What happens when a GRB goes off?

The first X-ray image of GRB 091221

Because Gamma Ray Bursts are so short-lived, we need to tell people about a new one very quickly. The traditional method of astronomy, where the astronomer publishes their data once they've fully analysed them and can present an interpretation as well, simply can't work for GRBs. We instead have to tell people very quickly about the burst so that they can go and gather data too. We still go onto analyse it and publish it (often working together with the people who followed up the Swift discovery) of course, but we can't keep the data private until then. This means that when a GRB is first detected, Swift scientists are involved in real-time astronomy, analysing data literally as it arrives, so that we can send a preliminary analysis of the burst to the scientific community within 20 minutes of the burst. So when I tweet something like, "Just found an X-ray counterpart." it really does mean that, we have literally just discovered a new object in the sky. The image to the right is an X-ray image of GRB 091221; the first image of a newly detected source. This image was produced just 72 seconds after Swift triggered on the burst.

But what do we actually do? What's going on in those first 20 minutes or so?

As soon as Swift detects a burst it sends a short message to the ground, which in turn gets distributed to scientists. For many people, such as the Swift team, this comes both as an e-mail and an SMS message to our mobile phones. As soon as we get this message, the people who are on duty dial into to a teleconference and log in to a web page. These on-call scientists are made up of specialists responsible either for the spacecraft as a whole, or one of the individual instruments (BAT, XRT and UVOT; see the Swift page for more information about the spacecraft and its instruments.) As Swift sends some of the data it's collecting down to the ground, those on-call scientists quickly analyse it, discuss the results and use the web page to write a circular giving what information we can. This includes things like:

Each of the instruments has its own paragraph giving this information (if they detected the burst), or, if the burst was not detected, a little bit of information about the observation. All of these things help other scientists to decide whether they want to look for the GRB.

Of course, it would be great to get a full and detailed analysis done in that first 20 minutes, but we can't because we don't have enough data. Swift only passes over the ground station once every 96 minutes, so when a burst goes off it has to send some data through a network of satellites run by NASA, which are shared by many missions. We only have limited band-width available, so only limited data comes down. The on-call scientists' job therefore continues for some time (maybe days, weeks or occasionally months) after the burst. When we first get the high-quality data from the ground station they need to analyse that and send a "refined analysis" circular giving more precise details of the GRb to the community. Then they need to repeat this analysis every day so that the Swift Science Team can decide, on a daily basis, whether to keep observing the GRB, and how much time each day should be spent doing this.

A typical time-line of the life of a GRB, from a Swift point of view, is shown below.

Time since the triggerEvent
0 secSwift triggers on the burst, and transmits the position to the ground. The Swift team get pages or SMS messages.
First few minutesIf it's safe to look at (not too near the Sun or moon), Swift automatically swings round so that the XRT and UVOT can look for the burst. They send limited data to the ground straightaway.
First 20 minutesThe on-call Swift scientists dial in to a telecon and analyse the limited data Swift is sending down. The BAT specialist works out a few details about the burst (how bright it was and how long it lasted) and the XRT and UVOT specialists decide whether their instrument has detected the afterglow and get a few basic details (position, brightness). Together these people write a circular e-mail and send it to the scientific community.
A few hoursSwift passes over the ground station in Malindi, Kenya, and downloads all of the data it has gathered so far. The on-call scientists analyse this data and send "Refined Analysis" circulars, which give an improved position, and detailed information like the spectrum (the colour, sort of), and how quickly it's fading.
Next few daysSwift keeps observing the GRB and sends data to the ground every few hours. The instrument specialists tell the rest of the Swift team what the GRB is doing.
Daily, at 9am (East Coat US time)The Daily Planning Telecon takes place. The BA will report on how bright the GRB is, and how it's changing, and the head of the mission will decide whether Swift will keep observing it, and for how much time each day.

The above is purely from the point of view of Swift. At the same time as this is going on, other astronomers will decide whether they can look at it with ground-based telescopes, and sending circulars giving their results. This may be reports of non-detections ("We don't see it, so it must be fainter than...") or a detection and brightness, or hopefully a redshift, which gives the distance to the burst.

So, this is what people like myself are doing in the minutes and hours after a GRB. Thankfully, a lot of the analysis is automated now (writing software to automatically analyse XRT data is one of my main jobs) so in those first few minutes the instrument specialists hopefully only have to check that the automatic results are correct, rather than try to, hurriedly but accurately, do all the analysis themselves!