Friday, April 5, 2013

Positrons In The Sky. Why?

Astronomers have long known that the Earth and everything in its vicinity are constantly bombarded with "cosmic rays" a rather misleading term that includes highly energetic particles such as electrons and positrons (the anti-matter equivalent of electrons) as well. The Alpha Magnetic Spectrometer (AMS-02) experiment, on the International Space Station, has been carefully measuring this bombardment of electrons and positrons before they hit the atmosphere, leading to complicated by interactions with the gases in it.

Their findings confirms those of several prior experiments, but with far greater precision which showed that the ratio of electrons to much more more positrons in cosmic rays is dependent on the energy of the particle (although positrons are always far more rare than electrons).

Positrons make up 5% of the lowest energy cosmic rays.  Positrons grow relatively less common up to particle energies of a bit less than 10 GeV where they make up about 1% of cosmic rays of that energy, and then grow more common again, reaching a proportion of 11% by the first bin starting at 100 GeV and 15% in each of the two bins covering 206 GeV to 350 GeV (although at a declining rate as energies increase) starting from 10 GeV through at least a bin of events in the 250 GeV to 350 GeV range.

There were 465 electrons and 72 positrons observed in this first round of ASM-02 data in the highest 260 GeV to 350 GeV data bin. The trend is smooth as illustrated on the graph below, and notably neither electrons nor positrons show any favored direction of origin in the galactic sphere at the time they reach the AMS in Earth orbit. We also know that there was at least one 982GeV electron and at least one 636GeV positron observed.

Fig. 1: AMS results (red dots) compared to previous results, of which the most important are PAMELA (blue squares) and FERMI (green triangles).  Within the stated errors, PAMELA, FERMI and AMS are essentially consistent; all show an increasing positron fraction above 10 GeV and as far up as 300 GeV or so.

Nobody is surprised that there are more electrons than positrons.

Lots of processes can rip electrons from atoms and send them hurling into space at high energies where they become cosmic rays. Few process emit high energy positrons, and when they do, they usually generate equal numbers of high energy electrons.

The natural assumption is that we have two data sets that need to be disentangled from each other. One set of phenomena that is producing roughly equal numbers of electrons and positrons at particular energy levels, and another set of phenomena that are producing only electrons. This is complicated a bit by the fact that electrons of all kinds can easily pick off positrons en route to the detector and electromagnetic attraction attracts stray cosmic ray electrons and positrons to each other.

Also, many processes that create positron-electron pairs also create pairs of quarks whose products are not reflected in the data, and so can't be a source of these positrons.

Just about everybody agrees that we're not certain why this positron excess is observed. It could be emissions from pulsars or active galaxies. It could be some process in astronomy that nobody's ever seriously considered. It may or may not be "new physics" beyond the scope of the Standard Model and General Relativity, but at the very least, this is a fairly new and now much better documented unexplained phenomena in astronomy and hence is a big deal.

Keep in mind also, that given what we know so far, one doesn't have to interpret the data as a high and low energy "positron excess" at all.  The money chart that invites the conclusions and analysis merely shows an electron-positron ratio. It is equally possible, given the existing and publicly available data, that we are mostly seeing an excess middling energy electrons in cosmic rays because there is some unknown process that generates electrons but not positrons in cosmic rays of that energy range, that tapers off at higher energies.  However, if the rumors discussed below are true, that can't be the whole story.

Matt Strassler and Resonaances confirm the summary above and point out that while the AMS-02 data is much more precise that it generally confirms facts that we already knew or strongly suspected about electron-positron ratios.

Are results being censored and why should we care?

The AMS did not release data from bins with higher energies, despite the fact that they did have higher energy events.  The project's spokesperson stated that those findings were too scattered to be statistically significant, but Lumos Motl argues that this is a bit of a conspiracy to sustain interest in the experiment which has years of activity ahead of it that must continue to be funded.

There are rumors that the data beyond the 350 GeV show a dramatic decline in the ratio of positrons to electrons, which would be quite pertinent to determining the possible source of the positron proportions that are observed. Motl goes so far in the post linked above as to sketch out some fake data estimates of what he thinks has probably been held back from the data.

Motl cares mostly because an analysis of the data from the earlier PAMELA experiment by SUSY theorists demonstrated that the positron excess that is observed could be caused by the annihilation of a SUSY dark matter candidate called a neutralino with a mass on the order of 300 GeV to 350 GeV, a range which, with the right assumptions, hasn't fully been ruled out by Large Hadron Collider data so far.

