an unwanted source of phototube signals. Like high energy gamma rays, they produce air-showers and are therefore accepted by the trigger. They must be rejected off line or by filters running in the data acquisition code. The rejection criteria are based on subtle differences in the nature of showers produced by photons (electromagnetic) and those produced by the protons and light nuclei of the cosmic rays (hadronic). In general the electromagnetic showers have a smoother light pool than do hadronic showers. The multiple heliostat detector, with its naturally pixellated nature, is well suited to implementing smoothness algorithms.
The cosmic ray rejection criteria improve the signal purity somewhat but the signal to background ratio is still a small fraction of a percent. Thus to measure the true gamma flux, STACEE uses the `on-off' technique. Data are recorded for 28 minutes while the detector tracks the putative source. Then the heliostats are slewed to a point in the night sky which is at the same declination but half an hour ahead of (or behind) the source. This takes about 2 minutes and when the new target has been acquired, another 28 minute run is started. This cycle is repeated through the night and during each night of the observing campaign. The result is a data set comprising a few dozen on-off `pairs'. It is the work of the offline analysis program to compare the numbers of accepted triggers in each set after applying any calibration data and tighter trigger cuts. A detection is reported if there are significantly more `on' triggers than `off'. STACEE is a photon counting detector but in a statistical sense only.
The STACEE collaboration [4] was formed in 1994 and has been doing feasibilty studies and prototype development for the last few years [5], [6], [7]. Construction of the final version of the detector began in earnest in 1997, with most of the hardware being produced at McGill and Chicago. By the fall of 1998 two thirds of the baseline detector were complete and an effort was made to detect the Crab nebula. The Crab is the brightest source of astrophysical gamma rays in the galaxy and is observed to have an essentially constant output. Thus it has become the standard candle of the field and is `required viewing' for new instruments. From November, 1998 to February, 1999 STACEE tracked the Crab during dark nights and while it was within 45 degrees of zenith. After cutting out data with poor weather conditions as well as various teething problems experienced with the new instrument, the collaboration was left with approximately 50 hours of `on-Crab' data and the same amount off-source. Comparison of the two data sets reveals an excess of 5438 showers in the Crab sample, a 5.8 sigma effect. After application of a tighter trigger, using off-line calibrations, and a rudimentary smoothness criterion, the excess increased to 7.5 sigma.
These results were presented at summer conferences [8] and a journal article is in preparation. To publish a measurement of the energy spectrum requires detailed knowledge of the energy scale and optical throughput of the detector and this is currently being studied. A preliminary estimate indicates that the peak energy of Crab gammas detected by STACEE is about 70 GeV. This is well below any of the first generation ACT telescopes.
For the immediate future the STACEE team is finishing the 48 channel detector that was in the baseline design. A third secondary mirror and camera have been built and are being installed and the final electronics are being integrated into the data acquisition system. Each channel will have its own 1 GHz digitizer, allowing the precise timing and multi-hit capability essential for lowering the detection threshold below 50 GeV. During the summer of 2000 the collaboration will expand STACEE to 64 channels for further sensitivity. The present plan is to run the device for at least three years. Initial targets include the Markarian galaxies 421 and 501 as well as the AGN 1219+285 and others that pass within STACEE's field of view. Further plans will be dictated by the data.
Canadian STACEE collaborators are supported by NSERC and FCAR. US funding of the collaboration comes from the National Science Foundation, the California Space Institute and the Research Corporation.
|
David Hanna
David Hanna started his professional life as a particle physicist, obtaining his PhD from Harvard in 1980. After a post-doc appointment at CERN he joined the NRC in Ottawa. Since 1985 he has been a faculty member at McGill and has pursued interests in high energy phenomena at both ends of the distance scale. |
 Motion
 FUSE
 Tables of Contents
|
Cassiopeia -- 1999-DS |