| |
Main
Date: 05 Aug 2006 10:53:09
From: BlagooBlanaa
Subject: Hubble Bubble Trouble
|
http://www.spaceref.com/news/viewpr.nl.html?pid=20539
Turns out that H0 may be smaller than thought - in one
direction anyways - towards M33.
Treating H0 as an homogeneous and isotropic constant is
about as smart as looking for dark matter, and finding it...
H0 is not scalar, and will vary depending on direction and epoch.
Get this right and dark matter disappears
cheers
|
|
| |
Date: 04 Aug 2006 18:04:40
From: Brian Tung
Subject: Re: Hubble Bubble Trouble
|
BlagooBlanaa wrote:
> http://www.spaceref.com/news/viewpr.nl.html?pid=20539
Funny. I already had the distance to M33 at 3 million light-years. I
could have scooped these guys.
More seriously, not all methods had the distance at 2.6 million light-
years. Some already measured it at 3 million.
> Turns out that H0 may be smaller than thought - in one
> direction anyways - towards M33.
Is M33 far enough to judge H0? We're only talking an error of 400,000
light-years. A difference of 15 percent would be more significant if
it applied to a galaxy tens of millions of light-years away.
> Treating H0 as an homogeneous and isotropic constant is
> about as smart as looking for dark matter, and finding it...
>
> H0 is not scalar, and will vary depending on direction and epoch.
> Get this right and dark matter disappears
Are you thinking of dark energy, perhaps? Dark matter is attested on
scales much smaller than those the Hubble constant applies to.
But even in that case, your argument fails. If it were true, then we'd
see some kind of systematic variation in Type Ia supernovae. In some
directions, they would be brighter than expected, and in some, they'd be
dimmer than expected. But in fact, they're dimmer than expected in
all directions, and in a way that varies smoothly and non-linearly by
distance. Because that effect is isotropic, variations in H0 by
direction can't make dark energy go away entirely; they can only suggest
changes in distribution. The fact that it's non-linear (even after
correcting for expansion) means that evenly distributed dust can't
explain all the extinction, a conclusion supported also by WMAP results.
But I'm sure you will continue believing that there is a conspiracy in
the ultra-high-paying field of cosmology. I expect to hear any day now
about Bill Keel's massive tithing campaign.
--
Brian Tung <brian@isi.edu >
The Astronomy Corner at http://astro.isi.edu/
Unofficial C5+ Home Page at http://astro.isi.edu/c5plus/
The PleiadAtlas Home Page at http://astro.isi.edu/pleiadatlas/
My Own Personal FAQ (SAA) at http://astro.isi.edu/reference/faq.html
|
| | |
Date: 05 Aug 2006 12:03:18
From: BlagooBlanaa
Subject: Re: Hubble Bubble Trouble
|
"Brian Tung" <brian@isi.edu > wrote in message
news:eb0qr8$e5m$1@praesepe.isi.edu...
> BlagooBlanaa wrote:
>> http://www.spaceref.com/news/viewpr.nl.html?pid=20539
>
> Funny. I already had the distance to M33 at 3 million light-years. I
> could have scooped these guys.
>
> More seriously, not all methods had the distance at 2.6 million light-
> years. Some already measured it at 3 million.
>
>> Turns out that H0 may be smaller than thought - in one
>> direction anyways - towards M33.
>
> Is M33 far enough to judge H0? We're only talking an error of 400,000
> light-years. A difference of 15 percent would be more significant if
> it applied to a galaxy tens of millions of light-years away.
>
the experimenters reckon they have 6% accuracy
>> Treating H0 as an homogeneous and isotropic constant is
>> about as smart as looking for dark matter, and finding it...
>>
>> H0 is not scalar, and will vary depending on direction and epoch.
>> Get this right and dark matter disappears
>
> Are you thinking of dark energy, perhaps? Dark matter is attested on
> scales much smaller than those the Hubble constant applies to.
>
attested? mighty strong words pardner
suspected, maybe... proven - not yet!
