From a www.theregister.co.uk news item the LargeSynopticSurveyTelescope to be built in Chile will capture 30 TB of data every night. How are they going to process all this data? Clearly a BOINC project is needed.
Tullio
There is a new status update on the Gravity Probe B web-site (which is actually more than a month old). What does "a 15% statistical uncertainty for the Frame-Dragging effect" mean? Is it the length of the missing pieces on the plots for the four gyroscopes?
"Everything should be made as simple as possible, but not simpler." A. Einstein
There is a new status update on the Gravity Probe B web-site (which is actually more than a month old). What does "a 15% statistical uncertainty for the Frame-Dragging effect" mean? Is it the length of the missing pieces on the plots for the four gyroscopes?
In my understanding the frame dragging effect is elicited by an accumulating difference in the direction of a guide star compared with the inertial behaviour of the gyros, with four giving redundancy. This has a number of attributable components from well known classical effects through to the general relativistic part being sought. Looking at
then from the first gyro down the presumed GR effect is ~ 40, 23, 18, 29 in the relevant units over the period shown - estimated by my eye. The gives a mean of 110/4 = 27.5 with standard deviation of 7.6 ( ~ 25% ), but that's with a simple equal weighting of the gyros. So I'd guess at the 15% being quoted as a measure of the spread of results over the gyros.
I guess they are so pleased as to get a result under unity ( 100% ) error!! :-)
Cheers, Mike.
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
Mike, I had a bit of difficulty understanding your post so I decided to look into it some more. The project set out with the goal of measuring frame-dragging with an uncertainty of 1% - for a while it looked like they weren't going to be able to make any statements about it, but recent breakthroughs in understanding their difficulties have allowed them to get it down to 15%. From what I read there's still a lot of work to be done and further analyses will be presented in three phases - Phase I in April, phase II in May and phase III in July if I'm not mistaken. (a final date of June 2010 is cited in the December 2008 paper to NASA, but I'm not sure how these dates are related exactly or even what year the 25 Februari 2009 presentation is referring to) Now that they understand the situation they're also speeding up their algorithms and preparing to throw more computing power at it, with a major goal being to greatly reduce the time steps between the data points used for analysis. (down to 2 seconds from once per 97 minute orbit, IIRC) They expect a final uncertainty of 3 to 5%.
Mike, I had a bit of difficulty understanding your post so I decided to look into it some more. The project set out with the goal of measuring frame-dragging with an uncertainty of 1% - for a while it looked like they weren't going to be able to make any statements about it, but recent breakthroughs in understanding their difficulties have allowed them to get it down to 15%. From what I read there's still a lot of work to be done and further analyses will be presented in three phases - Phase I in April, phase II in May and phase III in July if I'm not mistaken. Now that they understand the situation they're also speeding up their algorithms and preparing to throw more computing power at it, with a major goal being to greatly reduce the time steps between the data points used for analysis. (down to 2 seconds from once per orbit, IIRC)
Frame dragging in effect means the direction to a distant star ( well outside of Earth's gravity field ) will appear to wander when seen from well inside Earth's gravity field. General Relativity doesn't measure absolutes - only makes statements that compare observations between reference frames. A gyroscope will/ought maintain a given direction for it's spin axis according to the law of inertia, except that ( via GR's Equivalence Principle ) that 'inertia' is understood to mean as per local spacetime. Now the law of inertia means 'without any disturbance' ( which is a way of defining forces in a frame ), so to measure the distortion of spacetime alone ( a 'force free' path ) alot of other factors that influence a gyroscope need to be subtracted away. This is what their analysis is attempting to model and understand. It looks like electrical effects are the major bugbear to account for.
The general idea of the project is that after orbital insertion to try and fly the spacecraft around the gyroscopes ( each now a 'drag free' test mass ) without 'touching' them so that they follow the local spacetime paths which are attributable to gravity causing deviation or 'warping'. However four gyroscopes, while nicely redundant, can't all be in exactly the same place so any path which suits one of the gyroscopes for mapping it's 'untouched' path will necessarily be not-quite-right for the others as, strictly speaking, there will be a slight difference in the gravity field ( a 'gradient' ) within the craft itself. This induces a correlation between what's measured with one gyroscope to another.
Cheers, Mike.
