I am crunching SETI data without any hope of finding an alien signal, Einstein data with very little hope of detecting a GW, QMC data with at least a little hope of being helpful. What I like is to feel being part of a cooperative effort beyond race, language, religion and other dividing motives in the hope of contributing to scientific advances not restricted only to professional scientists but also to the common people.
Tullio
So, the new S5R3 run will cover a wider range of frequency band. But, will it cover the same band as S5R1 and S5R2 with the new sensitivity (or, as you wish, a new number of parameters) or a new band besides previous searches?
Knock-knock. Are there anybody who knows about this?
Now, after reading all your suggestions and links I want to ask some new questions.
Does GW have such a parameter as wavelength? Can it be measured this way?
If it has, than it explains why we should analyse such a huge data material with months or even years of observation. I guess, because the wavelength is very high. Is it truth?
Now, after reading all your suggestions and links I want to ask some new questions.
Does GW have such a parameter as wavelength? Can it be measured this way?
If it has, than it explains why we should analyse such a huge data material with months or even years of observation. I guess, because the wavelength is very high. Is it truth?
Yes, gravity waves have the typical attributes of waves, namely
* polarization
* propagation speed (assumed to be the speed of light )
* frequency, and therefore wavelength
Frequency of a GW depends on the source that is emitting it. At E@H, we are looking specifically for GWs emitted by rotating neutron stars with "bumps". Here the frequency, and therefore wavelength, has a direct relationship to the frequency of the rotation of the neutron star, which, surprisingly for the layman, can be a four digit number of revolutions per second (!). I think the frequency of the GW would be twice the spin rate? Or half?? You get the idea, tho.
The first number in the name of the result you are crunching is exactly this frequency, eg. : h1_0540.85_S5R2__114_S5R2c_0 ==> result looks at a small frequency band near 540.85 Hz.
There are other (believed) sources of GWs with higher and lower frequencies, but the ground based laser-interferometers that provide the data we are all crunching are most sensitive in a frequency band that happens to lie within that of acoustic waves (say 10 Hz ... a few kHz), so if we ever find a GW, one would be able to convert this 1:1 into something we could "listen to".
(Douglas Adams would have used this to describe how alien civilizations use the modulation of neutron stars as kind of galactical FM stations :-). Maybe we should filter for riffs rather than sine waves).
To detect GWs with lower frequencies = larger wavelengths, space based interferometers are planned, as they do not suffer from seismic effects that limit sensitivity on the low frequency end here on Earth.
...To detect GWs with lower frequencies = larger wavelengths, space based interferometers are planned, as they do not suffer from seismic effects that limit sensitivity on the low frequency end here on Earth.
CU
H-BE
That is an excellent response! :)
If I may ask, why are these last workunits taking so much longer to process? It is because of more calculations of course, but is that a function of us analyzing a different wavelength, or because there are new algorithms being run as we close out this last phase of S5R2?
If I may ask, why are these last workunits taking so much longer to process? It is because of more calculations of course, but is that a function of us analyzing a different wavelength, or because there are new algorithms being run as we close out this last phase of S5R2?
Thanks!
The computation time of the workunits was known quite accurately at the start of the run already, there was no fundamental change in the algorithms. The fact that some frequency-bands take longer than others must lie in the nature of the filtering and pattern matching algorithms used, and this is far beyond my knowledge.
In S5R2, all workunits did a full-sky search. In S5R3, the sky can be divided among several workunits which allows a finer control of run times for each result.
....In S5R2, all workunits did a full-sky search. In S5R3, the sky can be divided among several workunits which allows a finer control of run times for each result.
CU
H-BE
Thanks! :)
In addition, it sounds like S5R3 will utilize some new algorithms as well, presumably to "dig" more information out of the data?
In addition, it sounds like S5R3 will utilize some new algorithms as well, presumably to "dig" more information out of the data?
No, not that I know of. The magic is in the parameters that are fed into the analysis software and the workunit generator in the first place, so S5R3 will look into a much wider area of the "search parameter space", where S5R2 was more like scouting around to test the terrain ... sort of... . Bernd will be able to explain this much better but is probably quite busy the next few days with the transition from S5R2 to S5R3.
Frequency of a GW depends on the source that is emitting it. At E@H, we are looking specifically for GWs emitted by rotating neutron stars with "bumps". Here the frequency, and therefore wavelength, has a direct relationship to the frequency of the rotation of the neutron star, which, surprisingly for the layman, can be a four digit number of revolutions per second (!). I think the frequency of the GW would be twice the spin rate? Or half?? You get the idea, tho.
We experience a well-understood gravity-related wave-effect here on Earth. I'm referring to tides. It is the waves in the water that give a name to the waves of electromagnetic energy we normally think of when we use the word "frequency." But the idea that gravity exerts a force that varies over time is not particularly new. But in the most familiar case, the pull of the Moon causes a gravity effect on Earth due to the rotation of the Earth rather than due to a modulation of the gravity "pull" itself. The resulting tidal effect is twice the frequency (twice per day) of the rotation of the Earth (once per day). I would thus find it surprising if the frequency of the gravity wave were not twice the rotational rate of the target object.
And by that I mean, does this statement by Dr. Allen on the main project page remain true?
"Einstein@Home is currently searching the most sensitive 840 hours of data from LIGO's first science run at design sensitivity (S5)."
