No, there are no “hints” of dark matter in new experimental results

Yesterday was a frustrating day for me, and I imagine for many other scientists. For several months now, certain people had been talking up results from the Alpha Magnetic Spectrometer (AMS-02) particle detector mounted on the International Space Station (ISS), without revealing exactly what they were. However, they hinted strongly that AMS-02 was detecting signatures of dark matter annihilation, which—if true—would be a wonderful discovery.

Which brings us up to yesterday, with big press build-up by NASA and CERN, and no availability of the research paper or even where the paper would be published until right before the seminar announcing the results. That’s the kind of secrecy involved in big announcements like the Higgs boson or the Planck cosmic microwave background data. However, when the AMS-02 results were revealed, they seemed a bit…anticlimactic. You wouldn’t necessarily gather that from the press releases (or subsequent press coverage, which generally takes its tone from press releases), but to many of us watching, it seemed that this announcement was not quite worthy of the major build-up and suggestive hints.

I wrote about the results and some of their implications at Ars Technica, but here’s the brief summary. AMS-02 is a multipurpose particle detector, so it’s able to measure the relative amounts of different kinds of particles. Specifically, one set of dark matter (DM) models predicts excess positron emission due to annihilation of two DM particles. (Positrons are the antimatter partner of electrons. They have identical mass and opposite electric charge to electrons, but they are far less common.) Earlier data from the PAMELA and Fermi telescopes hinted at elevated numbers of positrons, but things were pretty unclear—certainly nothing you could point to and say, “This is from dark matter!”

This plot shows the excess of positrons as a function of their energy. The red dots are from AMS-02, while the other colors represent data from other detectors. [Credit: AMS collaboration]
This plot shows the excess of positrons as a function of their energy. The red dots are from AMS-02, while the other colors represent data from other detectors. [Credit: AMS collaboration]
The AMS-02 results announced yesterday exhibited smaller errors and more detail than the previous observations. They confirmed an elevated flux of positrons, whose origin is uncertain. However, the expected signature from DM annihilation is not present in the AMS-02 data. To put it another way: something is making excess positrons, but we can’t pinpoint exactly what, and AMS-02 didn’t—and may not ever be able to—tell us the answer. It might be dark matter, but if you weren’t looking at the data hoping to see DM annihilation, there’s absolutely no reason to see it there.

Don’t get me wrong: the AMS-02 results are very good looking, and provide a lot of details not present in earlier data. The experiment itself appears (to this non-particle physicist) well-designed. I applaud the goal of looking for DM annihilation signatures, and hope someone finds them. However, yesterday’s AMS-02 announcement doesn’t provide “hints” of dark matter: it provides refined measurements of the positron excess, and that’s it. Interesting, potentially exciting, possibly associated with DM annihilation—but nothing you could call a hint, and nothing to justify the hype.

Annihilate! Annihilate!

Hunting for dark matter particles is an inherently complicated thing; see my piece for BBC Future and an earlier blog post for some of the challenges. However, many physicists suspect DM belongs to a particle type predicted by supersymmetry (abbreviated as SUSY, pronounced as SOO-see) called neutralinos. SUSY is too complicated to try summarizing in one blog post, but its details aren’t really important right now anyway. Suffice to say that neutralinos are widely considered to be a good DM candidate, and they have one nice feature: they are their own antiparticle. If a neutralino meets a neutralino coming through the rye, they annihilate, producing (among other things) a positron.

The energy output from annihilation depends on the mass of the DM particles, thanks to Einstein’s E = mc2. Any positron emitted from annihilation will then have a maximum energy, connected to the DM particle mass. The positron spectrum, such as the one in the plot above, will have a steep drop-off at that maximum energy. The AMS-02 data lacks such a drop-off, and the results are too good for one to be hiding. Of course, it could be at a higher energy than yesterday’s data release shows; according to the press conference I listened to yesterday, the AMS-02 researchers are working on that regime next.

However, let’s put this in perspective. If dark matter is neutralinos or some other self-annihilating particle, and the product includes positrons, then the excess positron flux might be due to DM annihilation. Without a drop-off, though, we can’t say anything much. (We also have good reason to doubt the simplest versions of SUSY, but that’s another story.) In other words, some of the excess positrons could be from DM annihilation, those particles may or may not be neutralinos…or if DM is some other type of particle, we’re barking up the wrong tree.
Why the hype?

