Wednesday, October 30, 2013

Dark matter wars are over: LUX safely excludes all the hints

The dark matter wars are over.

Many TRF readers were watching the sequence of talks in Lead, South Dakota. A gold mine was rebuilt as a top laboratory, thanks to many folks including the philantropist Danny Sanford.

Today, we learned that LUX has instantly become the clear leader of the direct dark matter search experiments. I was immensely impressed by the methods, sensitivity, clever calibration, and so on. Some of the magic of this experiment is the accurate localization of the nuclear events inside the tank.

The results – as presented in the talks and the paper that should officially appear on the arXiv in hours (and was sent to PRL) – are clear. For the CoGeNT or CDMS II Silicon or similar "positive hints" to be approximately right, LUX would have to record about 1,500 events in a region. Instead, they imposed the upper limit of 2.4 events or so. Cool. In other words, about 60% of TRF readers in a poll were right: we were going to listen to a null result.

The exclusion and the agreement with the background expectations is spectacular (no phantom events appear in their data at all) and I have no doubts that all the "Yes signals" are due to some misunderstood background.

This result hasn't excluded all meaningful models of dark matter yet but upgrades of this experiment are able to get to the region where the cross sections start to become unnaturally small.

I personally find it implausible that the CoGeNT/CDMS and other signals are real and LUX doesn't see them because the dark matter is "xenophobic" – avoiding interactions with the xenon nuclei. Why? Simply because xenon contains many vastly different isotopes and it's impossible for all of them to have negligible interactions with the dark matter particle because of some numerology. So some of them would still have to interact comparably to the silicon and one would still have to see hundreds of events although the upper limit is 2.4, to mention the particular example.

LUX has become 20 times more sensitive than the next competitor, using some counting. Using another counting, the upper limits on the cross sections were reduced by a factor of 3 relatively to the most inclusive results of XENON100. LUX has established #clubzepto where other experiments measuring cross section comparable to a zeptobarn may enter in the future. Note that "zepto-" is \(10^{-21}\) and a barn is \(10^{-24}\,{\rm cm}^2\) so the measured cross sections reach \(10^{-45}\,{\rm cm}^2\) or so.

The South Dakota governor gave a charming, inspiring talk. He started by mentioning that he didn't quite understand what the physicians were saying (OK, I improved it by mentioning "physicians") and he enumerated all the knowledge in physics he knows. In particular, light is faster than sound (although he doesn't know what the speed could be) because the sunset came to South Dakota early in the morning. Also, the sum of the angles of a triangle is always three. ;-)

More seriously, we learned about the wonderful scientific history of the site, Ray Davis, the discovery of neutrinos in the Homestake experiment, and so on. South Dakota may be proud. Dark matter, if it exists at all, has to be really weakly coupled WIMPs, heavier WIMPs than expected, axions, some light gravitinos, or something else.

Congratulations to the LUX folks to this amazing job that has removed pretty much all the fog from this industry.


  1. What do you mean "if it exists at all", what is it if it isnt some new particles?

  2. I mean the option of some kind of MOND, although hopefully less contrived than all the types of MOND I have seen so far. ;-)

  3. ... Waterloo?

    I mean, I'd appreciate to see a more wordy TRF article about what these reulsts mean in a bigger picture?

    The popular media will certainly explode with sourball up to trolling news ... :-/

  4. Has this experiment reduced the probably of dark matter existing or merely a few possible kinds of it?

    Since the initial argument for dark matter was anomalously high galactic rotational speed, modifications to GR or quantum gravity likely if MOND is too flakey?

  5. I am very happy about the results. It's not clear to me what sort of Waterloo you mean.

  6. Of course that every experiment that eliminates some models of dark matter reduces the probability of dark matter as a concept - the only question is how much.

    If the prior probability of our belief in dark matter is close to 100 percent, it will stay close to 100 percent. If someone starts with the belief that the probability of any dark matter is a small number "p", then a similar null result reduces "p" by the factor of "f" which is the fraction of the "space of possibilities or parameter spaces" that have been excluded - in that case, the decrease of "p" may be substantial.

    Needless to say, I am still much closer to the first group for which the probability (of dark matter of any type) was and remains close to 100 percent. But I am thinking about other possibilities, too, and especially about other types of dark matter.

