The News From LUX
LUX Rules Out WIMP Dark Matter In The Narrow Sense
In the narrow sense, WIMP dark matter is matter that lacks color charge and electromagnetic charge, but does interact via the weak force, and is not a neutrino, generally with masses of 1 GeV to 1000 GeV.
LUX is the world's most powerful direct dark matter detection experiment and has ruled out the existence of dark matter over a very wide range of plausible dark matter masses to very tiny cross-sections of interaction with ordinary matter. The entire range of exclusions by all other experiments from 4 GeV to about 1000 GeV, except CDMSLite 2015 (which has a bit of an edge in the single digit GeV mass range for a dark matter particle) is simultaneous confirmed by LUX down to more faint cross-sections of interaction. Tweaks to the analysis of the 2013 data announced in December of 2015 imply that:
[W]e have improved the WIMP sensitivity of the 2013 LUX search data, excluding new parameter space. The lowered analysis thresholds and signal model energy cut-off, added exposure, and improved resolution of light and charge over the first LUX result yield a 23% reduction in cross-section limit at high WIMP masses. Reach is significantly extended at low mass where the cut-off has most effect on the predicted event rate: the minimum kinematically-accessible mass is reduced from 5.2 to 3.3 GeV/c2.LUX is getting an upgrade in early 2016 to improve its performance, which means that the exclusion range will be getting even better in a few years.
The LUX exclusions, for the most part, are enough to exclude interactions on the same order of magnitude as that of neutrinos (the cross-section of interaction of a neutrino with a nucleon, which a WIMP with weak force charge comparable to all other weak force interacting particles known to exist would naively be expected to share, is on the order of 4*10^-39 to 8*10^-39 per cm^2/GeV), which means that any interactions of dark matter with ordinary matter via the weak force would have to involve dark matter particles with a weak force charge that is a tiny fraction of that of all other weakly interacting particles (something with no precedent and no good theoretical motivation).
The LUX exclusion will soon be even stronger requiring even tinier fractions of the weak force charge areas where there is effectively already an exclusion and expanding the mass range over which there is an exclusion of particles with ordinary weak force charges.
Different approaches have to be used to attempt to directly detect dark matter particles significantly lower than 4 GeV (e.g. the keV mass range favored for "warm dark matter" or the MeV range favored for "dark photons" in self-interacting dark matter models) because the neutrino background gets to strong to make out a signal. CDMSLite 2015, for example, which uses a methodology similar to the LHC still gets down only to about 1 GeV (with a considerably weaker cross-section of interaction excluded).
The LHC and previous particle accelerator experiments rule out most lighter WIMPS.
The CMS experiment has provided exclusions more strict than LUX as low masses.
Particle accelerator experiments, likewise, strongly disfavor the existence of fundamental particles that have even very slight interaction with any form of Standard Model particles with masses in the 1 eV to hundreds of GeV mass range, overlapping with the direct dark matter detection experiment exclusion range and providing the most strict direct detection limitations at the low end of the mass range.
Astronomy Data Largely Rules Out Simple Non-Self-Interacting Cold Dark Matter
Astronomy data, in general, disfavors cold dark matter (on the order of 10 GeV or more), because it would give rise to far more small scale structure in the universe. This is because, assuming "thermal" dark matter (i.e. dark matter created in the very early universe with a mean lifetime on the order of the age of the universe), the mean velocity of dark matter particles is a function of dark matter particle mass. Mean dark matter particle velocity, in turn, influences the scale at which ordinary matter would be forced by dark matter to have a highly structured distribution at small distance scales (e.g. groups of galaxies or smaller).
For example, colder dark matter would produce more subhalos in large galaxies and more satellite galaxies around larger galaxies. This limitation is even stronger with dark matter predominantly made of particles in the 1000 GeV or more mass range (roughly the mass of four uranium atoms or more).
This robust exclusion applies even if dark matter interacts only via gravity, but could be avoided if dark matter is self-interacting, or is not a thermal relic (i.e. rather than having a tens of billions of year long mean lifetime or more, it is created and destroyed at rates that are basically equal).
It is unlikely that dark matter that interacts via the weak force with ordinary matter exists with particles of more than 1000 GeV although experimental designs can't rule it out for dark matter that has significant self-interactions (which prevent it from acting like ordinary "cold dark matter" particle theories).
It had been hoped that including ordinary matter-dark matter interactions in cold dark matter simulations would solve this problem, but, for the most part, efforts to do this have been insufficient to solve the problem, or to solve other related cold dark matter problems like the cusp-core problem, which notes that inferred dark matter halo shapes differ materially from the shapes inferred to exist from the dynamics observed in galaxies.
