Saturday, 1 December 2007

Auger, Centaurus and Virgo

I wrote there was no interesting seminar last week, but i should mention there was an interesting pre-seminar discussion before the Wednesday Cosmo Coffe. The authors of this article were yelling at the speaker of that seminar. The latter is a member of the Auger collaboration and the former just submitted a comment to ArXiv, in which they put in doubt some of the conclusions obtained by Auger.

The new results from the Pierre Auger Observatory were announced a month ago (see also the post on Backreation). Auger looked at cosmic rays with with ultra-high energies, above $6\cdot 10^{19}$ eV. By theoretical arguments, 90% of such high-energy particles should originate from sources less than 200 Mpc away. This is because of the GZK cut-off -- ultra-high energy particles scatter on the CMB photons, thus losing their energy. Auger claimed to observe a correlation between the arrival direction of the ultra-high energy cosmic rays and the positions of the nearby Active Galactic Nuclei (AGN). This would mean that we finally pinpointed who shoots these tennisball-energy particles at us.

In the aforementioned comment, Igor Tkachev i yego komanda points out that Auger in their statistical analysis did not take into account the $1/r^2$ decrease of the flux with the distance to the observer. The claim is that, once the decrease is taken into account, the AGN hypothesis is disfavored at 99% confidence level. The problem is illustrated on the figure to the right. The crosses mark the positions of the nearby AGN, the color shades indicate the expected flux of the ultra-high energy cosmic rays (if the AGN hypothesis is true) and the circles denote the hits registered by Auger. One can see that most of the ultra-high energy particles arrive from the direction of the Centaur supercluster, while none arrive from the Virgo supercluster. The latter is in fact closer to us, and we would expect at least as many hits from that direction. According to the authors of the comments, the more likely hypothesis is that there is a bright source of the cosmic rays that lies in the direction of the Centaurus supercluster.

It seems that we have to wait for more data to finally resolve the cosmic ray puzzle. In the meantime, here is the local map i found while preparing this post. Just in case you need to find your way home...


Anonymous said...

They did "not take into account the $$1/r^2$$ decrease of the flux with the distance to the observer"?!?!?

I'm flabbergasted. After this, can ANY claim coming from the astro people be taken seriously?

Anonymous said...

There was a time when scientists sought broad review by trusted associates before publishing anything. That avoided most of these publicly embarrassing errors. These days, the rush to publish is more important than reputation. It seems everyone wants to be the "Paris Hilton" of science now, where celebrity is more important than reputation. I blame it on the string theorists, since their goofy ideas have taken over the press while still full of an almost infinite number of errors.

Jester said...

I also choked when i heard it, but that's what Tkachev et al. claim. The quote as it stands in the comment is ''The flux of a given source decreases as 1/r^2 with the distance r to the observer. This is not taken into account in the method of positional correlations used in Ref. [AUGER] ´´

Anonymous said...

"I blame it on the string theorists, since their goofy ideas have taken over the press while still full of an almost infinite number of errors.'

Eeh.. you want to blame the string physicists for the low quality standards in other fields? Haven't you seen, for example, those phenonemological model builders? Or the recent "Lisi theory of everything" that has been dubbed fabulous by leaders of the quantum gravity scene?

The only connection to string theorists is that the former try to imitate the latter, however to no avail; but that's not at the fault of the string theorists.

Anonymous said...

Unbelievable, I don't understand how this could happen if its true. Not b/c errors like this can't happen (eg a grad student hands a nonbiased data set to a prof who expects a biased set)

but more because you would think it would be immediately obvious upon inspection.

Im off to eyeball the paper now, its got me curious.

Anonymous said...

I think the "unbelievable" responses here are pretty superficial. Since when were the events that produced these cosmic rays supposed to be non-directional? The people are making a valid objection but to me at least it is far from clear how much it would apply in this situation. The devil would be in the statistics.