    Xenophobic dark matter with CoGeNT- and CDMS-like huge cross sections with other nuclei isn't something I consider realistic because the exclusion is really strong and the numerology couldn't be sufficient to suppress the interactions with all the xenon isotopes.

  7. kashyap vasavadaOct 30, 2013, 7:42:00 PM

    Are neutralino couplings completely free parameters, so you can get cross sections much less than the experimental limits? Also I thought Mond was ruled out by terrestrial experiments. Do you think some modified MOND may be still a possibility? Roughly what kind of modifications?

  8. Dear Kashyap,

    neutralino couplings aren't quite free parameters (after all, these interactions are linked to interactions of the quarks with Z, gamma, Higgs by the supersymmetry: only "which combination" of them is the neutralino is really a free parameter) but the expected magnitude of the couplings for normal-mass neutralinos *is* still beneath the LUX experimental limits so the normal models preferred by theorists haven't been excluded yet.


  9. Maybe in 2014 then...or 2015 when the LHC is back on track...

  10. Exactly as I said. Dark matter and SUSY will never be detected.

  11. Thanks for this nice reassuring reply Lumo !

    Of course do I appreciate these great results and the nice work behind them too :-)

    Seems feeling slightly uneasy for different completely unrelated reasons made my comment a bit too sour ... :-/

  12. I hope the other things will get better. Just to be sure, I repeat the URL for Kashyap here as well because I feel that you think that you should have a bad conscience but it is just not the case. The very first graph

    shows you that most of the theorists-preferred points on the supersymmetric neutralino interactions are still below the lines where LUX is now reaching although it remarkably starts to bite into the interesting region that was a priori preferred by theorists (before some experimenters started to say that the parameters could be very different than theorists believed).

    But it's only a "small minority" of the theorists-preferred region that have been excluded which is therefore no surprise.

    The discovery may still occur next time.

  13. expected... I ignored most of the posts about this subject precisely because of this...

  14. Ok, but still I have a comment: of course none of these experiments could ever exclude completely a theoretical concept... none of them is designed to do that, but then, what experiment could exclude (if the outcome were positive or negative for whatever it would measure) SUSY or dark matter?

  15. After going though this paper, I've lost all faith that we'll ever find dark matter… because it probably doesn't exist. What we assume is dark matter is probably some kind of a space-time anomaly that we haven't accounted for yet.

    And it makes sense: how could space be so full of dark matter (~27% of the universe) and yet we've eliminated all of the lighter energy candidates (light WIMPs) and most of the heavy WIMPs? Could the universe be full of dark matter at even higher energy scales? What process could have created it and why can't we detect it?

    I think theories that explain dark matter effects without resorting to new particles will now be the dominant area of DM research.

  16. NOTHING. The Yukawa alpha-lambda fraud begins! "If only we had a few more decimal places. 26.8 mass-% of the universe disappeared this morning. 45 years of Wesley Crusher theory are empirically falsified. The only viable model of the Tully-Fisher relation is MoND and Milgrom acceleration.

    Milgrom acceleration is a trace chiral anisotropic vacuum background only active toward matter: Noetherian leakage of conservation of angular momentum, nothing violated. One geometric Eötvös experiment, 90 days. If physics cannot understand it, chemistry can.

    No, wait...Tully-Fisherinos! Their collision cross-section is jury-rigged just below future LZ sensitivity. C'mon, Luboš, give some love to chiral spacetime torsion and its visible symmetry-broken sequelae. (Good point about xenon isotopes' nuclear spins.)

  17. This is trolling gibberish, obviously generated by letting a random nonsense text generator act on some physics terms plus some trolling phrases often found in popular media as input.

    Do you have a link to this nonsense text generator, such that I can create my own spam ... :-D ?

  18. Ignorant git and his halcyon ephemerides of compulsory degradative egalitarianism omniscience

    Two geometric Eötvös experiments. 0.113 nm^3 volume/alpha-quartz unit cell. 40 grams net as 8 single crystal test masses compare 6.68×10^22 pairs of opposite shoes (pairs of unit cells).