We suspect it perhaps may remain true, because we are using the same data sets on our computers (at least initially) that were used with the final units of the S5R2 analysis.
That being said, 840 hours of data equals about 35 days, which is way less than the 182 days (6 months) of S5 data currently being collected, with data completion anticipated before the end of this year.
Question: If the 35 days of S5 data we are currently analyzing takes well over a year of S5R3 crunch time, then it would seem the remaining 5 months of S5 data available after the end of this year will take a very long time to crunch! If I am making proper assumptions, of course! ;)
I am crunching SETI data
)
I am crunching SETI data without any hope of finding an alien signal, Einstein data with very little hope of detecting a GW, QMC data with at least a little hope of being helpful. What I like is to feel being part of a cooperative effort beyond race, language, religion and other dividing motives in the hope of contributing to scientific advances not restricted only to professional scientists but also to the common people.
Tullio
RE: So, the new S5R3 run
)
Knock-knock. Are there anybody who knows about this?
Now, after reading all your
)
Now, after reading all your suggestions and links I want to ask some new questions.
Does GW have such a parameter as wavelength? Can it be measured this way?
If it has, than it explains why we should analyse such a huge data material with months or even years of observation. I guess, because the wavelength is very high. Is it truth?
RE: Now, after reading all
)
Yes, gravity waves have the typical attributes of waves, namely
* polarization
* propagation speed (assumed to be the speed of light )
* frequency, and therefore wavelength
Frequency of a GW depends on the source that is emitting it. At E@H, we are looking specifically for GWs emitted by rotating neutron stars with "bumps". Here the frequency, and therefore wavelength, has a direct relationship to the frequency of the rotation of the neutron star, which, surprisingly for the layman, can be a four digit number of revolutions per second (!). I think the frequency of the GW would be twice the spin rate? Or half?? You get the idea, tho.
The first number in the name of the result you are crunching is exactly this frequency, eg. : h1_0540.85_S5R2__114_S5R2c_0 ==> result looks at a small frequency band near 540.85 Hz.
There are other (believed) sources of GWs with higher and lower frequencies, but the ground based laser-interferometers that provide the data we are all crunching are most sensitive in a frequency band that happens to lie within that of acoustic waves (say 10 Hz ... a few kHz), so if we ever find a GW, one would be able to convert this 1:1 into something we could "listen to".
(Douglas Adams would have used this to describe how alien civilizations use the modulation of neutron stars as kind of galactical FM stations :-). Maybe we should filter for riffs rather than sine waves).
To detect GWs with lower frequencies = larger wavelengths, space based interferometers are planned, as they do not suffer from seismic effects that limit sensitivity on the low frequency end here on Earth.
CU
H-BE
RE: ...To detect GWs with
)
That is an excellent response! :)
If I may ask, why are these last workunits taking so much longer to process? It is because of more calculations of course, but is that a function of us analyzing a different wavelength, or because there are new algorithms being run as we close out this last phase of S5R2?
Thanks!
RE: If I may ask, why are
)
The computation time of the workunits was known quite accurately at the start of the run already, there was no fundamental change in the algorithms. The fact that some frequency-bands take longer than others must lie in the nature of the filtering and pattern matching algorithms used, and this is far beyond my knowledge.
In S5R2, all workunits did a full-sky search. In S5R3, the sky can be divided among several workunits which allows a finer control of run times for each result.
CU
H-BE
RE: ....In S5R2, all
)
Thanks! :)
In addition, it sounds like S5R3 will utilize some new algorithms as well, presumably to "dig" more information out of the data?
RE: In addition, it sounds
)
No, not that I know of. The magic is in the parameters that are fed into the analysis software and the workunit generator in the first place, so S5R3 will look into a much wider area of the "search parameter space", where S5R2 was more like scouting around to test the terrain ... sort of... . Bernd will be able to explain this much better but is probably quite busy the next few days with the transition from S5R2 to S5R3.
CU
H-BE
RE: Frequency of a GW
)
We experience a well-understood gravity-related wave-effect here on Earth. I'm referring to tides. It is the waves in the water that give a name to the waves of electromagnetic energy we normally think of when we use the word "frequency." But the idea that gravity exerts a force that varies over time is not particularly new. But in the most familiar case, the pull of the Moon causes a gravity effect on Earth due to the rotation of the Earth rather than due to a modulation of the gravity "pull" itself. The resulting tidal effect is twice the frequency (twice per day) of the rotation of the Earth (once per day). I would thus find it surprising if the frequency of the gravity wave were not twice the rotational rate of the target object.
What data set are we using on
)
What data set are we using on the new S5R3 run?
And by that I mean, does this statement by Dr. Allen on the main project page remain true?
"Einstein@Home is currently searching the most sensitive 840 hours of data from LIGO's first science run at design sensitivity (S5)."
We suspect it perhaps may remain true, because we are using the same data sets on our computers (at least initially) that were used with the final units of the S5R2 analysis.
That being said, 840 hours of data equals about 35 days, which is way less than the 182 days (6 months) of S5 data currently being collected, with data completion anticipated before the end of this year.
Question: If the 35 days of S5 data we are currently analyzing takes well over a year of S5R3 crunch time, then it would seem the remaining 5 months of S5 data available after the end of this year will take a very long time to crunch! If I am making proper assumptions, of course! ;)