What We Actually Know About Dark Matter
The thing nobody tells you when they start reading about dark matter is that for most of us who work in astrophysics and cosmology, it has become less exciting and more routine. Ten or fifteen years ago, every paper seemed to claim a new detection. Now the field is full of null results and conflicting signals. The mystery hasn't been solved, but the drama around it has faded into something more like a grinding engineering problem. I spent about four years looking at rotation curve data from SDSS galaxies and then moved into direct detection work, so I've seen both sides of this thing—the observational side where you chase small excesses in velocity dispersion, and the experimental side where you sit on underground data waiting for nuclear recoils that never come. Dark matter is the name we give to whatever is making galaxies rotate faster than they should based on visible mass. That's it. That's the observation. The evidence stacks up across multiple independent methods—galaxy rotation curves, gravitational lensing, the cosmic microwave background power spectrum, large scale structure formation—but each method only tells you something indirect about the thing itself. We know it exists because gravity behaves in ways that visible matter alone can't explain. We don't know what it is. The leading candidate is still the WIMP, or weakly interacting massive particle, but that's a category, not a discovery. The cross-section limits from XENONnT and LZ have ruled out the most straightforward supersymmetric models. I remember sitting in a collaboration meeting around 2019 when someone presented a refined limit on the spin-independent WIMP-nucleon cross section and the room was pretty quiet about it. We had pushed down to something like 4 × 10^-48 cm^2 and still nothing. That's the current state of play: increasingly sensitive detectors, increasingly fewer places for simple models to hide.
There are other possibilities too. Axions, sterile neutrinos, primordial black holes, modified gravity. Each has its own search strategy and each has failed so far. The MOND approach—Mod fied New tonian D ynamics—is worth mentioning because it actually predicts rotation curve shapes quite well for individual galaxies without invoking dark matter at all. But it struggles badly on cluster scales and with the CMB. No single alternative explanation currently fits all the data better than the standard cold dark matter model, Lambda-CDM, even though the model has known tensions like the missing satellites problem and the core-cusp problem. I've seen people get very dogmatic about either side of this debate, and honestly it's usually healthier to just accept that we don't have the answer yet.
How We Actually Search For It
Direct detection experiments like LZ, XENONnT, and the upcoming DarQ and SuperCDMS SNOLAB are sitting underground, mostly in old mines, looking for the tiny recoil of a nucleus when a dark matter particle bumps into it. The background noise from natural radioactivity and cosmic rays is the real enemy here, not the dark matter signal itself. You have to go deep, shield everything, and use materials with extremely low levels of uranium and thorium contamination. I worked on a project where we had to certify that every bolt and washer in the detector assembly came from a batch that had been screened for trace radioactive isotopes. Some of the welding wire we initially ordered had a slight uptick in potassium-40 and we had to send it back. Indirect detection looks for the products of dark matter annihilation or decay—gamma rays, neutrinos, positrons. The Galactic Center gamma ray excess from Fermi-LAT was a big deal for a while. A lot of people thought they might have found something. Turns out it could just be unresolved pulsars. I've seen this pattern repeat: a promising signal, a flurry of papers, then a more careful analysis that either explains it away with conventional astrophysics or reduces the significance to something harmless. Collider searches at the LHC look for missing transverse energy—basically events where momentum isn't conserved in the plane perpendicular to the beam, suggesting something invisible carried it away. Again, nothing conclusive. The limits on simplified models with mediators have been steadily tightening, and people who built their careers on specific benchmark scenarios had to pivot. That's a normal part of experimental science, but it's hard to see from the outside.
Get the Full Details
The Practical Problems Nobody Talks About
One issue that comes up constantly in this field and rarely gets discussed outside of it is how much the Milky Way's dark matter distribution affects everything. We assume a standard Navarro-Frenk-White profile or an NFW-like curve, but the actual local density could be significantly different from the canonical 0.3 GeV/cm^3. I ran into this head-on when calibrating a simulation for a direct detection sensitivity study. The local dark matter density value you assume shifts your expected event rate by a factor of two or three. That's not a small error bar. It's a fundamental uncertainty in interpreting any null result or claim. Another practical headache is the nuclear form factor. When you're calculating recoil spectra, you need to know how the dark matter particle scatters off the entire nucleus rather than a single nucleon. Different target materials—xenon, argon, germanium—have different form factors and different sensitivities to different interaction types. I once spent two weeks debugging a Monte Carlo where the form factor implementation was wrong for an isotope mixture. The code produced a recoil spectrum that looked physically reasonable but was off by about eight percent in the low-energy region. That eight percent matters when you're comparing to experimental limits at the edge of sensitivity. And then there's the solar modulation issue for charged particle searches. If dark matter does produce positrons or antiprotons through annihilation, the solar wind modulates the low-energy flux reaching Earth. You have to apply a correction model, and different models give different results below about ten GeV. The AMS-02 positron fraction data is real, but interpreting it in terms of dark matter requires assumptions about the solar cycle and the local interstellar spectrum that I wouldn't call tight.
Where The Field Actually Stands
The biggest shift in recent years has been away from the simplest WIMP paradigm. The "WIMP miracle"—the observation that a thermally produced particle with a weak-scale cross section gives roughly the right relic density—used to feel like a smoking gun. Now it feels like a lucky coincidence that guided us in the right direction but didn't lead to a discovery. The parameter space that was most aggressively explored is essentially closed for the basic model. That doesn't mean dark matter isn't a particle. It means the first easy target has been eliminated. We're now in a phase where the searches are branching out. Light dark matter below a few GeV is a real frontier, and experiments like DAMIC-M and SuperCDMS are getting there. Axion searches using haloscopes like ADMX have already excluded some interesting regions of parameter space but haven't found anything yet. There's also growing interest in self-interacting dark matter as a way to address the small-scale structure problems, and fuzzy dark matter with ultralight masses around 10^-22 eV, which would behave more like a wave than a particle on galactic scales. The CMB constraints from Planck are now precise enough that they essentially fix the dark matter density to within about one percent. That's remarkable precision for something we can't directly detect. The number is _c h^2 0.12. But pinning down the density isn't the same as identifying the particle. The gap between those two achievements is where the field sits right now.
What To Watch
The next generation of experiments will probably change the landscape more than the current ones have. The Vera Rubin Observatory will map the Milky Way's dark matter distribution through stellar kinematics at an unprecedented scale. The Roman Space Telescope will add gravitational lensing data that should help break degeneracies in how we model dark matter halos. On the detection side, any experiment claiming the first real signal needs to clear a very high bar—multiple channels, multiple experiments, consistent with astrophysical constraints. A single excess in one detector is interesting but not enough. I don't know when we'll figure this out. I've been around long enough to see several cycles of excitement followed by disappointment, and I'm not convinced we're closer now than we were twenty years ago in terms of actually identifying what dark matter is. But the tools are better, the constraints are tighter, and the community is less willing to declare victory over weak evidence. That's progress, even if it doesn't feel like it.