    [1] arxiv: 1309.7296, 1208.5288, 1204.1859, 0912.5057, 0905.1929, 0706.2031, 1006.1376, 1106.1068
    [2] arxiv:1208.1658, 0801.4566, 0808.0506, 0811.0181, 0907.2562; Phys. Rev. D 68, 104012 (2003)
    [3]; pp. 6-7.
    [4] J. Math. Phys. 40, 4587 (1999);
    [5], by crystal lattice sphere radius., by number of atoms enclosed in sphere, for γ-glycine
    [6] Appl. Cryst. 33, 126 (2000)
    [8]; vs. arxiv:1306.5534, 1306.3983
    [9] arxiv:1109.1963
    [10] (Section 2: eleven pairs of enantiomorphicspace groups); , Section 3ff.
    [11] Class. Quantum Grav. 29, 184002 (2012), arxiv:1207.2442

  19. Very hard to argue against experiment. However this wouldn't be the first false negative in history either. If we have to look toward xenophobia as an answer, then it would seem we would have to look at composition of all the layers of rock above the homestake mine vs the soudan mine. Soudan is an iron mine and homestake is gold mine. I will read the paper more carefully to see if the different interaction cross sections of the rock is also considered, but wonder if these sort of large scale deposit effects are fully controlled for in many experiments

  20. Wow! Slow down man! That kind of activity can't be good for your health. MOND comes is many flavors... you can try mine, BUT due to your already revved up brain activity it might be too damn dangerous :-( So be careful ;-)

  21. There you go -> Seriously, dark matter and dark energy are *not* needed in order to explain observations. Read the linked paper carefully and you'll realize why.

  22. theoretical concepts for physics, called models, have to be validated by experimental evidence , otherwise they are just mathematics, not physics, or video games if they are phenomenological.

  23. You guys are worse than any graduate of a hard-core Madrasa school and do as much damage. Your unquestioned 300 year old faith in the in the power of mass to either warp space or attract other mass has promulgated not only the idea of the undetectable dark matter but also the concept of dark energy which has has as bad or worse theoretical difficulties.

    Kepler believed that it was the sun's light (not its mass) that did the gravitational attraction of the planets. How can light be attractive? Well, now besides me there is someone else that has performed a good many experiments that show heat (and its opposite coldness) effects the weight of a test mass. His experiments are at:
    and mine are at:

  24. Stephen Paul KingOct 31, 2013, 7:46:00 PM

    Has anyone considered that dark matter may be a field or gas of black hole nugget-like "texture" of space-time left over from the breaking of symmetry at the Big Bang and not a particle?

  25. Please feel free to make use of the possibility to stop trolling about GR ... ;-)

  26. Stephen Paul KingOct 31, 2013, 8:01:00 PM

    Seriously! There is an epoch predicted by the Standard Model where all particles where of just one type, where the running constants converge... I recall reading something in Principles of Physical Cosmology Paperback by Phillip James Edwin Peebles on textures...

  27. At what point will redshift be called into question? It seems that some drastically new physics is going to be needed. LIGO has not found gravity waves, LHC has almost completely ruled out SUSY. And now this.
    If further experiments can't find dark matter at what point will people start to question some of the basic tenents like how redshift is calculated and what it actually represents.

  28. Yes, the forces are assumed to have been unified in the very early universe, but I fail to say what exactly you mean by the texture of space time, and the black hole nuggets seemed too funny ;-)

  29. Stephen Paul KingOct 31, 2013, 10:12:00 PM

  30. Stephen Paul KingOct 31, 2013, 10:14:00 PM

  31. LHC hasn't ruled out SUSY at all.

    People will start to question the way how redshift should be computed when a glimpse of a rational argument implying that there's something wrong in the current right way; or a sketch of the new way will be found.

    And it's extremely unlikely that such an event will ever occur (you surely didn't help this hypothetical event arrive earlier) because it would almost be like to find a contradiction in mathematics.

  32. cern or the dark matter experiments cannot rule out anything... they are not constructed to rule out but to confirm. It is a logical bias of experimentalists nowadays... They think in terms of "finding particles" and verifying theories. They do not think at theories, they think only at designing things that could *find* one of the predictions of some theories. The last experiment I remember that was pretty smart was the strong CP violation proof, that one got a well deserved nobel prize... Whatever CERN does is to build larger things and use less intelligence... I really think is is a good idea to have the nobel prize given to higgs and not to some anonimous cern worker...

