jump to navigation

Are We Asking This Question? August 8, 2008

Posted by gordonwatts in ATLAS, physics life, press.

ATLAS open end CPThere is a book review in the NYTimes right now by Overbye titled “Inside Story of the Telescope That Nearly Wasn’t Built” – about the Hubble.

The space agency wanted to make sure its  long-awaited and astronomically expensive telescope — soon to be launched into orbit above the turbulent fog of the atmosphere — made an appropriately cosmic splash. The advice from those of us in the press peanut gallery was always the same and simple: pictures — cosmic postcards like the live pictures of other planets being transmitted from the Viking and Voyager spacecraft — early and often.

This is PR 101 — everyone, including us scientists, is easily captured by pictures. Especially stunning ones. Sure — they may not be the best way to convey accurate scientific measurements – but they are very easy to relate to. Are we doing the right thing for the start of the LHC? Do we know what pictures – science pictures – we are going to be pushing to the public?

ATLAS has a whole outreach group (as does CMS, I’m sure). We have the ATLAS Book. We have a movie. But what cool picture are we going to give the press when the science starts? Another picture of our detector – like the one attached to this blog posting? Surely we can do better. Our event displays – most are tuned for us to look at as scientists, not for the press or the public. Do we have anything?

Enough of my ideas. What should we have ready when science starts to roll out? At the Tevatron we write these plain-English-summaries. They aren’t totally plain, unfortunately. But perhaps we should get someone from the PR office to work with every analysis that is published to work on something like that?

Other ideas?



1. Nick - August 8, 2008

So far? Yeah, building such a massive machine makes for some really good photo ops. Actual data is a little more difficult to make pretty (post-bubble chamber, at least). Still, I think actual pictures of the machinery will carry media attention well after the point that it turns on and science starts getting done.

2. Colin - August 8, 2008

Having a live technical status page would keep me entertained at least. Fermilab has an alright (but ugly) page that’s the same information broadcast on the TVs on site, but I haven’t seen any equivalent for the LHC.

3. Aaron F. - August 10, 2008

“Our event displays – most are tuned for us to look at as scientists, not for the press or the public. Do we have anything?”

Is the data that goes into event displays going to be made public in a timely manner? Is there any way for people who don’t work at CERN to learn how to read the event description files? There must be tons of people—myself included, if I’m not too busy with school!—who would love coming up with remixed event displays. There are definitely people who do amateur processing on space images, and if the right tools were available, doing amateur event displays shouldn’t be much harder!

Is there a standard file format for accelerator events? Is there any public event data from accelerators that are already running? Are any event descriptions small enough to stuff into a personal computer?

4. Anonymous - August 10, 2008

Aaron F. — that’s a fantastic idea. This would be a wonderful way to get the very large field of interested computer graphics experts from outside particle physics involved. IMO, you or Gordon should contact the CERN press office. If they’re too busy, I’m pretty sure (as an ATLAS collaborator) that no one could ever seriously mind if one were to post just one single random event on, say, a blog. (Best of course to try to get approval, though!) It’s definitely something that should be done. IMO, there should be a nice prize for the best event display from the public. Maybe you could even get a display with the head tracking you posted about earlier.

5. gordonwatts - August 10, 2008

Hey — this is a great idea! Why didn’t I think of this!?

So, as far as posting a random event — great way to get yourself kicked off the collaboration.

Now, I can’t see how the collaboration wouldn’t be willing to post some events – like some Z events or some top quark events, or any event for some analysis that was published.

Wouldn’t you need geometry files as well (i.e. what the detector looks like)? It could well be that people would make idealized geometries. And people have already done this with the Tevatron experiements – I know, for example, that one can load the D0 detector into ROOT. These things really are 3D – so you can move around them and look at them.

Now, standard formats. Hmmm, now that is a problem. 🙂 I guess we would have to make them up as we went along? I guess the most important thing would be that they can easily be read by a computer, right?

I need to contact our outreach guys to find out about this. They may already have something like this up their sleave…

6. gordonwatts - August 10, 2008

Oh — I meant to add. There is a status page right now — the really interesting thing is how cold are things (the answer is: very cold). You can check out the current machine state here: http://lhc.web.cern.ch/lhc/ I don’t know what they will do when they start putting beam in the machine, but my guess is that page will be a portal to it. When I find out I’ll be sure to post — because you are exactly right. Status pages are very useful for us users but also for everyone else that is interested in the LHC!!