This exclusion shouldn't be unduly exaggerated, however. Dark matter does a very good job of explaining observations at a cosmology scale and at explaining phenomena at scales larger than galaxies, while not contradicting solar system scale observations. And, while the simple singlet fermion cold dark matter particle with no non-gravitational interaction is not right on the mark for galaxy scale and smaller structures, it is a pretty decent first order approximation of it. So, the notion that a dark matter self-interaction of some type could fix this real problem isn't far fetched.
Higgs Portal Dark Matter
A particle that interacts via the Higgs force with a Higgs boson in addition to gravity is sometimes called Higgs portal dark matter, because Higgs boson decays in particle accelerators would reveal it providing the only Standard Model connection to the dark sector.
Experiments at the LHC have not yet pinned down the properties of the Higgs boson precisely enough to confirm that it could not possibly have interactions with "dark" particles that have no interactions via the weak, strong, or electromagnetic forces, so they can't rule out "Higgs portal dark matter." But, LHC data in run-2 will greatly narrow this window of the dark matter parameter space, and even tighter boundaries will exist by the time that the LHC has finished its work.
The 750 GeV Anomaly Cannot Itself Be Dark Matter
The 750 GeV anomaly at the LHC announced last December, even if it is real, is not itself a good dark matter candidate, because a dark matter candidate needs to have a mean lifetime on the order of tens of billions of years or more, and needs to be much, much lighter (although dark matter self-interaction models can weaken the particle mass constraints). But, a 750 GeV anomaly, if real, might imply a whole new menagerie of particles which could include a dark matter candidate.
Sterile Neutrinos In The Narrow Sense Are Increasingly Disfavored
Neutrino oscillation data, as it becomes more precise, also increasingly disfavors a "sterile neutrino" in senso stricto that oscillates with other neutrinos despite not having strong, electromagnetic or weak force interactions, although it wouldn't rule out, for example, a particle called a sterile neutrino in the weak sense, which has no strong, electromagnetic or weak force interactions but does have mass and interact via gravity.
Another indirect, but nearly direct, way of detecting dark matter is to see the signature of dark matter-antidark matter annihilation events, if they exist. The Fermi experiment, for example, is of this type. But, these experiments face the fundamental problem that the background is ill understood. It can exclude annihilating dark matter in areas where no signals are seen up to small annihilation cross-sections, but cannot really confirm that potential annihilation signals have a dark matter source.
Also, the notion that dark matter particles which lack electromagnetic charge would produce highly energetic photons in their annihilation, is itself problematic.
The Simplicity Constraint.
It also bears noting that very simple models of dark matter, with dark matter dominated by one kind of fermion and possibly interacting via one kind of boson, tend to be better fits to the data, almost across the board, than more complex models of the dark sector. This doesn't mean that the dark sector is really that simple if it exists (e.g. no one could guess from astronomy data alone, that there were second and third generation fermions, that there were W or Z bosons, that there were eight different kinds of gluons, or that protons and neutrons were composite particles), but it does mean that the dominant particle content of a dark sector must be very simple.
We can also infer this from the fact that modified gravity models can accurately predict the reality that we observe over many order of magnitude of scale, with one degree of freedom at the galactic scale and only about three degrees of freedom at all scales. The minimum number of degrees of freedom in a modified gravity model is an effective cap on the number of particles that contribute to the dominant particle content of the dark matter sector.
Warm dark matter (ca. keV scale matter) with a dominant dark matter singlet fermion, and self-interacting cold or warm dark matter models with a dominant dark matter singlet fermion and a dominant dark matter boson, which in either case do not interact at all with ordinary matter except via gravity, remain the best fits to the data.
Thus, it seems very likely that direct dark matter detection experiments are doomed to not see any dark matter signals and that LUX will merely find nothing and extend the exclusion range in parameter space for dark matter particles.
The exclusion of so much of the parameter space of plausible dark matter candidates is one of the important reasons that gravity modification theories to explain dark matter are more plausible now than they used to be in the 1980s when dark matter theories were formulated and became dominant.
In contrast, in the 1980s, SUSY provided theoretically well motivated dark matter candidates with the right properties in multiple respects to fit what was then known about dark matter from cosmology models and much more crude predictions about dark matter haloes, when small scale structure problems with the cold dark matter paradigm weren't known, and when LUX hadn't excluded so much of the WIMP (in the narrow sense) parameter space. The remaining dark matter parameter space is no longer a good fit to the SUSY particles that had been hypothesized to be dark matter candidates.
For example, even if warm dark matter with keV mass particles is the right solution, there are really no SUSY particles that can serve in this capacity.
Indeed, ultimately, LUX is more of a blow to SUSY, by ruling out light dark matter candidates that would have a signal at the tested dark matter particle masses if SUSY was correct, than it is to the dark matter particle hypothesis in general.