I would like to check the science we already have, without
resorting to dark doodads
> But even in that case, your argument fails. If it were true, then we'd
> see some kind of systematic variation in Type Ia supernovae. In some
> directions, they would be brighter than expected, and in some, they'd be
> dimmer than expected. But in fact, they're dimmer than expected in
> all directions, and in a way that varies smoothly and non-linearly by
> distance. Because that effect is isotropic, variations in H0 by
> direction can't make dark energy go away entirely; they can only suggest
> changes in distribution. The fact that it's non-linear (even after
> correcting for expansion) means that evenly distributed dust can't
> explain all the extinction, a conclusion supported also by WMAP results.
>
smoothly - hmmmm
have a look at the data processing pipeline for all the surveys that crank
out H0
they do an explicit and implicit smoothing of data to accomplish a nice
value
for H0 because they expect it to be "the same" in all directions
and how much do we really know about type 1a supernovae anyways?
how much dust is there? local? sure about that?
planets, ecliptic inclinations, oort clouds hmmmm?
I would love to see large scale survey data "deconvolved" of the assumption
that H0 is homogeneous and isotropic...
I would also like to see time resolved surveys
Say what are the resonant modes of the Universe due to remnant BB grav
waves?
> But I'm sure you will continue believing that there is a conspiracy in
> the ultra-high-paying field of cosmology. I expect to hear any day now
> about Bill Keel's massive tithing campaign.
The only conspiracy is the necessity to publish results that 'toe the party
line'
And cosmologists with a family and a mortgage are just as susceptible to
fudging in a zeitgeisty way as well as any other wage slave...
And everyone uses the same data reduction software and techniques don't
they?
IRAF MIDAS etc cross correlation, FFT etc , arc lamps, ccd's
standard stars from the HD catalogue, blah blah blah
Vega anyone?
I would love to see dark matter proven/disproven
H0 or H(f(x,y,z,t........)) proven/disproven
It is the rather convenient survey/jobs program that has sprung up
around H0 that may well be cranking out crap science that bothers me.
And who has actually done any real testing of the tools that do the data
reduction?
NIST???? I dunno - but I think that there should be!
>
> --
> Brian Tung <brian@isi.edu>
> The Astronomy Corner at http://astro.isi.edu/
> Unofficial C5+ Home Page at http://astro.isi.edu/c5plus/
> The PleiadAtlas Home Page at http://astro.isi.edu/pleiadatlas/
> My Own Personal FAQ (SAA) at http://astro.isi.edu/reference/faq.html
|
| | | |
Date: 05 Aug 2006 00:45:14
From: George
Subject: Re: Hubble Bubble Trouble
|
"BlagooBlanaa" <BlagooBlanaa@hotmail.com > wrote in message
news:44d3fc0e$0$24756$afc38c87@news.optusnet.com.au...
>
> "Brian Tung" <brian@isi.edu> wrote in message
> news:eb0qr8$e5m$1@praesepe.isi.edu...
>> BlagooBlanaa wrote:
>>> http://www.spaceref.com/news/viewpr.nl.html?pid=20539
>>
>> Funny. I already had the distance to M33 at 3 million light-years. I
>> could have scooped these guys.
>>
>> More seriously, not all methods had the distance at 2.6 million light-
>> years. Some already measured it at 3 million.
>>
>>> Turns out that H0 may be smaller than thought - in one
>>> direction anyways - towards M33.
>>
>> Is M33 far enough to judge H0? We're only talking an error of 400,000
>> light-years. A difference of 15 percent would be more significant if
>> it applied to a galaxy tens of millions of light-years away.
>>
>
> the experimenters reckon they have 6% accuracy
Umm, actually, their margin of error is 6%, which means that their
measurements have 94% precision.
George
|
| | | | |
Date: 05 Aug 2006 21:46:41
From: BlagooBlanaa
Subject: Re: Hubble Bubble Trouble
|
"George" <george@yourservice.com > wrote in message
news:Kv-dnSTvZOnHv0nZnZ2dnUVZ_s6dnZ2d@insightbb.com...
>
> "BlagooBlanaa" <BlagooBlanaa@hotmail.com> wrote in message
> news:44d3fc0e$0$24756$afc38c87@news.optusnet.com.au...
