When one thinks of high energy astronomy, satellites tend to come to mind right away, because the atmosphere of the Earth is entirely opaque to radiation at UV, X-ray, or gamma-ray wavelengths (which is a good thing for human vitality). In order to directly detect such radiation, which comes from a celestial object, one must therefore place the detector above the atmosphere. Although sounding rockets and balloons played a critical role in the early days of high energy astronomy, satellites were the ultimate vehicle to launch the field into prominence. Over the past several decades, numerous breakthroughs in the field have been enabled by space-borne observatories.
However, satellite-based experiments become increasingly ineffective in detecting gamma rays at increasing energies. When photons reach an energy of tens of GeV, it is, in fact, problematic to detect them directly, because of the practical difficulty in constructing a suitable detector to "stop" them. This is where ground-based gamma ray experiments come in and contribute. At such high energies, there is a unique window of opportunity to do high energy astronomy on the ground. Although the gamma rays cannot penetrate the atmosphere all the way down to the ground, the consequences of their interactions with the atmosphere can be observed and quantified to infer their spatial, spectral, and temporal properties. In essence, one uses the atmosphere as part of a giant gas detector to register gamma-ray radiation.
1.1. Experimental Principles
The interactions between GeV-TeV photons and particles in the atmosphere result predominantly in relativistic electron-positron pairs, as illustrated in Figure 1. These secondary electrons or positrons lose energy mainly by bremsstrahlung radiation to produce gamma rays. The latter may produce more pairs and the pairs produce more gamma rays ... on and on the cascading process goes, until ionization becomes the main channel of energy loss for relatively low-energy electrons and positrons. Therefore, upon the incident of each gamma ray on the atmosphere, a shower of charged particles is formed in the atmosphere. Moving downward, the density of shower particles increases until it reaches a maximum (known as the shower maximum), typically at an altitude of roughly 10 km above sea level, and then begins to decrease. The air showers may be observed through the detection of light emitted (or induced) by the shower particles or the detection of those shower particles that manage to reach the ground.
Figure 1. Schematic of air shower development. Note the presence of muons (and neutrinos) associated with hadronic showers. Reproduced with permission from Konrad Bernlöhr.
Unfortunately, not all air showers seen are initiated by gamma rays. In fact, only a tiny fraction of them are. This is because the showers are also formed when cosmic ray particles (mainly protons) interact with the atmosphere, as also illustrated in Fig. 1. Since cosmic rays outnumber cosmic TeV gamma rays by many orders of magnitude, picking out gamma-ray-induced showers is truly like finding a needle in a haystack! This is a main reason why it had taken about two decades of painstaking development before the first TeV gamma-ray source, the Crab Nebula, was convincingly detected (Weekes et al. 1989.) The key for the success lies in the formulation of an empirical procedure to separate the showers initiated by gamma rays from those by cosmic rays. There are physical differences between electromagnetic showers and hadronic showers. Unlike electromagnetic interactions, hadronic interactions result mainly in pions. Therefore, hadronic showers mainly contain the decaying product of pions, including muons, electrons, positrons, and neutrinos from charged pions (±), as well as gamma rays from neutral pions (0). While subsequent 0-induced events are indistinguishable from the gamma-ray events of interest, ±-induced events manifest themselves, e.g., in the associated muons or neutrinos. It should be noted that the background events induced by cosmic ray electrons are also electromagnetic in origin and are thus difficult to eliminate.
Two broad classes of experiments are designed to explore the differences between electromagnetic and hadronic showers. One is based on detecting shower particles (photons, muons, electrons, positrons, and neutrinos) that reach the ground, while the other is based on detecting Cherenkov radiation induced by superluminal charged particles in air showers. The main advantages of the former include a long duty cycle and a large field-of-view, while those of the latter include high sensitivity (due to efficient gamma-hadron separation), low energy threshold, and good energy and spatial resolution. To a large extent, therefore, the two types of experiments are complementary in practice (e.g., wide-field surveying vs narrow-field imaging). The examples of particle-based experiments include Milagro (which is no longer in operation) and AS and ARGO, both of which are located at the Yangbajing (YBJ) International Cosmic Ray Observatory in Tibet, China and are ongoing. Cherenkov experiments can be further divided into imaging experiments, such as CANGAROO-III, HESS, MAGIC, and VERITAS, and non-imaging experiments, such as CELESTE and STACEE. The imaging technique was pioneered by the Whipple Collaboration, which led to the detection of the very first TeV gamma-ray source, and was greatly enhanced by the HEGRA Collaboration through stereoscopic imaging with multiple telescopes. The stereo-imaging techniques is employed in the current (and future) generation of narrow-field imaging experiments. In general, the imaging experiments are far more sensitive than their non-imaging counterparts. The duty cycle of Cherenkov experiments is limited by the requirement of their operating under good weather on moonless nights. For example, VERITAS typically runs for 700-800 hours in a year, or a duty cycle of < 10%, compared to the > 90% duty cycle of, e.g., Milagro. More technical details can be found in a recent review article by Aharonian et al. (2008).
1.2. Development and Scientific Drivers
The primary driver for the development of TeV gamma-ray astronomy is to utilize the unique window of opportunity on the ground to push astronomy towards the uppermost end of the electromagnetic spectrum. As the history of astronomy has shown, a new window into the universe nearly always brings about new discoveries. The prospect of probing the most energetic and most violent phenomena in the universe provided strong motivation for decades of painstaking efforts to develop and perfect techniques for the field.
TeV gamma ray astronomy attempts to address many of the same questions that other branches of astronomy do. They include: cosmic sources of TeV photons, radiation geometries and mechanisms, properties of radiating particles and their environments, and so on. The field also offers an excellent example of interdisciplinary, collaborative efforts between astronomers and physicists, because it also explores topics that go beyond "traditional" astronomy, including acceleration of particles in various astronomical settings, the origin of cosmic rays, the nature of dark matter, and cosmology in general. The interdisciplinary nature of the field is also reflected in the data collection, reduction, and analysis procedures. As already described in Section 1.1, the experiments employ instrumentation that is familiar both to astronomers (e.g., optical telescopes) and particle physicists (e.g., detectors). In data reduction, cuts are made to separate gamma-ray and cosmic-ray events, which bears resemblance to event selection in a particle physics experiment. The products of data analysis are, however, quite standard in astronomy, such as light curves, spectra, and images.
Since the detection of the Crab Nebula, TeV gamma-ray astronomy has experienced steady, albeit slow at times, growth throughout the 1990s and in the early 2000s, and has matured significantly over the past five years or so, thanks to the availability of a new generation of Cherenkov gamma-ray observatories. The number of sources detected has grown rapidly from a handful to over 70 (Aharonian et al. 2008, and references therein). More significantly, an increasing number of classes of sources have been established as TeV gamma ray emitters, including BL Lac objects, radio galaxies, quasars, shell-type supernova remnants (SNRs), pulsar wind nebulae (PWNe), X-ray binaries, and stellar clusters, as shown in Fig. 2.
Figure 2. Distribution of discrete TeV gamma-ray sources (as of March 2009). The sky map is in galactic coordinates. The colors differentiate various classes of sources (for legend see http://tevcat.uchicago.edu). Courtesy of the TeVCat Team
Besides discrete sources, large-scale diffuse TeV gamma ray emission has also been detected along the "Galactic Ridge" (Aharonian et al. 2006d) and in the Cygnus region (Abdo et al. 2007), offering direct evidence for interactions between cosmic rays and molecular clouds. The latter might be a significant contributor to the reported excess of signals in the Cygnus region that could be associated with Galactic cosmic-ray particles and gamma rays (Amenomori et al. 2006).