7. Claire Lee - August 11, 2008

You can already do some playing around with simulated events in ATLAS, with HYPATIA, Hybrid Pupil’s Analysis Tool for Interactions in Atlas (these acronyms kill me 🙂 ): http://web.phy.bg.ac.yu/~hypatia/

Also, as far as I know where actual data is concerned, there are plans to do something like the “Hands_On-CERN” at Particle Physics Masterclasses ( http://www.physicsmasterclasses.org/physics/physics_hoc.htm ) with the LHC data (the original was with LEP data in the DELPHI detector). Though I guess that this will only come out in a few years, or once that particular data has been made public, but since the masterclasses are aimed at high school students and teachers it should come out in a relatively family-friendly format. Perhaps they’ll start with early data on Z decays and then, in a couple of years once results are announced, do something cool like Higgs searches?

I haven’t done any of the HYPATIA stuff myself so I can’t comment on it, but my advisor, fellow students and I did the Hands-On-CERN with about 70 high school science teachers from around South Africa in June as part of a conference they were on; in groups they looked at events of Z decays into jets and lepton pairs, classified each event, and calculated the branching ratios for each decay. The group that got the closest to the correct BRs got a prize 🙂 They thoroughly enjoyed the talks we gave them and the practical exercise they did, and learnt a hell of a lot along the way.

I think the Masterclasses are an excellent outreach tool for anyone interested in the LHC, but definitely not well known enough. And the best part about them is all you need is the internet (or the CD), and Java for the 3D interactive event displays (no ROOT necessary!) 🙂

8. Markk - August 11, 2008

Actually I kind of assumed there would be some 3D Youtube’d movie of a zoom through some of the first (interesting, beautiful) collisions recorded.

Or even an animation of collision with the detected tracks coming out with little flashes where the detectors reacted. Anyway something like that. That would probably literally be worth millions in public support over the years…

9. gordonwatts - August 11, 2008

They already have the ATLAS movie (which is even rendered in blue-red 3D). So that means they have converted the geometry files into the correct format. And since they have done that one should be able to easily convert them into ROOT format or any high end graphics format that can take advantage of a decent video card (which I don’t think ROOT can!?!).

10. Aaron F. - August 12, 2008

Gordon — thanks for your response! I’m glad I’m not the only one who is excited about this. 🙂

“Now, standard formats. Hmmm, now that is a problem.”

Maybe not… looking at the CMS data hierarchy, I get the impression that many CMS events will end up in the AOD (Analysis Object Data) format. These simulated ATLAS events are available in the AOD format too.

“Wouldn’t you need geometry files as well (i.e. what the detector looks like)?”

The AOD format contains collections of Track objects, which seem to have the detector geometry already taken into account. Each Track contains the x, y, and z coordinates of the outermost detector hit, for example.

“So, as far as posting a random event — great way to get yourself kicked off the collaboration.”

There’s a “live” feed (it’s about two weeks old) of images from the D0 control room monitors. Actual data is, of course, more sensitive than data displays, but judging from the zillions of events that accelerator experiments need to get significant results, I don’t think the shadowy cabal of CERN administrators would be too stingy about releasing a few dozen events. Of course, I don’t totally understand the aims and politics of data secrecy…

An even more secure approach, if it’s technologically feasible, would be to release a small number of events that were discarded by triggers. I don’t think CERN would mind the public going through a tiny fraction of its trash!

11. gordonwatts - August 12, 2008

Sorry Claire — it looks like the Spam filter nailed your posts. I un-spammed both of them.

12. gordonwatts - August 12, 2008

Clair — how easy is the Java 3D display to use? I’ve not tried it yet.

Aaron — So, “AOD” is, sadly, not a standard. And it contains way more information than the experiments are probably going to be willing to release. The CMS AOD and the ATLAS AOD, for example, are very different formats. One experiment can’t read the other’s. Worse, at least in ATLAS, a bunch of design decisions mean you need 1-2 GB of C++ source code to correctly read the AOD. This is marginal for doing real physics, and is even more marginal when it comes to sending it around!

You are right about detector geometry. What I meant was some file that described the physical layout of the detector. That way one could draw it and see things like the tracks in context. Drawing isn’t trivial, actually – for a charged track you have to propagate it through a magentic field, and, at least in ATLAS< given we have an open-core magnetic field that isn’t simple. However, we have a map of the field, so it can be done (well, we do it, therefore anyone with enough time on their hands can do it).