>>
>> "Brian Tung" <brian@isi.edu> wrote in message
>> news:eb0qr8$e5m$1@praesepe.isi.edu...
>>> BlagooBlanaa wrote:
>>>> http://www.spaceref.com/news/viewpr.nl.html?pid=20539
>>>
>>> Funny. I already had the distance to M33 at 3 million light-years. I
>>> could have scooped these guys.
>>>
>>> More seriously, not all methods had the distance at 2.6 million light-
>>> years. Some already measured it at 3 million.
>>>
>>>> Turns out that H0 may be smaller than thought - in one
>>>> direction anyways - towards M33.
>>>
>>> Is M33 far enough to judge H0? We're only talking an error of 400,000
>>> light-years. A difference of 15 percent would be more significant if
>>> it applied to a galaxy tens of millions of light-years away.
>>>
>>
>> the experimenters reckon they have 6% accuracy
>
> Umm, actually, their margin of error is 6%, which means that their
> measurements have 94% precision.
>
correct
> George
>
|
| | | | | |
Date: 05 Aug 2006 08:46:58
From: William Hamblen
Subject: Re: Hubble Bubble Trouble
|
On 2006-08-05, BlagooBlanaa <BlagooBlanaa@hotmail.com > wrote:
>
> "George" <george@yourservice.com> wrote in message
> news:Kv-dnSTvZOnHv0nZnZ2dnUVZ_s6dnZ2d@insightbb.com...
>>
>> "BlagooBlanaa" <BlagooBlanaa@hotmail.com> wrote in message
>> news:44d3fc0e$0$24756$afc38c87@news.optusnet.com.au...
>>>
>>> "Brian Tung" <brian@isi.edu> wrote in message
>>> news:eb0qr8$e5m$1@praesepe.isi.edu...
>>>> BlagooBlanaa wrote:
>>>>> http://www.spaceref.com/news/viewpr.nl.html?pid=20539
>>>>
>>>> Funny. I already had the distance to M33 at 3 million light-years. I
>>>> could have scooped these guys.
>>>>
>>>> More seriously, not all methods had the distance at 2.6 million light-
>>>> years. Some already measured it at 3 million.
>>>>
>>>>> Turns out that H0 may be smaller than thought - in one
>>>>> direction anyways - towards M33.
>>>>
>>>> Is M33 far enough to judge H0? We're only talking an error of 400,000
>>>> light-years. A difference of 15 percent would be more significant if
>>>> it applied to a galaxy tens of millions of light-years away.
>>>>
>>>
>>> the experimenters reckon they have 6% accuracy
>>
>> Umm, actually, their margin of error is 6%, which means that their
>> measurements have 94% precision.
>>
> correct
>
No, what this usually means is that the investigators believe
that the true value is within 6% of their result with some level
of confidence (usually 95% or within 2 standard deviations).
Bud
|
| | | | | | |
Date: 06 Aug 2006 00:30:41
From: BlagooBlanaa
Subject: Re: Hubble Bubble Trouble
|
"William Hamblen" <wrhamblen@comcast.net > wrote in message
news:MsqdnbXq0aLPPEnZnZ2dnUVZ_rKdnZ2d@comcast.com...
> On 2006-08-05, BlagooBlanaa <BlagooBlanaa@hotmail.com> wrote:
>>
>> "George" <george@yourservice.com> wrote in message
>> news:Kv-dnSTvZOnHv0nZnZ2dnUVZ_s6dnZ2d@insightbb.com...
>>>
>>> "BlagooBlanaa" <BlagooBlanaa@hotmail.com> wrote in message
>>> news:44d3fc0e$0$24756$afc38c87@news.optusnet.com.au...
>>>>
>>>> "Brian Tung" <brian@isi.edu> wrote in message
>>>> news:eb0qr8$e5m$1@praesepe.isi.edu...
>>>>> BlagooBlanaa wrote:
>>>>>> http://www.spaceref.com/news/viewpr.nl.html?pid=20539
>>>>>
>>>>> Funny. I already had the distance to M33 at 3 million light-years. I
>>>>> could have scooped these guys.
