In our Special Topics analysis of gamma-ray burst
research over the past decade, the work of Dr. Chryssa
Kouveliotou ranks at #4 by total cites, based on 114 papers
cited a total of 4,165 times. According to
Essential Science IndicatorsSMfromThomson
Reuters, her record includes 183 papers, the
majority of which are classified in Space Science, cited a
total of 5,998 times between January 1, 1999 and April 30,
2009. She has also been named a Highly Cited Researcher in
Dr. Kouveliotou is a member of the Gamma-Ray
Astrophysics Team at NASA's Marshall Space Flight Center in
In the interview below,
ScienceWatch.com correspondent Gary Taubes talks with Dr.
Kouveliotou about her work in gamma-ray
You've been working on gamma-ray bursts long
before they caught the imagination of the astrophysical community at
large. What prompted your research?
I first started working on gamma-ray bursts in 1978 when I was doing my
Ph.D. work in Germany, and I used to joke that there were two and a half or
three and a half Gamma Ray Burst astronomers at the time, and I was the
half. I was just very intrigued by these bursts. They were a brand-new
phenomenon and we didn't know anything about them. There were maybe three
or four instruments at the time capable of seeing them.
Remember, the phenomenon was only discovered in 1967, when the VELA
satellites were launched. These were designed to enforce the nuclear test
ban treaty by monitoring clandestine nuclear explosions from space. They
observed the first gamma-ray burst in 1967, and then many more, but all
they could say is that they didn't come from the solar system. The first
paper on the discovery of gamma-ray bursts came out in 1973.
When I got into the field there were a few instruments capable of observing
them; the field was very small and the interest was very small, although
slowly it started growing. Then the people here in Huntsville, Alabama, led
by Jerry Fishman, designed and built an instrument to fly on the Compton
Gamma-ray Observatory, which was launched in 1991, and that instrument was
dedicated to burst detection. It was called the Burst and Transient Source
Experiment—BATSE. I started working with that group when the
satellite was launched. When the instrument was turned on, data started
coming and it was a watershed—every day a discovery. Those were very
exciting and wonderful times.
What was it about the 2003 Nature paper on
the gamma-ray burst of March 2003 that made it such an influential
paper (Hjorth J, et al., "A very energetic supernova
associated with the gamma-ray burst of 29 March 2003," 423:
847-50, 19 June 2003)?
GRB associated with a SN
(April 25, 1998): (fig. 1)
is the AFTER and (fig. 2)
is the BEFORE. The circle
with the dot in the center
is where the explosion took
place in 1998 as can be
seen by the bright source
at the left at the same
An artist’s concept
of the exploding star that
creates the GRB jets as it
Well, you have to realize that what we do is look for these bursts first in
the gamma-ray part of the spectrum, where there might be one or two a day;
we try to establish the direction, and then we look in other wavelengths to
try to identify a counterpart. The question was after all these thousands
of gamma-ray bursts—with BATSE, we had 2,704 confirmed events in
almost 10 years—we still needed to identify the optical counterparts.
The way BATSE located gamma-ray bursts wasn't very accurate. Its error box
was large and the ground-based observatories would not spend their precious
observing time trying to cover the entire region of the error box. Nobody
was very much interested in looking for the optical counterparts except us.
And we didn't know what the critical point was in time. How quickly did we
have to look after seeing the burst? Did we have to do it fast, or would
the source still be visible for days afterward? And then how bright would
the optical counterpart be? And how would the brightness decay? What was
the limiting magnitude of a counterpart? Did we need a big telescope or a
small telescope that was fast enough? When I worked with BATSE, we worked
with a lot of other people; we started these collaborations and we gave
them all these directions we calculated, but we were not able to find
This changed with the launch of Beppo-Sax in 1997. Beppo-Sax had an
instrument, the Wide Field Camera, which was ideal for locating gamma-ray
bursts. It could locate them very accurately compared to BATSE. In February
1997, it discovered the first X-ray counterpart of a gamma-ray burst, and
the location was accurate enough that an optical counterpart was also
observed. The big question was what was the origin of these gamma-ray
bursts—galactic or cosmological—and, depending on where it was,
what could create something like that?