The prof down the office from me in Seattle worked on that live event field. When the accelerator is running it is 100% up to date (or should be). Perhaps 10 or 20 minutes behind reality. When there is no data arriving then it dips into a pool of “interesting” events it has seen go by. The real difference there is that it is showing pictures. You can’t take those pictures and get the full kinematics and extract the objects. Thus the collaboration has no trouble releasing this. I expect that ATLAS and CMS and LHCb and ALICE all to do something similar. I had it as my desktop background for a while. What I’d hope is that instead of a bunch of jpegs released for random events, a XML file (or similar) of a random event would be released. Anyone could connect to that feed. It would be sparse, so it wouldn’t be possible to do real physics on it, but it would be good enough that you could see jets, perhaps W’s and Z’s, etc.

The full data stream is huge. It would be seriously expensive for us to ship out. In fact, internally, we don’t deal with that stream much. The first thing we do is try to extract muons and electrons and jets from it. And then calibrate those. And then combine them into higher level objects (like W’s and Z’s and top quarks, etc.). And how we do the combination and how we do the calibration is constantly changing. The collaboration may interpret same event differently depending on what year it is. 🙂

13. Claire Lee - August 12, 2008

Hey Gordon, no prob. You can delete one though cause they’re the same.

The event display is okay. Once you get used to it it’s easy enough, but for a first-time (non-experimental physicist) user it takes a while to get the hang of it and understand what it’s all showing. Most of the questions the teachers had in the practical part were about understanding the actual display and how to use it to figure out what they needed to know. But once they understood they were fine.

As it stands on it’s own it’s well documented, though. There’s a section called “Wired Help” under “Picture Analysis” that explains it well enough, though I think it may be better in a tutorial format for people working on their own.

It’s also just a picture, of course – you can’t get the kinematics of the individual tracks, though they do give you the Ecms and measured energy. It’s not enough to do real physics, but it is enough to classify each decay. For what it’s aimed for I think it’s fine, but I think the interface could be made nicer if they want to use something like this for publicity purposes.

14. gordonwatts - August 12, 2008

Ok — I need to try out that event display. So far I’ve been too much of a whimp to run any of them.

15. tim head - August 12, 2008

My attempt to help visualise the scale of the ATLAS experiment:


If you visited the cavern last summer it was so full that you could not see how big it really was, all you could see was blue (lots of atlas is painted blue) metal right in front of you.


16. Aaron F. - August 12, 2008

Claire Lee — Very cool! I will definitely be playing around with HYPATIA when I have some more free time.

Hands-On CERN before sounds awesome! My school has quite a few people working on ATLAS, and also some people who are really into outreach, so I’m surprised I’d never heard of it before!

I love the stylized WIRED display that Hands-On CERN uses. It seems much more simplified than most of the event displays I’ve seen, and the coloring makes it easy to see patterns — it makes me think “physics is comprehensible” instead of “physics is incomprehensible.” If I’m reading the instructions right, each color represents a different particle jet. I want to ask how the tracks are sorted into jets, but I have a feeling that’s one of the questions that accelerator physicists spend years and years trying to answer… 🙂

gordonwatts —

“What I meant was some file that described the physical layout of the detector. That way one could draw it and see things like the tracks in context.”

Ahh, I see what you mean. I’d been thinking it might be fun to do more abstract representations of the data, which wouldn’t necessarily be precisely dependent on the detector geometry. For example, you could unroll one of the barrel detectors into a flat sheet and show particle strikes on that as rain hitting the surface of a pond, with high-energy gammas through the end caps represented as lightning flashes. Or, if you could figure out the initial 4-momentum of each particle tracked (I guess you’d probably need the magnetic field map for this), you could roll your own magnetic field to get interesting visual effects. You could send all the particles outward into a very strong uniform field, for instance, to get a starburst of corkscrew tracks reminiscent of old-fashioned bubble chamber photos. You could even virtually collimate the outgoing particles by giving them all the same initial position and direction, and use suitably designed electric and magnetic fields to show the energy distribution of the collision debris by making lower-energy particles curve out sideways earlier than higher-energy ones.

tim head — Great video!!! Seeing the detector rise towards the camera like that really gives a sense of the ginormity of the thing… I’ve never gotten such a strong impression from any of the other photos and movies I’ve seen. Plus, I’m a sucker for time lapse. 🙂

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: