At 5:00 EST, physicists on the Cryogenic Dark Matter Search Experiment (CDMS) are giving seminars at the nation's leading high energy physics laboratories, Fermilab (Chicago) and SLAC (Stanford), to report on their latest results. This experiment offers the best chance for laboratory detection of dark matter, and they are reporting the results of the most intensive search to date. The physics blogosphere is abuzz with rumors that they may have detected events which could very well be evidence for dark matter. The results are usually given at the end of the talk (around 6:00 EST).
Below the fold, I'll describe the evidence for dark matter and how one can look for it, either through astrophysical observations, in deep underground laboratory experiments, or in high energy collider experiments. I will watch the talks online, and when a result is known, will post it. The preprint giving all of the figures will appear at 8:00 Sunday evening, and I can download the relevant figures from the paper shortly thereafter.
I am a theoretical particle physicist with over 100 publications, many of which discuss dark matter and cosmology. In this diary, I won't give a lot of links to other sources--the Wiki article on the subject is actually pretty accurate.
UPDATE: They found 2 events (alas, a third event was just outside the signal region). The expected background is 0.5-0.8 events. So there is roughly a 20% chance that they have detected nothing, and just got a couple of background events. But there is an 80% chance that they have seen dark matter.
What is dark matter
Dark matter, simply, is matter that doesn't emit any form of electromagnetic radiation. Stars emit visible light, planets/people emit infrared radiation, hot gas emits ultraviolet radiation. Virtually everything we observe in our universe is observed by electromagnetic (EM) radiation. By definition, dark matter does not emit EM radiation, and thus can't be detected directly in telescopes. And yet physicists and astronomers are convinced that dark matter constitutes 80% of the matter in the Universe.
So how do we know it exists?
Evidence from clusters:The evidence has become more and more convincing over the years. In the 1930's, Fritz Zwicky, from Caltech, was looking at the motion of galaxies in galactic clusters. Using Newton's Law of Gravity, studying the motion of objects tells you how much mass is pulling on the objects. Zwicky found that one needed much more mass than was observed visually. He concluded that much of the mass of galactic clusters was invisible.
Evidence from rotation curves:But much stronger evidence came from galactic rotation curves. Vera Rubin (who IMHO should have won the Nobel for this years ago), measured the velocities of objects in orbit around edge-on spiral galaxies. The result was shocking. In freshman physics, one can calculate that the speed of an object in orbit varies as the inverse of the square root of the distance from the center of the orbit (for those who have had freshman physics, one equates centripetal force with Newton's Law...it's one line). This works for all of the planets in the solar system, binary stars, etc. From this, one can determine the mass inside the orbit (it is a very easy way to determine the mass of the Sun). But Rubin found that the velocity curve was flat, and did not fall off with distance.
The dashed line shows the predicted curve, assuming that all of the mass is visible. The solid line shows the measured curves. This is schematic, but actual rotation curves are below.
Clearly, if Newton's Law is correct, there is much more mass than we see. Looking in more detail, it appears that the dark matter forms a spherical halo around a spiral galaxy. Later, analysis of velocities of stars in elliptical galaxies also showed evidence of non-luminous matter.
Evidence from lensing:But there is even more evidence, from gravitational lensing. Suppose a very bright object (like a quasar) is behind a galaxy. Gravity bends light (we see starlike bend around the Sun during eclipses, and it is predicted by Einstein's Theory of Gravity), and thus when the light bends around the galaxy, we can get multiple images (as well as smearing of single images) of the bright object, as shown in the picture.
In the picture, a single object appears doubled. If one includes light rays out of the screen, one can get multiple images. In practice, they are a bit more involved:
.
Nonetheless, using spectroscopy, one can see some remarkable lenses:
. In this picture, spectral analysis shows that the blue streaks are all the same galaxy.
Hundreds, perhaps thousands, of such multiple images have been found. Using these multiple images, astrophysicists have even been able to map (in some detail) the density of dark matter in galactic clusters (and the results match what one expects from velocity measurements).
Evidence from the Bullet:The most direct observational evidence is the Bullet Cluster, shown in the picture.
These are two galaxy clusters that have collided, and are now moving apart. When galaxies collide, the dust and interstellar material in the two galaxies interact, and thus slow down. The dark matter, however, does not interact, and keeps moving. The red in the picture is the hot gas, and the blue is the dark matter (mapped from gravitational lensing). One can see that the dark matter has separated from its parent galaxy.
The amount of dark matter is now known, and it is around 80% of the total matter in the universe. It is remarkable that we don't know what 80% of the matter in the universe is made of...
Maybe the Law of Gravity is Wrong!
This was the leading alternative to dark matter for many years. In fact, there is precedent. In the 19th century, there were two discrepancies in the solar system with the law of gravity. The orbit of Mercury was a little bit off (very little, but measureable), and the orbit of Uranus was also a bit off. The discrepancy in Uranus' orbit was caused by "dark matter", i.e. previously unseen matter--known as the planet Neptune. The discrepancy in Mercury's orbit was caused by a modification of the law of gravity (when Einstein's Law modified Newton's Law). Thus it is quite possible that the galactic rotation curves are caused by a modification of the law of gravity. The most popular alternative was called MOND (Modified Newtonian Dynamics).
