There was a time I (Pavel Kroupa) was quite happy with the dark-matter cosmological model. Dark-matter cosmology made a lot of sense, since we were allowed to keep our simple Newtonian "equations of motions" which meant that modelling star clusters, galaxies and the whole universe was simple (but still challenging) computationally. The only price we had to pay was to accept the existence of exotic new "dark" particles which were not part of the otherwise overwhelmingly successful Standard Modell of Particle Physics. Since the neutrino had been predicted in 1933 to exist based on a missing mass problem in beta decay and was then discovered in the laboratory in 1956, the mental pathway had already been laid out towards accepting the existence of dark particles. The dark-matter cosmological model was enhanced over the past decade by the addition of dark energy, that is, the "Lambda Cold Dark Matter Concordance Cosmological Model" (the LCDM CCM) was born. It came to be celebrated as an exquisite physical model of the universe, and many of us astrophysicists knew we live in the era of precision cosmology, given that the CMB, SNIa and large-scale structure observations seemed to provide an excellent fit.
The issue was now merely to work out the details of galaxy formation and evolution, and to confirm the exitence of the dark matter particle through direct experimental detection, and to develop a theory for particle physics which would supercede the Standard Model of Particle Physics and naturally contain the dark matter particles.
Thus the LCDM CCM is currently a widely accepted or should I say "believed" description of the birth and evolution of the universe and of its contents. This comes as no surprise: The many parameters that define this model (see e.g. Lambda-CDM model) have been measured extremely precisely and at the same time the model excellently accounts for the large-scale structure of the universe, as is evident for example in the distribution of galaxies and galaxy-clusters and the microwave background temperature variations. So people were very impressed and started talking about the advent of precision cosmology.
At the beginning I too did not bother with the fundamental issues raised by some (see e.g. Prof. Dr. Tom Shank's paper). My own research was very much confined to the early version of the CCM (mid-1990's) when I began performing numerical experiments on the satellite galaxies of the Milky Way applying, as everyone else, Newtonian dynamics with dark matter. My computational work on the evolution of satellite galaxies as they orbit about the Milky Way within its large massive dark matter halo quite quickly demonstrated that there are very natural mathematical solutions for the observed satellites but without dark matter. I was stunned, because my results were based on observed young satellite galaxies that have no dark matter (tidal dwarf galaxies - TDGBonn) and they evolved into objects that resembled the observed ancient satellite galaxies of the Milky Way and appeared to be full of dark matter. My calculations showed, however, that this was fake dark matter - the stars that make up the satellites move around within the satellite and about the Milky Way, and the observer from Earth interprets these complex motions to be due to an unseen dark matter component in the satellite.
Thus, with time it became increasingly apparent that the CCM accounts poorly for the properties of the satellite galaxies and their distribution around the Milky Way, that is for our cosmological neighbourship. One major problem turned out to be the distribution of satellite galaxies about the Milky Way, in a disk-like structure, which was not in good agreement with the expected distribution within the CCM (Kroupa et al. 2005). The idea that the satellites are a group of dwarf galaxies that fell into the Milky Way dark matter halo (The "Custer's Last Stand" of the CCM) was ruled out as a viable solution to this Disk-of-Satellites problem because dwarf galaxy groups do not have the required properties (Metz et al. 2009).
If the Milky Way satellites can be derived from dark-matter free objects that we definitely know to exist through direct observation (tidal dwarf galaxies), then how can they at the same time be the dark matter sub-structures predicted by the CCM? My work proved that the there were more than one solutions to the satelite galaxies, and that the CCM predictions were therefore not unique. Warm dark matter (WDM) models fared no better. (Warm Dark Matter particles like gravitinos and sterile neutrinos are, to keep it simple for the moment, moving quicker than Cold Dark Matter particles like Axions and WIMPs.
Considering other major galaxies I began to realise that actually I do not know any single galaxy whatsoever which looks like an object that may be described successfully within the framework of the dark-matter CCM (for example in terms of a galaxy's distribution of dark matter within it, or in terms of the thinness and extent of the visible matter in galaxies, or even in terms of the star-formation behaviour of galaxies).
Perhaps the tide began turning significantly in 1999. Then I heard an excellent talk at Harvard University by Stacy McGaugh on his research on rotationally supported galaxies (see Prof. Dr. Stacy McGaugh for much information on this issue). Stacy explained in a most convincing manner that an alternative description via modified gravity (or "extended gravitational theory") actually leads to a far superior understanding of galactic properties than the CCM. See McGaugh's MOND pages and frequently asked questions for an introduction to Modified Newtonian Dynamics or Prof. Dr. Mordehai Milgroms popular article in Scientific American (2002) article or its version for german readers in Spektrum der Wissenschaft (only html, http://www.spektrum.de/artikel/829190).
Basically "Modified Newtonian Dynamics" or MOND means that Newtonian gravity has to be modified at very small accelerations. If accelerations are small, the gravitational force is a bit stronger than would otherwise be supposed. This may either be due to a change in gravity, or due to a change of inertial mass of the particle (in which case the equivalence between gravitating mass and inertial mass would be broken). Both possibilities offer extremely exciting avenues of research within mathematical physics.
So why should this idea replace the concept of dark matter? Both can possibly account for the same fundamental problem: Apparently many stars rotate too quickly around their galaxies. They rotate so quickly that they would fly away if their galaxies had only the mass we can observe. This observation is quite universal and also holds for galaxies orbiting in clusters of galaxies. The most obvious solution to the problem: There must be more mass than we can observe. Having ruled out other candidates for the missing mass, in the 1980s the notion of "dark matter" became generally recognized: There must be a new kind of matter which should be unobservable via direct measurements (that is: "dark"), but interacting via gravity force.
But until today dark matter has not been observed. Worldwide there are several experiments underway to measure its signals but results have been negative all along. Favoured dark matter particle parameters have by now been excluded. And the consistent failure to find any trace of the dark matter particle is increasingly suggesting that dark matter may not exist at all.
Indeed, in the last few years it has become rather clear, based purely on astronomical evidence, that the CCM (with cold or warm dark matter with at most very weak coupling to baryonic, i.e. "normal" matter) is ruled out as a viable description of the universe. Our research paper published in 2010, Local Group tests of dark-matter concordance cosmology: Towards a new paradigm for structure formation, leads to this conclusion (the Local Group is a typical environment of galaxies and must therefore be contained in the CCM).
While this is a very strong conclusion, it rests upon a two-decade-long study towards understanding the dynamical behaviour of the hypothetical dark matter particles and the physical processes at work in normal matter. This study involved working on how stars form and how their birth structures (star clusters) dissolve, and how purely gravitating matter (dark matter) collects in self-gravitating structures with stars and gas. This study compiles the most recent high-resolution models calculated on supercomputers worldwide by various teams with the specific aim to solve the small-scale problems and to explain the properties of satellite galaxies within the logical framework of the CCM. Basically it can now be stated with much confidence that the properties of galaxies and the arangement of their satellite galaxies about them cannot be understood in a universe with cold or warm dark matter and Newtonian dynamics plus the physics of normal matter.
In this research paper five problems for the CCM are found, in addition to the well-documented previously known problems that had mostly not been resolved. Each of the five problems poses a challenge for CCM, and together they exclude it with very high confidence indeed. The Disk-of-Satellite issue is but one of the problems.
To be continued.
by Pavel Kroupa (22.07.2010): "A challenge for Dark Matter" in "The Dark Matter Crisis - the rise and fall of a cosmological hypothesis" on SciLog. See the overview of topics in The Dark Matter Crisis.