>>>>>
>>>>> More seriously, not all methods had the distance at 2.6 million light-
>>>>> years. Some already measured it at 3 million.
>>>>>
>>>>>> Turns out that H0 may be smaller than thought - in one
>>>>>> direction anyways - towards M33.
>>>>>
>>>>> Is M33 far enough to judge H0? We're only talking an error of 400,000
>>>>> light-years. A difference of 15 percent would be more significant if
>>>>> it applied to a galaxy tens of millions of light-years away.
>>>>>
>>>>
>>>> the experimenters reckon they have 6% accuracy
>>>
>>> Umm, actually, their margin of error is 6%, which means that their
>>> measurements have 94% precision.
>>>
>> correct
>>
>
> No, what this usually means is that the investigators believe
> that the true value is within 6% of their result with some level
> of confidence (usually 95% or within 2 standard deviations).
>
correct
no -one has stated precisely how the uncertainty was estimated
or quoted - 1 sigma, 2 sigma (assuming normal distribution)
blah blah blah relative error, absolute error blah blah blah
doesn't really matter does it - quite frankly I find most error
estimates quoted in astronomical circles to be literally pi in the sky
:))
ps: you would be amazed and hoorified to see an FFT based cross correlation
whip around in response to poisson distributed data contaminated by by
random
and correlated noise.
if you carefully do this (go on use IRAF and try out some simple red shift
calculations
on synthetic spectra in the presence of noise) you will see that a lot of
the error
estimates you see are just rank utter bosh - as are the velocity
estimates...
go on try it!
> Bud
|
| | | | | | |
Date: 05 Aug 2006 17:14:39
From: George
Subject: Re: Hubble Bubble Trouble
|
"William Hamblen" <wrhamblen@comcast.net > wrote in message
news:MsqdnbXq0aLPPEnZnZ2dnUVZ_rKdnZ2d@comcast.com...
> On 2006-08-05, BlagooBlanaa <BlagooBlanaa@hotmail.com> wrote:
>>
>> "George" <george@yourservice.com> wrote in message
>> news:Kv-dnSTvZOnHv0nZnZ2dnUVZ_s6dnZ2d@insightbb.com...
>>>
>>> "BlagooBlanaa" <BlagooBlanaa@hotmail.com> wrote in message
>>> news:44d3fc0e$0$24756$afc38c87@news.optusnet.com.au...
>>>>
>>>> "Brian Tung" <brian@isi.edu> wrote in message
>>>> news:eb0qr8$e5m$1@praesepe.isi.edu...
>>>>> BlagooBlanaa wrote:
>>>>>> http://www.spaceref.com/news/viewpr.nl.html?pid=20539
>>>>>
>>>>> Funny. I already had the distance to M33 at 3 million light-years.
>>>>> I
>>>>> could have scooped these guys.
>>>>>
>>>>> More seriously, not all methods had the distance at 2.6 million
>>>>> light-
>>>>> years. Some already measured it at 3 million.
>>>>>
>>>>>> Turns out that H0 may be smaller than thought - in one
>>>>>> direction anyways - towards M33.
>>>>>
>>>>> Is M33 far enough to judge H0? We're only talking an error of
>>>>> 400,000
>>>>> light-years. A difference of 15 percent would be more significant if
>>>>> it applied to a galaxy tens of millions of light-years away.
>>>>>
>>>>
>>>> the experimenters reckon they have 6% accuracy
>>>
>>> Umm, actually, their margin of error is 6%, which means that their
>>> measurements have 94% precision.
>>>
>> correct
>>
>
> No, what this usually means is that the investigators believe
> that the true value is within 6% of their result with some level
> of confidence (usually 95% or within 2 standard deviations).
>
> Bud
Right. As I said, their measurements have 94% precision.
|
| | | |
Date: 04 Aug 2006 21:17:06
From: Brian Tung
Subject: Re: Hubble Bubble Trouble
|
BlagooBlanaa wrote:
> > Is M33 far enough to judge H0? We're only talking an error of 400,000
> > light-years. A difference of 15 percent would be more significant if
> > it applied to a galaxy tens of millions of light-years away.
