Gamma ray astronomy is a difficult science. Like X-ray, ultraviolet, and infrared astronomy, it requires observatories in space, because the Earth's atmosphere is opaque to these wavelengths of light.(*) However, it is unlike most other forms of astronomy in that "imaging" -- focusing photons using mirrors or lenses to form an image -- is impossible. Gamma rays are so energetic, they pass straight through any material you might use to focus them. So to observe gamma rays, you need very specialized telescopes.

The Compton Gamma Ray Observatory (CGRO) was the second of NASA's Great Observatory satellites, after the Hubble Space Telescope. As its name suggests, it was a gamma ray astronomy mission, designed to observe the most energetic events in the universe. CGRO was named in honor of the American physicist Arthur Holly Compton, winner of the 1927 Nobel Prize in Physics for his discovery of Compton scattering. It was ordered by NASA, and TRW was the prime contractor who assembled the satellite; the (four) scientific payloads were supplied by various institutions around the world. The satellite was carried into orbit by the Space Shuttle Atlantis (mission number STS-37) on April 5, 1991, and flew until it was deliberately de-orbited on June 4, 2000. It was the largest science satellite ever put into orbit, weighing nearly 35000 pounds (16000 kg). CGRO performed very sensitive and detailed surveys of the gamma ray sky, and catalogued several thousand gamma-ray bursts, but it failed in its key mission to discover the origin of these mysterious bursts.

CGRO had four separate instruments:

  • Burst and Transient Source Experiment (BATSE)
  • Oriented Scintillation Spectrometer Experiment (OSSE)
  • Imaging Compton Telescope (COMPTEL)
  • Energetic Gamma Ray Experiment Telescope (EGRET)

Each of these instruments used a slightly different technique to observe gamma rays, though all telescopes used Compton scattering in one way or another. They all observed gamma ray photons of different energies, and were meant for different purposes. I'll describe each briefly below; I know much more about BATSE (having worked with it in the past) so my descriptions of the other instruments will be shorter.


BATSE was composed of eight different detectors, fixed at the eight corners of the spacecraft. BATSE was meant to be an all-sky monitor, to detect gamma ray bursts coming from any direction. Each detector was composed of two separate instruments: a Large Area Detector (LAD), and a Spectroscopy Detector (SD). The LADs were meant as the workhorses of BATSE, designed mainly as light buckets to catch gamma ray photons. The SDs were a bit more discriminating, as they provided a means to actually measure the energy of photons as they came in, and not just count them. Both detectors used large, thallium-doped sodium iodide crystals (I mean large, like a cubic foot) which scintillate when hit by a gamma ray. The gamma ray was then detected by measuring the flash of light created by the gamma ray hitting the crystal.

Because the eight BATSE detectors could see gamma rays coming from any direction, it was thought to be an ideal way of detecting gamma ray bursts. And it was -- over its lifetime, it catalogued over twenty seven hundred of them. Unfortunately, BATSE wasn't very good at telling us exactly where the bursts occurred. You could get a rough idea of where the gamma rays came from by measuring the intensity of the burst seen by each detector; a detector pointing right at it would have a stronger signal than one pointed in a different direction. But BATSE never provided enough spatial sensitivity to actually locate a burst well enough to find it with an optical telescope. This wasn't for lack of trying -- CGRO established the "BACODINE" network, in which preliminary burst coordinates were sent to astronomers around the world via pager or email, and these astronomers would then scan that patch of sky for any optical transients. Unfortunately the BACODINE positions were never precise enough, nor the alerts fast enough to make an optical detection based on BATSE coordinates. These coordinates could be combined with observations from other satellites (particularly the Ulysses solar observatory) to improve the localizations, but this took days to accomplish. Gamma ray bursts fade on timescales of minutes so bursts would have long since faded by the time the source locations could be improved.

This was somewhat of an embarrassment for NASA as they had hoped to discover the origins of gamma ray bursts with this satellite. It was even more of an embarrassment when a small Dutch-Italian satellite called BeppoSAX finally managed to catch a gamma ray burst in the act and localize it in 1997. They are now believed to be massive hypernovae or the merger of two neutron stars or black holes at cosmological distances (millions or billions of light years away).