What were the options for what could cause one of
There weren't many models. One viable one was the collapsar model,
suggested by Stan Woosley in 1993. He suggested that gamma-ray bursts are
the products of the collapse of a massive star, which ejects its hydrogen
envelope and is collapsing into a black hole. Another model was the merging
of two neutron stars—or two compact objects. Once the second or third
event was identified with a counterpart, we were able to put the Keck
telescopes on the sources and measure the Doppler shifts, determine the
redshifts of the events—the distance—and establish that they
were cosmological in origin.
Is this is what you did with the 29 March 2003
This paper was about the third burst that could be identified with a
supernova. The whole idea is that when a collapsar dies, you probably see
an effect like a supernova explosion. People looked to see if a supernova
explosion was associated with a gamma-ray burst. The very first one found
was in April 1998—April 25th. It was found
serendipitously. We looked at the location of the burst with an optical
telescope and saw this very, very bright source. This is the nearest ever
gamma-ray burst to have been reported—37 megaparsecs from here. No
wonder it was so bright. But that event was disputed. People weren't
convinced it was really associated with the gamma-ray burst because the
original error box was so large. There was a good probability it was
associated, but not 100%, maybe only 99%.
Then in 2003 we had the third such event—030329 came along. It was
probably among the top two percent of the brightest bursts ever detected.
It was very bright and also nearby. And now we had
collaborations—huge teams—ready to go to work from the ground
and follow up quickly on the optical counterpart. In this case, we were
able to detect the optical effect. When we looked at the spectrum, we saw
lines there that were broadening every day; the expansion velocity was
making them broader; the ejecta from this supernova were expanding at a
very high velocity; one of the fastest exploding ejectas ever recorded. The
shape was very similar to the spectra of that event reported in 1998.
So in one fell swoop this event affirmed that a supernova can be associated
with a gamma-ray burst and it confirmed the 1998 association with a
supernova. It also established that a collapsar was a pretty good candidate
for these longer events.
What do you mean by "longer events?" Does that mean
there are shorter events, as well?
This is one of the major contributions I've made to this field. I
established the existence of this bimodal distribution—a group of
short events, roughly less than two seconds, and a group of longer events
that average around 30 seconds. One of the problems with Beppo-Sax, for
example, is that it didn't have the capability to detect short events, so
all the counterparts identified for quite a while were exclusively of these
longer events. Then the Swift satellite was launched, and that can point
the satellite at the counterpart in less than a minute—that's the
strength of the satellite, hence the name—and so it could find
counterparts for the short gamma-ray bursts as well.
So what is the best explanation for the shorter
events, and are they distinctly different than the longer
We used to think the longer events were associated with collapsars and the
shorter events with mergers. Now I'm not entirely sure we can say that.
We're still waiting to see more to be sure.
Why the uncertainty?
It used to be that we thought mergers only happened in very old
populations, but now models say that you can have them in younger
populations, and some of these short events have been located in younger
galaxies—star-forming galaxies, which didn't fit the old models. And
now there are collapsar models that could also conceivably create short
events if the right conditions are met. So it's looking like a medley.
We can't say one model makes all the long events and this other model makes
all the short events. Other events may also create these bursts.
Magnetars—magnetic stars—are a possibility. This is one of my
other contributions to this research. Magnetars are neutron stars
conjectured to have extremely high magnetic fields—about
1015 gauss or about a thousand times higher than so-called
"...when we look at these
high-energy transients, we can expect
These were also found serendipitously. They're also pulsating stars;
they're rotating and their peculiarity is that they are slowly rotating. So
far, the dozen we've discovered have a very narrow period range—from
5 to 12 seconds. Other pulsars have a much wider range.
When were these magnetar-associated events first
discovered, and what was serendipitous about it?