>
> the experimenters reckon they have 6% accuracy
Missed my point. I'm not doubting their accuracy. I'm doubting that
it has relevance to the Hubble constant. The Andromeda Galaxy, which
is only 20 percent closer, perhaps, is actually approaching us. By a
strict application of Hubble's Law, that would mean the distance to it
is negative. Obviously, that's not true. I'm certain a similar kind of
consideration applies to M33; applying Hubble's Law to it just doesn't
make sense.
> attested? mighty strong words pardner
> suspected, maybe... proven - not yet!
Attested does not proven. It just means evidence shown for. Sorry if
there was any confusion about that.
> smoothly - hmmmm
> have a look at the data processing pipeline for all the surveys that
> crank out H0 they do an explicit and implicit smoothing of data to
> accomplish a nice value for H0 because they expect it to be "the
> same" in all directions
No, they do not. In the original source papers, they give data points
plus confidence bars (typically 95 percent). They actually give several
smooth curves, which are derived from theoretical models, including one
which has no dark energy in it at all. It can be seen from these graphs
that the data, taking into account the confidence intervals, match
curves that include dark energy considerably better than they match the
ones without.
Of course, when they print redactions in places like Sky and Telescope,
Scientific American, American Scientist (though less so here), they
adjust for their audience. They remove some of these aspects.
> and how much do we really know about type 1a supernovae anyways?
> how much dust is there? local? sure about that?
You missed my comment about the dust. The dust would have to be
distributed unevenly in order to be sufficient *on its own* to account
for the effects.
> The only conspiracy is the necessity to publish results that 'toe the party
> line'...[snip]
Thanks for confirming my prediction.
--
Brian Tung <brian@isi.edu >
The Astronomy Corner at http://astro.isi.edu/
Unofficial C5+ Home Page at http://astro.isi.edu/c5plus/
The PleiadAtlas Home Page at http://astro.isi.edu/pleiadatlas/
My Own Personal FAQ (SAA) at http://astro.isi.edu/reference/faq.html
|
| | | | |
Date: 05 Aug 2006 22:01:03
From: BlagooBlanaa
Subject: Re: Hubble Bubble Trouble
|
"Brian Tung" <brian@isi.edu > wrote in message
news:eb1642$f64$1@praesepe.isi.edu...
> BlagooBlanaa wrote:
>> > Is M33 far enough to judge H0? We're only talking an error of 400,000
>> > light-years. A difference of 15 percent would be more significant if
>> > it applied to a galaxy tens of millions of light-years away.
>>
>> the experimenters reckon they have 6% accuracy
>
> Missed my point. I'm not doubting their accuracy. I'm doubting that
> it has relevance to the Hubble constant. The Andromeda Galaxy, which
> is only 20 percent closer, perhaps, is actually approaching us. By a
> strict application of Hubble's Law, that would mean the distance to it
> is negative. Obviously, that's not true. I'm certain a similar kind of
> consideration applies to M33; applying Hubble's Law to it just doesn't
> make sense.
>
lets not get too free in applying an averaged out parameter here :)
this is part of the problem isn't it?
when we talk about galaxy recession, we talk about galaxies as though
they were 'points' - well obviously they are not! if we look at Type 1a
supernovae as markers there is obviuosly some consideration to be made
for where the progenitor star of the type 1a was in the parent galaxy, etc.
Obviously when the galaxy is close - andromeda say, it is ludicrous to
apply H0 to it no?
>> attested? mighty strong words pardner
>> suspected, maybe... proven - not yet!
>
> Attested does not proven. It just means evidence shown for. Sorry if
> there was any confusion about that.
>
>> smoothly - hmmmm
>> have a look at the data processing pipeline for all the surveys that
>> crank out H0 they do an explicit and implicit smoothing of data to
>> accomplish a nice value for H0 because they expect it to be "the
>> same" in all directions
>
> No, they do not. In the original source papers, they give data points
> plus confidence bars (typically 95 percent). They actually give several
> smooth curves, which are derived from theoretical models, including one
> which has no dark energy in it at all. It can be seen from these graphs
> that the data, taking into account the confidence intervals, match
> curves that include dark energy considerably better than they match the
> ones without.
hah - you are naive about how professional researchers go about selecting
'outliers' in their data...