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
From a www.theregister.co.uk
)
From a www.theregister.co.uk news item the LargeSynopticSurveyTelescope to be built in Chile will capture 30 TB of data every night. How are they going to process all this data? Clearly a BOINC project is needed.
Tullio
RE: There is a recent
)
That is a nice site. If GP-B can help shed light on inertia I may finally know why I am so lazy!
Latest news on Gravity Probe
)
Latest news on Gravity Probe B:
IEEE Spectrum
Tullio
There is a new status update
)
There is a new status update on the Gravity Probe B web-site (which is actually more than a month old). What does "a 15% statistical uncertainty for the Frame-Dragging effect" mean? Is it the length of the missing pieces on the plots for the four gyroscopes?
"Everything should be made as simple as possible, but not simpler." A. Einstein
RE: There is a new status
)
In my understanding the frame dragging effect is elicited by an accumulating difference in the direction of a guide star compared with the inertial behaviour of the gyros, with four giving redundancy. This has a number of attributable components from well known classical effects through to the general relativistic part being sought. Looking at
then from the first gyro down the presumed GR effect is ~ 40, 23, 18, 29 in the relevant units over the period shown - estimated by my eye. The gives a mean of 110/4 = 27.5 with standard deviation of 7.6 ( ~ 25% ), but that's with a simple equal weighting of the gyros. So I'd guess at the 15% being quoted as a measure of the spread of results over the gyros.
I guess they are so pleased as to get a result under unity ( 100% ) error!! :-)
Cheers, Mike.
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
Mike, I had a bit of
)
Mike, I had a bit of difficulty understanding your post so I decided to look into it some more. The project set out with the goal of measuring frame-dragging with an uncertainty of 1% - for a while it looked like they weren't going to be able to make any statements about it, but recent breakthroughs in understanding their difficulties have allowed them to get it down to 15%. From what I read there's still a lot of work to be done and further analyses will be presented in three phases - Phase I in April, phase II in May and phase III in July if I'm not mistaken. (a final date of June 2010 is cited in the December 2008 paper to NASA, but I'm not sure how these dates are related exactly or even what year the 25 Februari 2009 presentation is referring to) Now that they understand the situation they're also speeding up their algorithms and preparing to throw more computing power at it, with a major goal being to greatly reduce the time steps between the data points used for analysis. (down to 2 seconds from once per 97 minute orbit, IIRC) They expect a final uncertainty of 3 to 5%.
RE: Mike, I had a bit of
)
Frame dragging in effect means the direction to a distant star ( well outside of Earth's gravity field ) will appear to wander when seen from well inside Earth's gravity field. General Relativity doesn't measure absolutes - only makes statements that compare observations between reference frames. A gyroscope will/ought maintain a given direction for it's spin axis according to the law of inertia, except that ( via GR's Equivalence Principle ) that 'inertia' is understood to mean as per local spacetime. Now the law of inertia means 'without any disturbance' ( which is a way of defining forces in a frame ), so to measure the distortion of spacetime alone ( a 'force free' path ) alot of other factors that influence a gyroscope need to be subtracted away. This is what their analysis is attempting to model and understand. It looks like electrical effects are the major bugbear to account for.
The general idea of the project is that after orbital insertion to try and fly the spacecraft around the gyroscopes ( each now a 'drag free' test mass ) without 'touching' them so that they follow the local spacetime paths which are attributable to gravity causing deviation or 'warping'. However four gyroscopes, while nicely redundant, can't all be in exactly the same place so any path which suits one of the gyroscopes for mapping it's 'untouched' path will necessarily be not-quite-right for the others as, strictly speaking, there will be a slight difference in the gravity field ( a 'gradient' ) within the craft itself. This induces a correlation between what's measured with one gyroscope to another.
Cheers, Mike.
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
Read this latest update today
)
Read this latest update today - figured it was worth some thread necromancy :)
RE: Read this latest update
)
Thanks for that.
There's some phenomenal modelling and processing to unscramble that lot.
... If only they had made the rotors and housing conductive?...
Regards,
Martin
See new freedom: Mageia Linux
Take a look for yourself: Linux Format
The Future is what We all make IT (GPLv3)
RE: Read this latest update
)
It's an impressive analysis for sure. GR-wise it's a higher order test and thus quite subtle in low field strength ie. around here.
Cheers, Mike.
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