Enough work for crunchers until the Advanced LIGO launch?

Kotulic Bunta
Kotulic Bunta
Joined: 22 Jan 05
Posts: 33
Credit: 859204
RAC: 0
Topic 194708

I was wondering whether there is enough work for crunchers until the Advanced LIGO is put into operation.

From the available documents I understood that this launch is planned for 2014. Now we are analysing the S5 data, and currently the S6 data are being gathered/measured with the interferometers. Will those S6 data be enough for the next four years of analysis?

And what is the association with the launch of Arecibo binary pulsar search this year? Altough definitely useful and connected with the GW search, it still evokes thoughs that it drains some power which otherwise could be used for direct GW search.
(although of course I am pretty sure that this activity is well prepared and reasonable).

I apologize if I'm asking questions which are answered elsewhere...

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6591
Credit: 327588262
RAC: 214258

Enough work for crunchers until the Advanced LIGO launch?

Quote:

I was wondering whether there is enough work for crunchers until the Advanced LIGO is put into operation.

From the available documents I understood that this launch is planned for 2014. Now we are analysing the S5 data, and currently the S6 data are being gathered/measured with the interferometers. Will those S6 data be enough for the next four years of analysis?

And what is the association with the launch of Arecibo binary pulsar search this year? Altough definitely useful and connected with the GW search, it still evokes thoughs that it drains some power which otherwise could be used for direct GW search.
(although of course I am pretty sure that this activity is well prepared and reasonable).

I apologize if I'm asking questions which are answered elsewhere...


There will always be plenty to do! :-)

The base task of E@H is to 'de-Doppler' the signals from the LIGO and associated IFO's. The Earth's rotation and it's passage in orbit around the Sun alter the reception of whatever are the source's emitted waves. It's an endless task of looking at data sets and trying to match to some putative pattern within it. Since we are looking for continuous waves here at E@H, then those signals are going to be repetitive ( ie. pulsar like ) at some frequency, with some profile or shape, plus maybe either gradually slowing down or speeding up too. Oh, and we don't know where they are either!

While the analysis cannot be any more confident than permitted by the quality of the recorded data, I don't think we have as yet fully explored all possible permutations ( and granularity of divisions ) within the parameter space available. Indeed the hierarchical nature of the current search strategy can be seen as an effort to grab the 'low hanging fruit' that might be about with some expectation of reasonable efficiency given what resources that E@H volunteers have made available.

However it is well to note that E@H has already yielded results of valuable significance, that being the placement of upper bounds upon certain astronomical behaviours. Theorists will have noted this and I expect have adjusted their thinking and models accordingly. Further analysis at E@H can thus refine those limits even in the absence of a confirmed detection - which of course remains the grail to seek.

[ E@H has also been implicitly testing the entire pipeline from spacetime wiggle to statistical plot, through the mechanism of hardware and software injections into the data stream. We need to know the whole enterprise works correctly to gain confidence in any future detection. ]

In any case Virgo and GEO aren't part of the AdLIGO upgrade timeline you mention, so their data taking will remain for us to examine as well in the interim.

As for the ABP search, it was included for a number of reasons. Firstly their is some bloody good science in that - finding new or characterising known pulsars from radio data. Secondly the astronomical targets are likely to be of the same type that will give us gravitational waves - very valuable for cross confirmation when the time comes. Thirdly, but not necessarily least, is to maintain enthusiasm & interest of we E@H contributors during a period when it is about even odds as to whether a confirmed gravitational wave source signal will arise. I guess people like/need milestones and prizes to score on! We are just on the cusp of delving into the region/edge/limit of hard detection with the IFO's ( from a quantum mechanical limit viewpoint ) - AdLIGO is meant to properly bag that. Enhanced LIGO is yielding some pretty notable lessons already ( see my Detector Watch threads in the Science board ) .

Cheers, Mike.

( edit ) Actually now that I'm answering this, another obvious strategy comes to mind. We know where some pulsars are. Thousands in fact. Some have got to be in binary systems - even if you don't yet know fully which ones. Why don't we 'point' our signal search algorithm at them specifically and see what pops it's head up? All the data is both time stamped and from known terrestrial position(s). That yields Doppler components to any given sky position and the incoming vector from the source upon any given IFO arm arrangement. You don't know the pulse profile but heck you could have a couple or five stabs at that. Hmmmm .....

I have made this letter longer than usual because I lack the time to make it shorter ...

... and my other CPU is a Ryzen 5950X :-) Blaise Pascal

Kotulic Bunta
Kotulic Bunta
Joined: 22 Jan 05
Posts: 33
Credit: 859204
RAC: 0

Thanks Mike for your prompt

Message 96248 in response to message 96247

Thanks Mike for your prompt and detailed reply.

Quote:

There will always be plenty to do! :-)
...

I am really glad to hear that..

Quote:
( edit )We know where some pulsars are. Thousands in fact. Some have got to be in binary systems - even if you don't yet know fully which ones. Why don't we 'point' our signal search algorithm at them specifically and see what pops it's head up? All the data is both time stamped and from known terrestrial position(s)...

Well, it seems that sort of that has already been attempted (although not for all 2000+ but "only" for 116 known pulsars):
Searches for gravitational waves from known pulsars with S5 LIGO data

Anyway, I am really looking forward to the next years of analysing data from this fascinating project... regardless of whether we would find GW or whether we would let this pleasure for the next detectors/generations...

DanNeely
DanNeely
Joined: 4 Sep 05
Posts: 1364
Credit: 3573885105
RAC: 741009

RE: ( edit ) Actually now

Message 96249 in response to message 96247

Quote:

( edit ) Actually now that I'm answering this, another obvious strategy comes to mind. We know where some pulsars are. Thousands in fact. Some have got to be in binary systems - even if you don't yet know fully which ones. Why don't we 'point' our signal search algorithm at them specifically and see what pops it's head up? All the data is both time stamped and from known terrestrial position(s). That yields Doppler components to any given sky position and the incoming vector from the source upon any given IFO arm arrangement. You don't know the pulse profile but heck you could have a couple or five stabs at that. Hmmmm .....

A few years back one of the scientists posted a few posters showing design vs actual sensitivity and design sensitivity vs known pulsars. With the possible exception of the crab pulsar none are detectable with current hardware. The crab may be depending on where on the +-2 order of magnitude error bar the actual signal strength is and if there's enough separation between it's 54hz signal and the enormous (IIRC 4 order of magnitude) 60hz noise spike.

Advanced ligo's projected sensitivity should be high enough to readily detect several; and lisa should bag lots of them if it gets funded.

DanNeely
DanNeely
Joined: 4 Sep 05
Posts: 1364
Credit: 3573885105
RAC: 741009

Regarding the original

Regarding the original question, how long will ligo need to be shut down for the upgrades?

Running out of data isn't likely to be an issue. Even if the outage is very extended the current search app is only looking at the best few percent of the data; the analyzed dataset can always be expended to include some of the almost as good data. Alternately since the detection threshhold is (among other things) dependent on the length of the data samples being integrated in each WU new WU's with longer integration times could be created. Unless most of the existing data in WUs is chronologically continuous this would also mean increasing the fraction of the data crunched in total.

I wouldn't worry too much about running out of data unless the shutdown is extremely long.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.