The official energy range of BATSE was 20 keV to 1 GeV, but in reality the low energy boundary of the SDs could be dropped to less than 7 keV by changing the gain on the photomultiplier. They could also be used for imaging, but this requires a bit of trickery. Since the satellite orbits the Earth, some sources which lie within the plane of the orbit are occasionally occulted by the Earth. If you measure the change in the total gamma ray flux as the target gets eclipsed by the Earth, you have a good estimate as to how many gamma rays per second were coming from that target. In principle, you could do this for just about any source on the sky, but in reality, only the strongest sources produced enough flux to be measurable in this way. For example, I was involved in a project to measure a five-year-long light curve of Scorpius X-1 using BATSE. It was difficult to do, but it produced interesting results.(**)

BATSE was designed and built by NASA's Marshall Space Flight Center, along with the University of Alabama (Huntsville), and the Universities Space Research Association.


OSSE was a set of four detectors, each able to point in a different direction. Unlike BATSE, the OSSE detectors could be slewed on their own to point at individual targets. Also unlike BATSE, they had a much narrower field of view -- a few degrees rather than half a steradian -- provided by tungsten collimators. The detectors are otherwise similar to BATSE, with the use of doped sodium iodide and cesium iodide crystals as the scintillators. The detector design is called a phoswich -- which simply means you have two different crystals attached to one another, and you can study the incoming photons by measuring how the incoming gamma ray interacts with each. The detector was able to distinguish gamma ray photons of different energy depending upon the pulse of light observed when the crystal scintillated, and thus OSSE was able to perform spectroscopy (with an accuracy of about eight percent in energy).

Because OSSE was able to slew its four detectors in almost any direction and it had a small field of view, it was primarily used to observe point sources like neutron stars, x-ray binaries, AGN, and quasars. The detectors acted like photometers and spectroscopes simultaneously -- photometers because they measured the light curve, and spectroscopes because they could actually measure the energies of the incoming photons. OSSE had an energy range between 50 keV and 10 MeV. The energy range covers energies common in many nuclear reactions (especially the electron-positron annihilation emission line at 511 keV). OSSE concentrated on observing places where we expect to see these nuclear reactions, particularly supernova remnants, and the spallation-generated nuclear fusion emission from solar flares. It also mapped the 511 keV annihilation emission from the Milky Way, including the Galactic center.

OSSE was designed and built by the Naval Research Laboratory of the United States Navy.


COMPTEL was the core of CGRO, despite the satellite's main mission being the search for gamma ray bursts. This experiment, along with EGRET, performed a detailed sky survey of the cosmos at gamma ray wavelengths. COMPTEL was an odd design, consisting of a liquid at the top of the instrument, and sodium iodide crystals at the base. Gamma rays passed through the outer shielding of the telescope, and underwent Compton scattering in the liquid. The energy released by this first scattering was then analyzed. However, the gamma rays were scattered by the liquid down into the sodium iodide detectors where they would then scintillate, as in the BATSE and OSSE detectors. The data from the two interactions was then combined to determine where the gamma ray came from (to within a degree or so), and what its energy was.

COMPTEL had a large field of view (about a steradian), but by combining the data from the two detectors, you could pinpoint the source location of the gamma ray to within a few degrees. COMPTEL was used to survey the entire sky in gamma rays with energies between 0.8 and 30 MeV, but it was also used to point at individual sources once the sky survey was complete. COMPTEL's energy range was also well-suited to observe high-energy solar flares, so it conducted many observations of the Sun as well. Like BATSE, COMPTEL was fixed to the spacecraft, and so pointing COMPTEL meant reorienting the entire spacecraft. In fact, most spacecraft reorientations were made to service COMPTEL or EGRET observations, and observations with the other instruments were scheduled around this. On a few occasions, a gamma-ray burst would occur in the field of view of COMPTEL, and thus these bursts were slightly better localized than those detected by BATSE alone. (But it still wasn't good enough to determine their exact location. Oh well.)

COMPTEL was an international collaboration, led by the Max Planck Institute in Garching, Germany, along with the Space Research Organization of the Netherlands, the University of New Hampshire, and the European Space Agency.