>
> Of course, when they print redactions in places like Sky and Telescope,
> Scientific American, American Scientist (though less so here), they
> adjust for their audience. They remove some of these aspects.
hell yeah - they remove all the really interesting stuff!
>
>> and how much do we really know about type 1a supernovae anyways?
>> how much dust is there? local? sure about that?
>
> You missed my comment about the dust. The dust would have to be
> distributed unevenly in order to be sufficient *on its own* to account
> for the effects.
so what does this tell you - what do you *know* about dust around type 1a
supernovae hmmm?
>
>> The only conspiracy is the necessity to publish results that 'toe the
>> party
>> line'...[snip]
>
> Thanks for confirming my prediction.
>
perhaps I should have put 'conspiracy' in quotes Brian - so you may have
felt less justified in jumping on it...
Just as a little side note - from another branch of scientific endeavour to
be sure -
what is the causative agent of stomach ulcers?
let me answer - Helicobacter Pylorii - a bug... It took decades to get this
bug recognized
because the Medical Fraternity pooh-poohed any research along those lines -
the conventional
wisdom was that ulcers were caused by 'stress', and anyway the standard
stains and tests did
not show a bug - so obviously there couldn't be any there....
I claim no such 'conspiracy' save the normal cowardice extant in the career
researcher...
> --
> Brian Tung <brian@isi.edu>
> The Astronomy Corner at http://astro.isi.edu/
> Unofficial C5+ Home Page at http://astro.isi.edu/c5plus/
> The PleiadAtlas Home Page at http://astro.isi.edu/pleiadatlas/
> My Own Personal FAQ (SAA) at http://astro.isi.edu/reference/faq.html
|
| | | | | |
Date: 05 Aug 2006 08:16:25
From: Brian Tung
Subject: Re: Hubble Bubble Trouble
|
BlagooBlanaa wrote:
> Obviously when the galaxy is close - andromeda say, it is ludicrous to
> apply H0 to it no?
Yes, that was my point.
> hah - you are naive about how professional researchers go about selecting
> 'outliers' in their data...
Actually, I am a professional researcher. But that's OK, I'm part of
the conspiracy.
You live in a strange world.
--
Brian Tung <brian@isi.edu >
The Astronomy Corner at http://astro.isi.edu/
Unofficial C5+ Home Page at http://astro.isi.edu/c5plus/
The PleiadAtlas Home Page at http://astro.isi.edu/pleiadatlas/
My Own Personal FAQ (SAA) at http://astro.isi.edu/reference/faq.html
|
| | | | | | |
Date: 07 Aug 2006 11:54:42
From: BlagooBlanaa
Subject: Re: Hubble Bubble Trouble
|
> You live in a strange world.
Indeed, as do we all.
Great isn't it?
And how well have you calibrated your research tools Brian?
As a minimum I would suggest that NIST run some tests on IRAF
(different versions too!) and other tools used to determine red shift
and see if the precisions and accuracies quoted in learned journals
are really justified.
How robust is a least squares fit? How robust is a robust estimator?
Just exactly how much blurring does a survey data processing pipeline
do anyway? Its a lot of work, and the work is not done by the Prof
(thankfully) but by others using tools whose correct realm of
applicability is lost in the sands of time (zero padding???? windowing)
There is a lot of scope (sorry) for data with real science in it to be
blurred and averaged into a nice safe tenured job.
There is also a lot of scope for noisy data to be processed to yield
any damn thing the Prof wants...
cheers
|
| |
Date: 05 Aug 2006 02:41:49
From: canopus56
Subject: Re: Hubble Bubble Trouble
|
Brian Tung wrote:
<snip all >
Here's the pre-print on Banonos' article:
http://arxiv.org/PS_cache/astro-ph/pdf/0606/0606279.pdf
It's pretty amazing that using 1.2 through 10 meter telescopes, they
can image eclipsing binaries _in other galaxies_.
Suffice it to say that one data point a statistically significant trend
does not make. I'll wait for the 31st measured star before deciding
that HST Team Hubble constant needs another revision.
- Canopus56
|
|