Credit: Zach Weiner. Thanks to Gabrielle Rabinowitz for reminding me about this comic.

My colleague Ethan Siegel guessed a while back that the AMS-02 results were being oversold, and he was right. After the press events yesterday, I suspect two things are involved: the personal ego of one of the lead researchers, and the need for the various entities running the International Space Station to show that it’s scientifically useful. I’m sympathetic to the latter point: this after all is the era of “austerity measures” and budget sequesters, whose purpose is to push the blame for financial problems onto the poor and onto programs that take up relatively little of a country’s expenditures. (NASA’s entire budget is a mere drop in the bucket of the US budget, but you rarely hear that in discussions of belt-tightening.) However, overhyping results doesn’t help their case, and the outcome is regrettable. The ego of the researcher, which led him to oversell the results starting several months ago, may have forced NASA’s hand, however.

Either way (as Ethan wrote) this is not how science operates. Let’s say what these results really are: an interesting refinement of earlier observations, showing an excess of positrons we need to explain. We should make ordinary science interesting and exciting, but it benefits none of us to make such results sound more revolutionary than they really are.

9 responses to “No, there are no “hints” of dark matter in new experimental results”

  1. Voices for scientific accuracy like yours need to be heard by more. Well done, Matt!

    1. Thanks, Ethan! Same to you.

  2. […] I published my rant over at Galileo’s Pendulum, explaining exactly why I’m grumpish about the way these results were announced and characterized in much of the […]

  3. And today is a frustrating day for me. It is clear from the paper that their source model is robust, even with other’s data as regards fluxes (flux ratios), and “hints” at dark matter through exhibiting the required cutoff. (And in the most interesting regime of TeV WIMPs to boot!)

    But they are way from having the confidence required of a significant observation. No data from beyond the projected peak, even.

    So you can play games and say that the raw data doesn’t show the cutoff. (I’m not sure you can claim it doesn’t “hints” at a cutoff though, seeing how it tapers off nicely.)

    But what you shouldn’t do is play games. I’m surprised, actually frustrated, to read that this is considered “accurate”. There is a testable difference between showing a putative cutoff (fitting of a low confidence model – here even robust) and having an observation (high confidence model).

    The AMS claim will be an extraordinary claim and need extraordinary data. They don’t have even ordinary data for the claim as of the moment (2 sigma – when 2 sigma “hints” come and go). Do we need to say more?

    1. I’m not quite sure what your point is on this, but to be clear: the AMS team was not claiming dark matter discovery, just that the data suggested there could be DM annihilation there. That’s the aspect they played up.

      As for the shape of the positron flux, the leveling-off at higher energies is expected, and there should be a decline in flux at even higher energies—no matter what the source is. After all, positrons can’t have just any energy: something has to accelerate them to these high velocities, and only so many candidates are available. If you measured positron flux at very high energies, you wouldn’t find any at all, whether DM annihilation is responsible for them or not. In other words, the shape of the power spectrum doesn’t tell us anything about DM either way.

  4. Torbjörn Larsson, OM Avatar
    Torbjörn Larsson, OM

    To be precise, the AMS source claim, which is not extraordinary, is not elevated to the extraordinary claim of observing DM annihilation by the AMS team as of yet (at least in the paper, I missed the press conferences).

    So there is also the interesting middle ground of having a diffuse isotropic source if and when they get that far. That is likely when the DM hype will start to take off, as opposed to this whatchamacallit – “anti-DM” (?) – hype.

  5. I believe the current craze is dark matter. It is only a matter of time before some fortune teller finds hints of dark matter in her tea leaves.

  6. […] particles have been announced over the years, only to be squashed by later examination. The recent Alpha Magnetic Spectrometer (AMS-02) results were heavily hyped, and while they might have dark matter hiding in them, the data are far too ambiguous to say […]

  7. […] was u.a. seither auch Papers hier, hier und hier, eine DLR PM, Science@NASA und Artikel hier, hier, hier, hier, hier, hier und hier in Worte zu fassen […]

%d bloggers like this: