skip to content

Kavli Institute for Cosmology, Cambridge



Studying the clustering of matter at small separations allows cosmologists to test theories of dark matter.


Despite abundant evidence of the presence of dark matter in the Universe, the direct searches in laboratories have found no viable candidate that would explain the dark matter. It is thus increasingly important to explore astrophysical and cosmological probes of the nature of dark matter. Currently the most stringent constraints on the nature of dark matter candidates comes from studies of the clustering of the intergalactic medium (IGM; see link) in spectra of distant quasars. The transmission feature, called the Lyman-alpha forest, traces the hydrogen density structure to quasars, providing cosmologists with a map of density structure. In the presence of light dark matter species, the clustering of hydrogen atoms at small distances is suppressed. As shown below for a particular dark matter model called fuzzy dark matter (FDM; ultra-light axions), state-of-the-art cosmological simulations show smoother cosmic web in the case of the FDM model, compared to a more standard heavy particle dark matter model (CDM; cold dark matter).


Fig.1: These images depict the distribution of hydrogen gas within the IGM, with bright areas indicating high gas density. On the left is a simulation based on the standard cold dark matter model. On the right is a simulation based on fuzzy dark matter. The amount of structure visible on the left disappears in a fuzzy dark matter model. (Image credit: Vid Iršič)


Utilizing the instrumental capabilities of high resolution spectrometers at VLT and Keck allowed KICC researchers to measure the strength of this suppression in clustering of Lyman-alpha forest and compare it to different theoretical models of dark matter. The measure of clustering is typically presented as shown below, plotted as an amplitude, giving the strength of the clustering, against a wavenumber, an inverse distance measure in velocity units. As such clustering at large distances is represented on the left hand side of the plot (at lower wavenumbers), and clustering at small distances shown on the right hand side of the plot (at higher wavenumbers). KICC researchers use numerically extensive hydro-dynamical simulations to predict the models that are then subsequently tested against the data in a likelihood based approach. An important aspect of such an analysis is to disentangle the amount of suppression at small distances coming from the evolution of the IGM, and that of the dark matter model. To achieve this, researchers look at the evolution of the clustering with time. As the Universe evolves from early times (higher redshifts) the temperature and pressure of the IGM follow a pattern established by the re-ionization history (see link). Whereas the dark matter imprints a characteristic distance scale on the distribution of matter that has a dramatically different evolution with time.


Fig.2: Measured Lyman-alpha flux power spectrum obtained from high resolution data sets (XQ-100 - circles, MIKE - squares and HIRES - triangles). Different colours represent different redshifts, that are indicated on the right handside of the plot, ranging from z= 3 to z= 5.4. The solid line is the best-fit model, and the dashed lines show a model where the mass of the fuzzy dark matter particle is decreased to low mass that is excluded at more than 2-sigma. (Image credit: Adopted from Iršič et al. 2017)

In the case of the fuzzy dark matter model, this characteristic scale is related to the particle's wave-like nature that is one of the concepts of quantum physics called wave-particle duality. Cosmological bounds on such a scale provide important information on the mass ranges that direct search experiments focus on. These constraints will improve with more observations of distant quasars that will be provided through surveys such as WEAVE and DESI, which KICC researchers are heavily involved in. Aside from the Lyman-alpha forest observations, future intensity mapping instruments targeting the early Epoch of Reionization and Cosmic Dawn, e.g. SKA, will be able to access even earlier times, when the clustering at small distances is less affected by the physics of the IGM, thus pushing the frontier in unveiling the nature of dark matter through observations of the large-scale structure.

During the Cosmic Dawn, the evolution of fuzzy dark matter is subject to quantum effects leading to unique interference patterns in the structure of first galaxies. KICC researchers have, in collaboration with colleagues from Princeton University, conducted a series of first-of-the-kind simulations to compare the buildup of first galaxies in a fuzzy dark matter universe to the more familiar cold dark matter scenario. Even though in the late Universe these different dark matter models often predict similar shapes of galaxies, the look of the first galaxies formed at the dawn of star formation would be strikingly different.



Fig.3: Anatomy of a cosmic filament, for cold dark matter (top) and fuzzy dark matter (bottom) cosmologies: (a) the projected dark matter distribution in the simulation domain at redshift z = 5.5; (b) projections of dark matter, gas, and stars in a filament; and (c) slices of the dark matter through a filament. In the cold dark matter model, the dark matter fragments into subhalos on all scales. On the other hand, fuzzy dark matter exhibits interference patterns at the scales of the de Broglie wavelength, which regularize caustic singularities. These differences in small-scale structure will help constrain the elusive nature of dark matter. (Image credit: Adopted from Mocz et al. 2019)


In fuzzy dark matter, cosmic filaments, and not spherical halos -- as in the case of cold dark matter, are the sights of primordial star formation. The filaments are unstable, and where gravity is strong enough, the interference overdensities form stable structures, called solitons, in the centres of filaments. These solitons are unique to fuzzy dark matter model. Their characteristic features are imprinted in the distribution of gas and stars, and are a potential observable feature that could help unravel the nature of dark matter.

Current KICC researchers involved in this area are:

  • Vid Iršič
  • Anastasia Fialkov
  • Martin Haehnelt
  • Prakash Gaikwad