The discovery of a SUSY particle that makes up dark matter in the universe would be a huge deal that would instantly falsify the Standard Model, and even if the discovery of a new particle could not be confirmed from this data alone, the AMS-02 positron-electron data would strongly point LHC researchers towards the precise mass and particle properties to be looking for to discover a new SUSY particle if one is out there and producing this effect.  The dramatically narrowed search parameters for the suspected 300-400 GeV SUSY neutralino could greatly improve the power of the LHC to prove or disprove that particular hypothesis in future runs.  And, a 300 GeV-400 GeV SUSY particle ought to be something that the LHC has the capacity to observe without any major modifications sometime in the next few years.  String theory would be saved and the Standard Model would be resigned to a mere low energy effective theory for a supersymmetric higher energy effective theory that manifests below the TeV electroweak scale.

But, if I were you, I wouldn't invest in a stash of party balloons and confetti yet.

What could produce the results allegedly being censored?

For what it is worth, I don't rule out the possibility that positrons in high energy cosmic rays may have origins in some sort of exotic particle or process, although it isn't entirely clear what kind of exotic particle would produce positrons over so wide an energy range in so smooth a manner.

In the case of a fermion or a short lived bosons like a W or Z bosons or a Higgs boson, you would expect a distinct bump at a particular, discrete energy threshold and not a long slope.

Decays into positrons and electrons over such a wide range of energy levels would be more characteristic of a relatively stable boson that can have a range of energies before decaying predominantly to positron-electron pairs.

Do we know of any such particles?

It turns out that we do. They are called photons, and at high energies (in high end of the gamma-ray wavelengths of 100keV or more) they can spontaneously produce high energy positron-electron pairs in an exceedingly well understood way, while being relatively disinclined to create protons (just the leptophillic pattern one would like to see given the data we see at AMS-02).

One would have to figure out the source of photons that decay to the positrons and some of the electrons that we observe in cosmic rays to explain the data, but many processes that produce gammay-rays and surely we don't have a full accounting of all of them.

Ordinary beta decay of stationary heavy atoms typically produces gamma-rays of 10 MeV or less.

Pair production starts to become possible at a bit more than 1 MeV that encounter electric fields.

Obviously higher energy pairs of the kinds seen with hundreds of GeV of mass-energy by AMS-02 take more energy to produce. Pair production is the predominant form of absorption for photons in the GeV energy range or above that encounter electric fields.

But, there is nothing terribly exotic about photons in particle physics. If the positrons observed are a result of pair production by gamma rays, then the interesting question is what is producing such a spatially uniform, but energetically non-uniform distribution of them.

On the theory that extraordinary claims require extraordinary proof, I'd hope that one would try very hard to exclude positrons created by gamma-rays perhaps from new previously unknown sources, before resorting to explanations requiring 300-400 GeV neutralinos that also happen to make up dark matter.

Dark matter generally shouldn't decay to positron-electron pairs

We wouldn't expect dark matter that lacked an electromagnetic charge to couple to positron-electron pairs at all.

For example, Standard Model neutrinos do not do directly, although if energetic enough a neutrino and antineutrino that collide can give rise to a Z boson that can decay to a positron-electron pair, as well as into pairs of particles and antiparticles of every other possible type with equal probabilities, more or less, for all energetically permitted pairs. Likewise, Z bosons do not decay directly into photons which don't couple to the weak force and would violate conservation of spin if produced in pairs by a single vector boson like a decaying Z boson.

Astronomy data disfavors heavy dark matter particles that easily decay to positrons and electrons

But, whatever is generating the high energy positrons observed in cosmic rays, the notion that this is an annihilation of exotic, non-baryonic dark matter particles such as neutralinos is exceedingly implausible.

As I've observed in other posts, the astronomy data increasingly strongly favors dark matter particles with masses in a narrow 1000 eV to 2000 eV mass range (called "warm dark matter") and increasingly strongly disfavors particles with masses in the hundreds of GeV mass range (called "cold dark matter") that would annihilate or decay to particles in the energy range observed by AMS-02 followed by a dramatic drop off in predicted positron ratio at higher energies as an important component of dark matter.

SUSY does not have any plausible dark matter candidates in the keV mass range. Even the low GeV mass range is a challenge for a SUSY theory to assign a particle to without creating observable phenomena that haven't been detected so far in laboratories.

This suggests that SUSY does not have a viable exotic dark matter candidate, which deprives that theory of one of its important justifications for being studied.

