Thursday 19 November 2009

Fermi says "nothing" sure sure?

I wrote recently about a couple of theory groups who claim to have discovered intriguing signals in the gamma-ray data acquired by the Fermi satellite. The Fermi collaboration hastened to trash both these signals, visibly annoyed by pesky theorists meddling in their affairs. Therefore a status update is in order. Then I'll move to realizing the holy mission of yellow blogs, which is spreading wild rumors.

The first of the theorist's claims concerned the gamma-ray excess from the galactic center, allegedly consistent with a 30 GeV dark matter particle annihilating into b-quark pairs. The relevant data are displayed on this plot released recently by Fermi, which shows the gamma-ray spectrum in the seven-by-seven degrees patch around the galactic center. There indeed seems to be an excess in the 2-4 GeV region. However, given the size of the error bars and of the systematic uncertainties, not to mention how badly we understand the astrophysical processes in the galactic center, one can safely say that there is nothing to be excited about for the moment.

The status of the Fermi haze is far less clear. Here is the story so far. In a recent paper, Doug Finkbeiner and collaborators looked into the Fermi gamma-ray data and found an evidence for a population of very energetic electrons and positrons in the center of our galaxy. These electrons would emit gamma rays when colliding with starlight, in the process known as inverse Compton scattering. They would also emit microwave photons via synchrotron radiation, of which hints are present in the WMAP data. The high-energy electrons could plausibly be a sign of dark matter activity, and fit very well with the PAMELA positron excess, although one cannot exclude that they are produced by conventional astrophysical processes. But Fermi argues that there is no haze in their data. During the Fermi Symposium last week the collaboration was chanting anti-haze songs and tarred-and-feathered anyone humming Hazy shade of winter. Interestingly, it seems that each collaboration member has a slightly different reasons for doubts. Some say the haze is just heavy cosmic-ray elements faking gamma-ray photons. Some say the haze does exist but it can be easily explained by tuned-up galactic models without invoking an energetic population of electrons. Some say the haze is LOOP-1 - a nearby supernova remnant that happens to lie roughly in the direction of the galactic center. But none of the above explanations seems to be on a firm footing, and the jury is definitely out. In the worst case, the matter should be clarified by the Planck satellite (already up in the sky) who is going to make more accurate maps of photon emission at lower frequencies that will lead to a better understanding of astrophysical backgrounds.

And now wild rumors... which, let's make it clear, are likely due to daydreaming over-imagination of data-hungry theorists. The rumors concern Fermi's search for subhalos, which is one of the most promising methods of detecting dark matter in the sky. Subhalos are dwarf galaxies orbiting our Milky Way who are made almost entirely of dark matter. Two dozens of subhalos have been discovered so far (by observing small clumps of stars that they host) but simulations predict several hundreds of these objects. The darkest of the discovered subhalos has a mass-to-light ratio larger than a thousand, indicating large concentration of dark matter. Because of that, one expects dark matter particles to efficiently annihilate and emit gamma rays (typically, via final state radiation or inverse Compton scattering of the annihilation products). Although the resulting gamma-ray flux is expected to be smaller than that from the galactic center, the subhalos with its small visible matter content offer a much cleaner environment to search for a signal.

So, Fermi is searching for spatially extended object away from the galactic plane that steadily emit a lot of gamma rays but are not visible in other frequencies. The results based on 10-months data have been presented in this poster. Apparently, they found no less than four candidates at the 5-sigma level!!! However, according to the poster, these candidates do not fit the spectra of three random dark matter models. For this reason, the conclusion of the search is that no subhalos have been detected, even though it is not clear what astrophysical processes could produce the signal they have found.

Well, I bet an average theorist would need fifteen minutes to write down a dark matter model fitting whatever spectrum Fermi has measured. On the other hand, the collaboration must have better reasons, not revealed to us mortals, to ditch the candidates they have found. On yet another hand, the fact that Fermi is not revealing the positions and the measured spectra of these four candidates makes the matter very very intriguing. So, we need to wait for more data. Or for a snitch :-)


Luboš Motl said...

I think it is a very healthy attitude that the experimenters say "nothing" as long as they don't see statistically significant signals for "something". Experimenters should be conservative in this sense, and they should have some respect for the well-established theories (and even their most straightforward extrapolations to so-far untested regimes!) that should define their expectations and null hypothesis that will be compared with their new observations.

Of course, the experimental "nothing" means something else than the theoretical "strictly zero". Experimenters can never measure a "strict zero", they can only improve limits claiming that an effect is small.

It's also OK if theorists are trying to find new scenarios and possible phenomena: that's phenomenologists' main job, after all. Of course, the falsified ones have to be abandoned. Those that are not yet falsified must still be filtered and the extremely awkward ones should be ignored, too.

If played well, it's a fair game in which theorists and their theories compete against each other, even before convincing new experimental signals are found, and it's not surprising that experimenters and theorists (or phenomenologists) stay on different sides of an isle.

Unknown said...

Having been at the meeting in Washington DC two weeks ago, I think you're not being qwuite fair about the Fermi reaction to the "Fermi Haze (though they do universally seem unenthusiastic about the name!). The majority of Fermi collaborators I spoke to seemed to think it was something interesting and worth understanding.

It's true that a couple of vocal Fermi people have doubted the haze's existence. And it may be also true that I simply talk to a different segment of the collaboration than you do. But I still think it's inaccurate to say the collaboration as a whole is dismissive...

Anonymous said...

"stay on different sides of an isle."

An what the Hell is an "isle"?

chris said...

"an average theorist would need fifteen minutes to write down a dark matter model fitting whatever spectrum Fermi has measured."

sure, but how long does it take to predict an unknown spectrum? :-) much more fun to see the spoiled model builders struggle to make an actual prediction.

Anonymous said...

Isle \Isle\, noun

1. An island. [Poetic]

Imperial rule of all the seagirt isles. --Milton.

2. (Zo["o]l.) A spot within another of a different color, as upon the wings of some insects.

3. [Obs.] See {Aisle}.

Jester said...

Chris, what I meant is that there are zillons of dark matter models predicting various gamma-ray spectra. So dumping the subhalo-like signal because the spectrum does not match three particular models is a silly thing to do. As I said, I suspect they must have a better reason...but they dont say it.