However, MOND has great difficulty explaining gravitational lensing and is completely incapable of explaining the Bullet Cluster. Thus it is not taken as seriously now as it was years ago.
OK, so what is the Dark Matter?
In principle, anything that doesn't emit EM radiation is dark matter. If you were to teleport Glenn Beck into deep space, then when his corpse cooled off to the temperature of ambient space (3 degrees Kelvin), he would no longer emit any kind of EM radiation and would be dark matter. It is unlikely that the dark matter is composed of corpses, but could it be small cold planets/rocks? No. It turns out that when cosmologists calculate the production of nuclei in the early universe, everything works out beautifully if the total density of neutrons and protons is roughly what we observe today. If the dark matter were composed of neutrons and protons, these calculations would be way, way off. Furthermore, we know of no mechanism that could hide all of those protons and neutrons in the amount necessary. So the dark matter is not made of atoms.
It could be neutrinos, which are very light elementary particles that do not interact electromagnetically. They do interact with the "weak interaction" which is the interaction responsible for radioactive beta-decay; a typical neutrino might go through the earth without interacting, yet we not only have detected and studied them, but now make beams out of them. But they are too light and thus move so fast that they wouldn't clump into galaxies.
The most likely candidate, according to the overwhelming majority of physicists, is something with the horrible acronym WIMPs, which stands for Weakly Interacting Massive Particles. In many, many theories of particle physics, there exists one or more particles that are massive (roughly 50-200 times heavier than a proton), have zero charge (so they don't interact with EM radiation), stable (so they don't decay) and only have weak interactions. When one calculates their abundance today, they have just the right properties and evolve in the right way to consitute the dark matter.
My own personal favorite dark matter candidate (and I'd guess that 50% or more of particle physicists share this view) is the "LSP" or lightest supersymmetric particle. Supersymmetry is a remarkable symmetry which requires that every known particle (electron, quark, photon...) is accompanied by a "partner" (called the selectron, squark, photino...). It turns out that local supersymmetry theories automatically contain gravity, they emerge from string theory, they solve many of the outstanding problems with the Standard Model of Particle Physics, and there are close to 100,000 papers written on supersymmetry in the past 20 years. The search for supersymmetry is one of the two main targets of the Large Hadron Collider starting up in Geneva (the other is the Higgs---but that's another story). It also turns out that the lightest of the partners (probably the "photino") is absolutely stable and has precisely the properties one expects of dark matter.
What is the new experiment, CDMS?
Other than astrophysical observations, are there any other ways to detect dark matter? At CERN, the Large Hadron Collider will be able (for almost all possible dark matter candidates) to produce dark matter particles directly, if they are WIMPS. The dark matter particles would appear as missing energy--they would be able to say that a particle with a mass of A and an interaction rate of B was produced. This could be a smoking gun. But that will almost certainly take until 2012 at the earliest.
Another method is direct detection. The earth is moving through the dark matter, and many billions of dark matter particles go through your body every second. But they barely interact. The CDMS experiment is designed to detect them. It is located a half mile underground in an old mine in northern Minnesota--by being that deep it avoids backgrounds from cosmic rays. It consists of hockey puck size wafers of silicon and germanium, cooled to one one-hundredth of a degree above absolute zero. When a dark matter particle interacts, it will shake the crystal, and the vibration will propagate from the interaction point to the surface, where there are sensitive detectors (and exquisite timing, to distinguish from external vibrations). It has been running for several years. The most recent results were presented a little less than two years ago. The result presented then was:
The bottom panel has background events removed, and signal event are between the red lines. You see that there are no signal events, and so they did not detect dark matter. They then published their summary results:
This is a busy figure. The x-axis is the mass of the WIMP, and the y-axis tells how strongly it interacts in their detector. The relevant line is the black line--CDMS claims that there is no dark matter with a mass and interaction strength above the black line. The relevant shaded region is the green region. This is the region predicted by supersymmetry. One can see that close to half of the region is already excluded. With more data, the black line will drop down, and could eventually excluded supersymmetry (although that will take years). Of course, with more data, a discovery could be made. They are now announcing the results based on many more years of data, and we will see, very soon, what their results are.
The following has been stated as the two scenarios:
* Scenario #1
CDMS has detected 2-3 events with the expected background of order 0.5. All eyes will turn to XENON100 - a more sensitive direct detection experiment that is kicking off as we speak - who should provide the definitive answer by the next summer. In the meantime, theorists will produce a zillion of papers fitting their favorite recoil spectrum to the 3 events.
* Scenario #2
All this secrecy was just smoke and mirrors. CDMS has found 0 or 1 events, thus setting the best bounds so far on the dark matter-nucleon cross section. Given the expectations they raised in the physics community, the Thursday speakers will be torn to pieces by an angry mob, and their bones will be thrown to undergrads.
UPDATE: Scenario #1 is correct. They found 2 events (alas, a third event was just outside the signal region). The expected background is 0.5-0.8 events. So there is roughly a 20% chance that they have detected nothing, and just got a couple of background events. But there is an 80% chance that they have seen dark matter.