Even if some SUSY theory compatible with existing particle collider experiments was correct, given that we know that cold dark matter is too heavy to be consistent with current astronomy observations, none of the SUSY particles could be stable, not even a lightest supersymmetric particle (i.e. LSP), because even an LSP would be too heavy to be dark matter. So any large quantity of stable LSPs would be contrary to astronomical observations.

A hitherto unknown long range force operating between dark matter particles (probably with a U(1) gauge group similar to that of electromagnetism that affects only dark matter or has highly suppressed interactions with other kinds of matter) might mitigate the problems with a cold dark matter model for cold dark matter that is almost light enough to be warm dark matter, but not quite.  But, when you are talking about particles with masses of 300 GeV and up, it is virtually impossible to include those particles in any dark matter model that can fit the astronomy data.

The astronomy data are a much better fit to a 1 keV to 2 keV sterile neutrino that interacts only via gravity and some force that operates only between sterile neutrinos with a massless or very light carrier boson (as suggested, for example, by graviweak unification theories), not a hundreds of GeV WIMP candidate.  So, even if annihilations of dark matter particles and antidark matter particles of the right masses, if this could produce photons at all (perhaps indirectly through a chain of decay reactions), they would tend to produce X-rays, rather than gamma-rays with sufficient energy to produce high energy positrons of the kinds observed at ASM-02.

Is the positron energy range related to nuclear binding energy curves and atomic mass distributions?

I would also note that the cutoff in positron frequency that Motl supposes exists coincides pretty closely with the mass-energy of the most massive atoms in the periodic table and that the slope of the electron-positron ratio curve fairly closely mirrors the inverse of the first derivative of the nuclear binding energy per nucleon v. number of nucleons curve (the curve itself hits its peak at about 50 GeV of particle mass but the peak of the curvature comes much earlier), although probably with a bit of amplification (perhaps squared or cubed).

Putting two and two together, the positron v. electron ratio in cosmic rays could reflect something as benign as the natural distribution of high energy photons released from some subset of nuclear reactions in stars or supernovas. The abundance of positrons at low energies could reflect the relative abundance of hydrogen and other light atoms in the universe (and maybe even the surprisingly low levels of lithium in the universe), while the abundance of positrons at high energies could reflect interactions involving the heaviest atoms.


andrew said...

See here finding the results consistent with gamma-ray emissions from pulsars and a stretch to match to previously considered dark matter anhillation models.

andrew said...

Another paper examining dark matter options states that the dark are inconsistent if this hypothesis is correct and that the DM mass must be very high.

"In light of the first release of high precision measurement on the positron fraction from the AMS-02 experiment, we perform a detailed global analysis on the interpretation of the latest data of PAMELA, Fermi-LAT, and AMS-02 in terms of dark matter (DM) annihilation and decay using the GALPROP code. The allowed regions for the DM mass and its annihilation cross section or decay life-time are obtained for leptonic final states 2e, 2mu, 2tau, 4e, 4mu and 4tau. We show that for DM annihilation into electron and muon final states, the AMS-02 data alone favour a DM mass 300-400 GeV, which is significantly different from that favoured by the Fermi-LAT data on the total flux of electrons and positrons. The allowed regions by the two experiments do not overlap at 99.99999% C.L., which is insensitive to the changes in the normalization and slope of the conventional astrophysical background. Consistent fits are obtained for 2tau and 4tau final states, which favours TeV scale DM with large cross sections above 3times 10^{-23}cm^3*s^-1. For all the considered leptonic channels, the current data favour DM annihilation over DM decay, and show no indication of charge asymmetric DM decay."

andrew said...

Yet a third paper on the same theme.

The first and the third conclude that the data are consistent with pulsar sources, which the second does not consider. All three papers imply that a dark matter annihilation that fits the data much be very heavy (600 GeV+) and must have a very high annihilation cross section (10^-23 or 10^-24 rather than the 10^-26 generally assumed for these kinds of dark matter), and some point out difficulties in fitting the data consistently. The low cross-section is required for relic densities not to disappear by the current era.

Given that very heavy CDM is strongly disfavored experimenally and other problems with the DM hypothesis, the case that the positron excess is pulsar sourced rather than dark matter sourced is high.

andrew said...

Another recent paper confirms that an ordinary astronomy source fully describes the data.

andrew said...

New data has been released from AMS-02 and contrary to Motl's speculations at the Reference Frame, there is no sudden drop off in positron fraction in the 300-350 GeV range as one would expect in a WIMP dark matter hypothesis.