  33. the major problem is that experimentalists do not have the minimum of theoretical knowledge required in order to develop meaningful experiments, and that is a shame, as it is slowing down progress... "we" need experimentalists with a good theoretical background ...

  34. Huh? LHC ruled out SUSY? What nonsense? -1.

    Read this article, for example, where experimental successes of Supersymmetry are listed.

  35. Agree ...

    Are those not called phenomenologists ... ;-) ?

  36. Husserl was a phenomenologist... we need physicists, not various subcathegories that have no clue about the work of others and are insulated away from anything except their latest paper... Feynman was a good physicist... nowadays there are nonsense categories like "phenomenologists" or "theoreticians" or "experimentalists and all sort of other -ists used as an excuse for ignoring valuable information!

  37. I think dark matter will never be detected just as Hawking's radiation even if we know that they are true.

  38. LHC is the church of physics.

  39. Then please let me clarify. From numerous articles the LHC was supposed to find support for SUSY at the energies the tests were run at. So far no new SUSY particles have been found at these energy levels. That in part rules out the SUSY theroies that predicted lower energy SUSY particles.
    LIGO was supposed to find gravity waves. So far it has not. Advanced LIGO will push this further.
    And now we have LUX and dark matter. While it is true that experiments can't rule out a theory at some point with no positive findings there comes a time that the theory is in serious trouble and should be revised or even discarded.
    As for redshift I have read many articles that showed numerous anomalies that call into question redshift. And redishit is one of the reasons dark energy has been "discovered" to explain it.
    My point is simple. How many experiments will be necessary to force change when these experiments are not finding support for the theories that are currently in vogue.

  40. I thought they were called extinct ..:-)

  41. I assume that you want to say by this, that in the same way as the LHC is a testimony of what modern civilization is able to achieve, impressive famous cathedrals make visible what earlier civilizations have been able to do...?

  42. A texture is a type of topological defect, like magnetic monopoles and cosmic strings. I believe, but am not certain, that textures have been ruled out by CMB observations.

  43. A greatly interesting development.

    Question: Im gathering, based on the lack of mention by the physics sources I read in discussing this result, that it doesn't yet overlap much with what AMS might be seeing. Will LUX ever meaningfully cover the region that AMS is looking at?

  44. SUSY has been expected at the LHC based on the criteria that unknown dimensionless parameters should not be too far from order 1. The Standard Model alone, if it is assumed correct up to the Planck scale, would require fine-tuning of its parameter setting the Higgs mass at the level of 10^{-32} or so. The LHC, broadly and roughly speaking has probed much of the parameter space where SUSY is low enough to totally solve this problem, such that the fine-tuning of the analogous parameter is reduced to about 10^{-2}. So if you interpret this "naturalness" that strictly, you could indeed be growing confident that your expectations were wrong, but thats not the same thing as falsifying SUSY, and as these basic numerical consideration shows, it hardly even reduces this one motivation of SUSY (among many) since the Standard Model is still worse by a factor of 10^30.

    If you're one of the people who never liked or understood SUSY, or someone who was inclined to listen to such a person, then you might claim that such a result is a sign that we're all on the wrong track. Thats their (and your) right, and certainly many more mundane or dubious ideas would justifiably loose interest if they were demonstrated to be cornered to the tuning level of 10^{-2} or worse, especially if there were some standard alternative that had no such tuning at all. But the fact is that SUSY is not some random idea, it is a uniquely special and powerful idea in QFT and theoretical physics, one that solves lots of problems and answer lots of questions without requiring any new concepts at all. Just as importantly there is no "standard" scenario where this tuning problem is absent, the Standard Model is exponentially worse off in that regard. So quite simply it is not at all rational to loose interest in supersymmetry or expect it to go away based on the LHC exclusions with any sane perspective of the situation, conceptually or numerically.

    So the answer is simple: The theories that are "in vogue" will change when something about our current understanding – theoretical and experimental – changes dramatically enough to warrant it. Despite numerous bloggers and authors trying to shout the opposite, there just isn't the evidence to justify your "sky is falling" assessment.

  45. Well, wikipedia considers them "not notable enough", so they're probably not extinct yet, because dodos and lonesome georges are WP:notable : )