EGRET was the final experiment on board, and it observed the highest energy photons in CGRO's spectral range -- 30 MeV to 30 GeV. It utilized a combination of detectors, like COMPTEL. The upper stage of the detector was a spark chamber -- a wire-mesh chamber filled with an inert gas. The chamber was designed to convert the highest-energy gamma rays into a cascade of electrons, positrons, and lower-energy gamma rays. At the top of the chamber, the gamma ray produces an electron-positron pair, which ionize the inert gas. These ions then result in a spark being generated in the wire mesh. You can determine the direction the initial gamma ray came from by following the trail of the sparks. At the bottom of the detector, a sodium iodide detector catches the remaining gamma rays from the cascade, to measure their total energy. Because it could measure a photon's path and energy, EGRET could generate images and spectra simultaneously.

Because EGRET was designed to observe the highest-energy photons, its purpose was to study the most energetic phenomena in the universe. These include processes which occur near the surfaces of neutron stars, and near the event horizons of black holes, thus giving us a view of these extreme forms of stellar matter. On larger scales, supermassive black holes lying at the centers of quasars, Markarian galaxies, and other AGN also create incredibly energetic gamma rays -- some billions of times more energetic than even EGRET could detect! EGRET let us measure the spectra of these objects at gamma ray energies with very good precision, which in turn helps us build models of how such energetic photons are created.

EGRET was also a multinational collaboration, involving NASA's Goddard Space Flight Center, Stanford University, the Max Planck Institute, Northrop Grumman, Hampden-Sydney College (Virginia), and the University of Heidelberg.

The CGRO mission was considered a success, and deservedly so. Although it never discovered the source of gamma ray bursts, CGRO provided a us with a wonderful new view of the heavens. BATSE generated a nearly complete catalog of gamma ray bursts over the course of its operation. The structure of the BATSE light curves told us much about the physics of these events, and the distribution of gamma ray bursts nearly evenly on the sky strongly suggested a cosmological origin. The EGRET and COMPTEL telescopes produced very sensitive maps of the universe at gamma ray energies, much better than those that had previously been done. The EGRET map showed that much of the high energy gamma radiation in our sky comes not from point sources or explosions like supernovae, but is a diffuse emission arising from the interaction of high energy cosmic rays with the interstellar medium. EGRET also provided us with a view of the very highest energy gamma rays, allowing us to view some of the most energetic events in the universe, like the centers of active galaxies. In fact, it discovered a new class of super-powerful AGN, the gamma-ray emitting blazars. COMPTEL generated a map just like EGRET, but at lower energies. This map also provided us with a clearer view of the gamma ray sky, particularly of gamma rays generated in our own galaxy. For example, COMPTEL mapped the distribution of the radioactive isotope Aluminum 26 in the Milky Way, helping us understand how matter ejected from supernovae is spread throughout the galaxy. Finally, OSSE detected the presence of an antimatter source in the center of the Milky Way, strongly suggesting the presence of a supermassive black hole there. These discoveries, and many others, greatly expanded our understanding of the high-energy universe.

However, there was one more disappointment with CGRO besides the gamma ray burst mystery, namely the failure of the on-board data tape recorders. CGRO was meant to perform observations semi-autonomously, taking data, storing it onboard, and then sending it by radio to a receiving station. When the data recorders failed, data had to be sent live, directly to ground station. This was done via NASA's Tracking and Data Relay satellite system (TDRSS), and it worked ok. Because TDRSS couldn't be used exclusively for CGRO, only about 80 percent of the data taken by the spacecraft was actually recorded. A little bit more data was lost as a result of the spacecraft passing through the South Atlantic Anomaly -- the charged particles there would overwhelm the signals from gamma rays, so the spacecraft was usually turned off for a short time on each orbital passage.

One last note, on the death of CGRO. As I said at the top, and as you almost certainly heard on the news, CGRO was deliberately brought down from orbit by NASA. This was done because the spacecraft only had two functioning gyroscopes left to stabilize the craft, and one of these was not in good health. Because the spacecraft was so massive, parts of it would have survived a re-entry, much like parts of Skylab survived to crash-land in remote Australia in 1979. To avoid inadvertently causing injury or death if the spacecraft were to make an uncontrolled re-entry over a populated area, it was brought down in a "controlled descent". The best guess for the re-entry location was in the eastern pacific, several hundred kilometers off the northwest coast of South America.

(*) Please see my writeup in Air Cherenkov Telescope for a gamma ray telescope that doesn't fly in space.

(**) see (I am author #4).

I used many resources for this writeup, but the best semi-technical site was is an excellent site for young people and those interested in non-technical and historical information.