What's up with the data?
8 May 2005 18:11:27 UTC
Topic 189135
(moderation:
Hi folks,
I've seen people asking about what's been done and where this project is going. The first part of the answer is: If you've been clicking through the front page of the project, go back and read the recent announcements.
The second part is a more verbose version of what's on the front page: The first search of the S3 data didn't find a pulsar, but it did find a lot of strange effects from the instrument in the data. These need to be cleaned out before there's any chance of detecting an astrophysical signal. You folks, with your hot-running CPUs, did the important work of finding them so that they could be removed. We were expecting some of this (like 60Hz and all its friends from the electric company), but there was more than expected. So now you are all running S3 data again, but it is a cleaner version of S3.
And now an answer to the related question of where this is all going: LIGO has been operational for several years now, in the sense of being able to measure strains. But it is a very complicated instrument (absolutely unprecedented in some respects) and therefore it is not a matter of just switching on. The first time you do that, it's not very sensitive. You track down one of the many subsystems which is giving the most trouble and beat down the noise from that subsystem. Then the instrument is a little more sensitive, and you look for the next thing, and so on. Sometimes you have to invent something completely new to get past a sticking point.
The result of this is that LIGO has been shooting toward its target sensitivity for the last few years. It is now very close to that sensitivity. S3 is within a factor of a few. The S4 data, which is in the can and being prepped for analysis, is within a factor of two. S5, which will start later this year, will be at (or at some frequencies a little beyond) target sensitivity. Also, S5 will not be a few weeks' run between equipment upgrades; it will be a solid year, and that also improves the chances of detection.
Hope this helps,
Ben
What's up with the data?
)
Thanks for the update, Ben.
So now you are all running S3 data again, but it is a cleaner version of S3.Is the S3 run with the added false source signals completed already?... or is that rolled into the cleaner version?
Hope this helps,
BenGlad I COULD help. I dumped protein@home... I like real science/engineering.
"No, I'm not a scientist... but I did stay at a Holiday Inn Express."
A regular post like this one
)
A regular post like this one is probably one of the most important things that the developers could do to "reward" the faithful user community for their hard work and dedicated contribution to the cause. It should have a very distinctive and unmistakeable title - like "Progress Report - Week 27" and it should be done either weekly or fortnightly, even if there is nothing new to report (there's always something to keep the users informed). It should be on someones "To Do" list and it should get done on time so that the users feel a sense of "connection" with the project and the people behind it. If some minion in the project wants to become famous, here is your golden opportunity :).
Anyway, thanks very much Ben. The information was very much appreciated.
Cheers,
Gary.
Hope this
)
Hope this helps,
Ben
Yes.
Thanks.
You have my heartfelt
)
You have my heartfelt gratitude for giving me another solid reason to stay with this project in the foreseeable future.
A regular post like this one
)
A regular post like this one is probably one of the most important things that the developers could do to "reward" the faithful user community for their hard work and dedicated contribution to the cause. It should have a very distinctive and unmistakeable title - like "Progress Report - Week 27" and it should be done either weekly or fortnightly, even if there is nothing new to report (there's always something to keep the users informed). It should be on someones "To Do" list and it should get done on time so that the users feel a sense of "connection" with the project and the people behind it.
I'll second this sentiment. Viewing the thousands of volunteers who crunch data and provide critical help with this research as co-collaborators is a good way to look at it. We are all in effect members of the Project Team.
Thanks very much for the detailed update.
A regular post like this one
)
A regular post like this one is probably one of the most important things that the developers could do to "reward" the faithful user community for their hard work and dedicated contribution to the cause.
Yes, I agree with this too. There are loads of us out here who are crunching 'for the science' (terrible phrase, sorry) who would love to know what progress is being made. It would probably be better as an article on the project front page, rather than in the message board, though.
Many thanks for keeping us informed.
B
Thanks Ben, So we at least
)
Thanks Ben,
So we at least know a tiny bit what is going behind the scenes, It will at least me and others keep motivated to help improve the instruments involved and hopefully will detect some 'waves'.
Many thanks,
John,
I would also like to thank
)
I would also like to thank Ben and company for the update. The importance of periodic communications such as this cannot be exaggerated. Cheers!
Hi folks, I'm too fried to
)
Hi folks,
I'm too fried to answer everything one by one, so here is a generic answer to the thread.
First, we're glad to hear you like this sort of thing. As Bruce said elsewhere, there will be some more progress reports coming up. Weekly may be a bit much, since not that much changes from week to week - it's only after the whole data run has been processed that we can start sifting, and then it takes a while. The CPU-intensive part that you are doing is basically making a first pass through the data for anything that looks suspicious; after that there is followup that is less CPU-intensive but more scientist-intensive.
There are efforts underway to make the analysis a bit more "real-time", but there will always be somewhat of a delay. Bruce had a nice analogy at a conference years ago: A lot of astronomy is looking for gold nuggets. Pulsars are more like panning in a stream. You're unlikely to get in one day a big lump of what you're looking for, but the little bits come in steadily and add up over time.
Second, the "new" S3 data does contain a few fake signals. We don't have any way for these (or even a real signal) to flash on your screen saver, partly because they get picked up after the fact as I said above. But they do calibrate the effectiveness of the analysis, as in "If there was a pulsar at this location, frequency, etc., we would pick it up with this probability." We might be able to work this into an "effective range" or "effective volume of the galaxy" for your entertainment, but right now we're still trying to come up with a version that's comprehensible to other scientists let alone the general public. It's more complicated than it looks because it depends on the pulsar frequency, the physics of how it's emitting, the statistical confidence level, all sorts of things.
If a "once per data run" progress report isn't enough and you'd rather have once per week or two, you can ask questions about gravitational waves and relativity in general. In fact there have been enough of those to set up something dedicated just to them, which I will open another thread to describe.
Enjoy,
Ben
Thanks again for your timely
)
Thanks again for your timely input, Ben. Any information you pass along gives us crunchers more insight into the science and problems you may be having. It makes the project more intresting and is certain to keep everyone focused on the project's success. Cheers, Rog.