[ History : ] From Modern Astronomer, 1997, by Josh Bloom.

Ray Tracing: A Cosmic Detective Story
     Colliding neutron stars within our galaxy, or super-bright
     objects at the edge of the observable Universe --- what are those
     mysterious gamma-ray bursts? Josh Bloom investigates...

Bursts of gamma-rays are observed to come from seemingly random directions
in the sky about once a day. Since the announcement of their discovery in
1973 scientists world-wide have been mystified by the cosmic events and,
until very recently, there was little consensus even on the distances to
the bursts, let alone what produces them. 

Many initially believed that the bursts originated from violent processes
on compact neutron stars within our own Galaxy. Then, as results from the
space-based Burst and Transient Source Experiment rolled in, most of the
community switched to believing them to originate at the edge of
observable Universe. Now, after a series fortuitous and remarkable
discoveries, the latter distance scale has been confirmed, making
gamma-ray bursts one of the most energetic phenomena in the Universe. What
follows is a remarkable tale of these bursts from discovery to present


Because our atmosphere is virtually opaque to gamma-radiation from outer
space it wasn't until the 1960s, when gamma-ray detecting satellites were
flown, that the first gamma-ray bursts (GRBs) were detected. The year 1963
saw the landmark signing of the Partial Test Ban Treaty by the world
superpowers. It was a promise that nuclear weapons would not be tested
underwater, in the atmosphere, or outer space so as to protect the
environment from radioactive fallout.  As a result, that year the US Air
Force sent into orbit the first of a series of satellites designed as a
verification of international adherence to the ban. The satellites, called
Vela (see Fig. 1), had X-ray, gamma-ray, and neutron detectors on-board.
As all three such types of emission are expected from a nuclear blast, a
coincident detection of all of them would have been a clear indication of
a nuclear test. 
Fig. 1: The Vela 5 satellite in low-Earth orbit. The Vela series was
the first group of satellites to detect gamma-ray bursts.

Credit: Laboratory for High-Energy Astrophysics at Goddard Space Flight Center

The satellite detectors were triggered many times over the lifetime of the
Vela program but (luckily) none appeared to have the tell-tale signatures
of a nuclear detonation. In 1972, a team at Los Alamos National Laboratory
in New Mexico reanalysed the data from the previous decade and determined
that a number of bursts of gamma-rays must have originated from somewhere
other than the Earth or the Sun. This was accomplished by noting the small
difference in the trigger times of the same event, as seen by different
satellites. Since light has a finite travel speed, a timing difference
translates directly into an angle of radiation incidence with respect to
the satellite positions. Knowing where the satellites were at the time of
the trigger allowed a crude position to be found. 

The spectacular discovery of GRBs (see Fig. 2) was announced in 1973 in
the famous paper of Klebesadel, Strong, and Olson called 'Observations of
Gamma-Ray Bursts of Cosmic Origin.' Dr. Ray Klebesadel, who still works at
Los Alamos National Laboratory, often refers to the events as *gamma
bursts* --- conspicuously leaving out his namesake --- perhaps out of
modesty for having been the co-discover of what has become one of the most
mysterious events known to humanity. 

Fig. 2: (left) Time-history of the first ever gamma-gay burst,
detected in 1967. It has what we now know to be a typical profile: its
intensity rises quickly and tails off slowly for about 10
seconds. (right) Time-history of a bright burst seen by BATSE in 1998.

Credit: Klebesadel, Strong and Olson; BATSE Science Team
Fumbling in the dark
The 1970s and 1980s saw an age of tremendous excitement about the new
phenomena.  Numerous theories cropped up as to their origin and distance
from us --- ranging from the very nearby (as close as the Sun) to the
very distant (from cosmic strings). At last count, there were over 200
distinct theories to explain the bursts. But most believed them to arise
from some process involving exotic stellar remnants (such as white dwarfs,
neutron stars, or black holes) in the disk of our own Galaxy.

Despite the intense efforts to understand the bursts, their origin
remained elusive.  An interplanetary network (IPN), headed by Kevin Hurley
and Tom Cline, was initiated whereby numerous satellites probing the Solar
System were outfitted with GRB detectors. In its heyday, six near-Earth
and four interplanetary satellites comprised the network. By 1987, some
200 positions of bursts from the IPN were found to unprecedented accuracy. 
BATSE: A light in the distance

On April 5, 1991 the Burst and Transient Source Experiment (BATSE) was
launched by the Space Shuttle *Atlantis* as part of the 17 ton Compton
Gamma-Ray Observatory, named after Nobel Laureate Arthur Halley Compton.
Since BATSE was 10 times more sensitive than previous GRB detectors, it
began to find bursts at a rate of about one per day. Of the nearly 2000
such bursts seen in the 30 years since their discovery, no two look
exactly alike. Some bursts last only a few milliseconds while others last
for hours. Many are dim, and flare just slightly above the background
level of normal gamma-ray flux in the sky, while others are briefly the
brightest, most energetic objects visible. 

But scientists have found clues to the nature of GRBs, and their distance,
by examining their global properties. One of their most striking
properties is that they appear almost perfectly randomly distributed on
the sky. This so-called 'isotropy' feature of the bursts, that there is no
preferred direction, implies that the Earth is very near the centre of the
bursting population. Copernicus revealed that Earth is indeed not at the
centre of most interesting astronomical systems: we live thankfully far
from the gravitational centre of our Solar System and nearly two-thirds
the way outside of the Milky Way centre. In fact, the only other isotropic
distributions of astronomical objects are nearby bright stars and very
distant quasars, near the edge of the observable Universe. 

The isotropy of GRBs had been known for decades from the IPN studies, but
since the earlier satellites only detected the brightest bursts, it was
hypothesised that a more sensitive instrument such as BATSE would find an
excess of faint bursts towards the Galactic Centre, confirming GRBs to be
a relatively 'local' phenomenon. But the discovery of the isotropy of even
the faintest bursts by BATSE meant that the local model of GRBs was in
serious trouble --- there was no statistically significant excess in any

Two viable competing distance scales emerged. The cosmological distance
scale, which had proponents from a minority of supporters in the 1970s and
1980s, became the model of choice. The isotropy of the bursts could be
explained with the basic tenet of modern cosmology --- the Cosmological
Principle --- which holds that the Universe should look the same to all
observers on large-scales.
The isotropy became a difficult observation to explain in the Galactic
model where GRBs were thought to come from isolated neutron stars in the
Milky Way. Since the Sun is offset from the Galactic Centre by some 30,000
light years, the only way out was through the proposition that the neutron
stars occupy a massive halo around our Galaxy more than half a million
light years in radius. Even in this scenario, a small deviation from
isotropy was expected, but since only a limited number of bursts had been
detected, the observed distribution remained consistent with the model.

Many stayed attached to the Galactic model since neutron star processes
explained some of the more controversial features of GRBs, such as some
reports of repetition. Moreover, by 1991 it was known that a distinct
sub-class of GRBs, called soft gamma-ray repeaters (SGRs; three are
known), do originate from neutron stars embedded within supernova
remnants nearby. 

By looking at the number of bursts within a certain brightness range,
BATSE also confirmed that we are seeing out nearly to the 'edge' of the
GRB population. Indulge in a terrestrial analogy: If one stands in a vast
field full of fireflies, each of which blink with the same intensity then,
assuming the bugs to be evenly distributed in the field, you would expect
to see only a few bright flashes from the ones nearby and several more
faint flashes from the many more bugs far away. The increasing number of
flashes with decreasing observed brightness is seen because (a) sources
appear dimmer the farther away they are and (b) the dimmer sources occupy
more volume of space around you so there are more of them. The 'edge' of
both fireflies and GRBs is found when fewer-than-the-expected number of
sources are detected at the faint end of the distribution (see Fig. 3). 


logn1.jpg here


In the two competing scenarios for the GRB distance scale, the 'edge' has
two very different interpretations. In the Galactic picture, the number of
neutron stars decreases with increasing distance from the galactic centre
out to a point where there are virtually no more left. In the cosmological
scenario, the decline is due to the expansion of the Universe: at earlier
times in the history of the Universe the volume of space was smaller than
it is now. As we look out into space, and back in time, we see some GRBs
that originated when the Universe was less than half its present age, so
we really look back to a time when the Universe was much smaller.  Hence,
there are fewer and fewer bursts the farther back in time we look! 
A New Chapter Begins

How could we begin to constrain the models of GRB production if we did not
know from how far they originated? It was a truly frustrating time. GRBs
just came and disappeared at random leaving no trace behind. Despite years
of intensive searches no one found unambiguous evidence for emission in
any other part of electromagnetic spectrum (some X-rays were known to
accompany a GRB, however). The bursts were mysterious orphans, seemingly
unconnected with any other known astronomical object.  And since the
positions of most bursts were highly uncertain (most BATSE burst positions
are only known to several square degrees), it was virtually impossible to
comprehensively study the surrounding sky field looking for a counterpart.

In March 1995, a debate between the two most ardent supporters both sides,
Bodan Paczyn\{'}ski (cosmological) and Don Lamb (Galactic), was held at
the Smithsonian in Washington, D.C. Touted as the possible turning point
in GRB research, it was the 75th anniversary of the Curtis and Shapley
'Great Debate' (where the size of the Universe was contested just three
years before Edwin Hubble made his seminal discovery about its expansion
--- and thus the size --- of the Universe). The parallel of the
Curtis/Shapley and GRB debates could not be more poignant: just as few
changed their minds about the size of the Universe after the 1920 debate,
most of us GRB aficionados left feeling even more confident that we were
upholding the side of truth. Just as in 1920, it was recognised that an
observational breakthrough was required to finally lay the distance scale
to rest. 
BeppoSAX and the 'Smoking Gun'

On 28 February 1997, the burst that would finally produce the 'smoking
gun', the long sought after observational breakthrough, arrived. It was a
typical burst in duration (about 30 seconds) but one of the brightest.
Several months earlier, the Italian-Dutch Satellite BeppoSAX was launched.
It had the unprecedented capability of pinpointing a GRB position by
localising the X-ray emission during a GRB. Eight hours after the event
(dubbed GRB 970228), the BeppoSAX team in Rome turned the more sensitive
instruments of the satellite towards the area located by the coarser wide-
field camera. What they discovered was the first fading X-ray afterglow of
a GRB.  More important, they had a highly refined position of the X-ray
emission (see Fig.  4). 


beppo.jpg here


Since other satellites were also able to observe the burst, there was a
small overlap region of the BeppoSAX positions and the IPN annulus from
which the burst could have originated. Intense scrutiny of this region
turned up a fading visible light source (see Fig. 5). The results were
reported by a large international team headed by Jan van Paradijs at the
University of Amsterdam and joined by UK astronomers, Dr. Nial Tanvir, Dr.
Max Pettini, and myself (*Nature*, volume 386, page 686). The recently
refurbished Hubble Space Telescope then followed the afterglow as it faded
(*Nature*, volume 387, page 476). 


ot.jpg here


Early controversial reports about the apparent proper motion of the
optical afterglow was a strong indication that the GRB must be Galactic in
origin. Even the fuzzy extension found by the HST near to the optical
transient was controversial --- was it the host galaxy or a mere
statistical blip? In the end, most researchers believed another burst
would come to settle the score. 

The prophetic hopes of the community came true on 8 May. Again, BeppoSAX
found an afterglow of a GRB and reported it to the world. On 10 May,
Howard Bond (Space Telescope Science Institute), working on Kitt Peak,
reported a brightening variable optical source within the error box of GRB
970508 (IAU Circular 6654). Then, on 11 May, just as the afterglow begin
its steady decline in brightness, a team from the California Institute of
Astronomy (headed by Mark Metzger) took a spectrum of the source with the
powerful 10-metre Keck telescope. In an IAU Circular (6655) that has
solidified the view of GRBs as one of the most energetic events in the
Universe, Metzger et al. reported the first discovery of a redshift of GRB
by observing absorption due to magnesium and iron. The redshift of GRB
970508 (z=0.835) places a lower limit to the GRB distance of several
billion light years. 


Panel1 here

The Theoretical Picture

The very recent measurement of a redshift has enormous implications for
the theory of GRB production. At the very least, such large distances
imply that the bursts are some of the most energetic phenomena in the
Universe, rivalling that of the fabulously brilliant quasars. During just
the first few seconds of the initial explosion, a typical burst liberates
the amount of energy that the Sun (about 10^{52} ergs total) will take 10
billion years to release. 
What are they?
From a *light travel-time* consideration (see panel) it is believed that the complex time-structure of most GRBs (see Fig. 2) must arise from a relatively small volume *of space*. Moreover, the amount of energy required implies that the mass of these progenitor objects must be at least that of our Sun. The small size and the large mass taken together imply that one of the most viable production mechanisms is through a violent merger of two massive compact objects such as neutron stars or black holes. The amount of energy required if the bursts were local to our Galaxy is (a puny) 10^{41} ergs. +++++++++++++++++++++++++++++++++++++++++++ Panel2 here +++++++++++++++++++++++++++++++++++++++++++ Tracing the afterglow

The decaying optical curves of GRB 970228 and GRB 970508 are now believed
(in a recent paper by Wijers, Rees, and Me\{'}sza\{'}ros) to be the
product of a high- speed blastwave slamming into the surrounding material.
There is further evidence to support this. Five days after the 8 May
burst, a faint radio signal popped up in the location of the optical glow.
Since radio emission of this sort is swallowed up if the emitting region
is too small, the emitting region must have grown at a tremendously fast
rate, very near the speed of light, to get to the necessary size within 5
days of the burst. 
Future Prospects

The history book of GRBs reads like a riveting mystery novel, full of
serendipity and surprise. Those that have spent many years trying to
understand the nature of their origin have done so with the deft hand of a
cosmic detective. But even as the great Mr. Holmes found himself lacking
the vital clue to solve the puzzle from time to time, so too did the
astronomy community require an observational breakthrough to finally
understand the nature of the bursts. 

The distance-scale debate of thirty years has ended with the discovery of
a burst redshift in May. But a vocal few remain sceptical that the optical
afterglow is unambiguously associated with GRB 970508. Perhaps the
decaying optical and X-ray emission was due to some unrelated phenomena.
In truth, astronomers do not really know how much objects vary down at
faint 20th magnitudes, so the association (and thus the redshift) may only
be a coincidence. We'll see. 

If the association holds though, and more GRBs are confirmed to reside on
the outskirts of the Universe, then it may be possible to infer that they
lie outside of galaxies. Knowing this can give important clues as to the
type of objects that create GRBs: if they are found to occur far outside
of galaxies, then merging neutron stars may become the paradigm since it
is expected that such systems can travel large distances before colliding
and exploding in a flurry of gamma-rays. 

Instead, GRBs may be found to originate from events as yet unknown ---
perhaps exploding mini black-holes or maybe 'exhaust from alien warp-drive
engines'. The possibilities are truly endless. This year has closed only a
chapter of one of the greatest mysteries in modern astronomy. Now, we
begin to ask new questions and the real fun begins again. 

*Josh Bloom has finished his M.Phil at Cambridge University and moves to
the California Institute of Technology in September to continue his
gamma-ray burst research. He has worked at Los Alamos National Laboratory
for several years and graduated from Harvard University (BA, magna cum
laude) where he won the Thomas Hoopes prize for writing and excellence in




What is a redshift?

A redshift is the measure of the change in frequency of light from a
distant receding object. Recall the *Doppler Effect* when the pitch of a
passing ambulance drops as it goes by. Certain molecules and atoms have a
characteristic frequency at which they vibrate. The ratio of the true
vibration frequency (or pitch) to the measured (lower) vibration frequency
is called the redshift (denoted by 1 + z). A cosmological redshift tells
us how fast something is moving away from us because of the expansion of
the Universe. And, since distance is related to recession velocity by the
Hubble constant, finding the redshift of an object or event is as good as
knowing its distance from us. 


Light travel-time consideration

The effective size over which an event takes place can be an important
tool in understanding the physical processes causing it. This can be
constrained very simply by multiplying the characteristic timescale of
observed intensity changes by the speed of light. Thus, since GRBs can
flicker on a timescale of a few milliseconds (thousandths of a second), we
guess that the initial size of the emitting region should be less than 100



Fig. 3: The so-called log N-log P brightness distribution of GRBs from BATSE and the
Pioneer Venus Orbiter (PVO). The bend away from the -3/2 power law slope towards the
bright end of the cumulative distribution could be caused by the expansion of the
Universe (if GRBs are cosmological in origin) or because we are seeing out to the
'edge' of a distribution of GRBs in the halo of our own Galaxy. 

Credit: E. Fenimore and J. Bloom


Fig. 4: Eureka! Discovery images of a fading X-ray source by BeppoSAX. The image on
the left was taken on 28 February and the image on the right on 8 March. 

Credit: BeppoSAX Science Team


Fig. 5: An R-band image taken at La Palma on 9 March 1997. Shown are the various
error box regions from the IPN, the BeppoSAX Wide Field Camera (WFD), and the more
sensitive instruments (MECS). The optical transient (marked with an arrow) is
visible in the intersection of all three error boxes. 

Credit: N. Tanvir and J. Bloom



'The bursts were mysterious orphans, seemingly unconnected with any other known
astronomical object'

'the bursts are some of the most energetic phenomena in the Universe,
rivalling that of the fabulously brilliant quasars'

'the emitting region must have grown at a tremendously fast rate, very
near the speed of light'