In 1986. You realize that in principle we should never see a burst from the
same place twice, at least not according to the burst models. This
shouldn't happen, since the event that causes the burst either blows up the
star or merges two stars together. Then in 1985-86, Kevin Hurley was
working with data from a Russian satellite and he realized that some of the
events seemed to come from the same quadrant of the sky, the same general
direction. That's all he could say. He sent around messages to people who
had X-ray detectors, and I was then working on the SolarMax mission looking
for gamma-ray bursts. He pointed out one burst that came at a particular
time and said see if your instrument detected it, and we did.
Then the detector on ISEE-3/ICE satellite, which was still alive, detected
it again. We were able to triangulate the signal from the sky and we
pinpointed it to the same source. A meeting was held and this was one of
the main subjects—what is this repeater? Why is it repeating? What
kind of source is it? A different type of star? We decided to call it a
soft gamma-ray repeater, because the average energy of the photons in these
repeating bursts was much lower than the average energy of photons
associated with other gamma-ray bursts.
Then we went back and found another repeater coming from N49, a supernova
remnant in the Large Magellanic Cloud. So it all came together. These
sources created a new group, the so-called soft gamma-ray repeaters. When a
new satellite was launched in 1996, the Rossi X-ray Timing Explorer, I
wrote a proposal to look long and hard at one of those sources, which by
that time had been identified as a faint X-ray source, to see if it was
rotating. These sources exhibited another characteristic: they can be
dormant and then suddenly they are active and emit multiple bursts. When
this happened in the source we were studying, we identified a period.
We can now say these sources are neutron stars and it's a totally different
phenomenon. Right now, as I said, we have about a dozen of these sources,
and there's a lot of development in the field. I'm now leading a project
with the Fermi Observatory, which was just launched a year ago. Whenever
one of these sources becomes active—four did in the last
year—we do a lot of analysis on what exactly is happening.
How would you summarize what you've learned about
gamma-ray bursts over the past ten years?
That when we look at these high-energy transients, we can expect anything.
There are a lot of different objects out there and there are probably
phenomena we've yet to identify. No single model can describe the whole
What research are you doing now and what do you
have planned for the future?
Well, I'm still working on gamma-ray bursts, and on soft gamma-ray
repeaters, these magnetars. But I'm also very interested in different areas
of research. We've just submitted a proposal to the Astrophysics Decadal
Survey. Every 10 years the field of astrophysics and cosmology is reviewed
by a committee, which determines what should be done in the next 10 years.
We've proposed to study cosmic chemical evolution: star formation and
evolution, when the first stars formed, when the first galaxies formed.
What is this cosmic web, in effect? To do this, we're going to use
gamma-ray bursts as a tool.
So instead of being the goal of the research, these bursts are going to be
one of our tools. We're going to study structures by using the spectra from
gamma-ray bursts that explode behind these structures. We're going to study
what the absorption lines in the spectra are, and then, as the gamma-ray
bursts die away—they're very helpful in this sense—we're going
to look to see if we can detect the same lines in emission in the spectra
and then tie the two together with the structures at that location and so
study the cosmic web—in the intricate network of stars and matter and
clusters of galaxies. The mission we've proposed is called Xenia—from
the Greek word for "hospitality"—and it's a very big collaboration.
It's at least 50 institutions, with 90 team members from all over the
When do you expect to hear if it gets
The only thing we're going to hear from the Decadal Survey is whether it
should be a priority in astrophysics. We'll get a ranking, that's all. Then
if Congress decides to fund NASA sufficiently, NASA will open the field for
proposals for missions. If we're ranked high scientifically, if we fall
close to the top in this ranking, then we have a chance to do this. It all
depends, of course, on whether NASA gets the money.
If there was one thing you could go back and do
differently in your research career, what would it be?
I would probably just spend more time learning how to build
detectors—doing hands-on development of the instruments. Spending
time building these kinds of instruments is the best way to understand
their strength and limitations. I never did that. I wish I
Chryssa Kouveliotou, Ph.D.
Gamma-Ray Astrophysics Team
Science Missions and Systems Office
Huntsville, AL, USA