Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Galaxy formation

Gone with the wind?

Windy weather is forecast where stars are forming. Numerical simulations show that these winds can reshape dwarf galaxies, reconciling their properties with the prevailing theory of galaxy formation.

The smallest things often cause the most trouble. The smallest galaxies are no exception: they have long caused difficulties for modern cosmology. Neither the number nor the appearance of small 'dwarf' galaxies conforms to the predictions of the otherwise highly successful cold dark matter (CDM) theory of galaxy formation. Such predictions, however, are usually derived from simulations that don't include stars. This apparently glaring oversight is justified because visible matter plays a minor part in the much larger drama of galaxy formation. The hidden actor is dark matter, whose effects are felt only through gravity. On page 203 of this issue, however, Governato and colleagues1 demonstrate that visible matter also has a key role in modifying the properties of dwarf galaxies, thereby simultaneously solving two long-standing problems with CDM galaxy formation.

The precise nature of dark matter is unknown. The favoured 'cold' dark matter is a hypothesized particle that became non-relativistic (slow-moving) soon after the Big Bang2. Yet despite this uncertainty, the prevailing model of galaxy formation based on CDM is tremendously successful, predicting a vast range of observational data.

Dwarf galaxies are small accumulations of stars with total masses one-tenth or less that of our Galaxy. Unlike the Milky Way, with its grand spiral arms and prominent central spherical bulge, dwarf galaxies have no central bulge and have a poorly organized spiral structure3. They are also less skilled than their massive counterparts at making stars and retaining gas. Stars and gas in a dwarf galaxy typically make up less than 10% of the total mass, with the remainder composed of dark matter.

A fundamental prediction of CDM galaxy formation is that galaxies have central 'cusps' — the density of matter increases steeply towards a galaxy's centre. These cusps are thought to arise because galaxies form hierarchically, with small structures merging to create larger structures4,5. As merging proceeds, matter with low angular momentum sinks to a galaxy's centre, creating a cuspy profile. Even dwarf galaxies form out of many smaller individual objects, and should thus have steep central density profiles. This is in sharp contrast to observations. Dwarf galaxies are often seen to host large regions of nearly constant density in their centres6,7: their profiles are 'cored'. The standard CDM theory simply cannot explain cored profiles in such galaxies. Many solutions have been proposed to solve this 'core–cusp' dilemma8 — most dramatically, abandoning CDM altogether in favour of an entirely new theory.

But before such a drastic step is taken, the accuracy of the cusp prediction must be verified. Cusps are predicted from numerical simulations that include only dark matter3,4. On the face of it, dwarf galaxies should be well represented by dark-matter-only simulations, because their mass is dominated by dark matter. On the other hand, these galaxies are sensitive to energetic processes such as star formation, which can release energies comparable to that binding the entire dwarf galaxy together. If star formation affects the central regions of dwarf galaxies, it might solve the core–cusp problem. Testing this hypothesis requires simulating the dark and luminous matter in a galaxy simultaneously — a challenge for even the most powerful supercomputers.

The race to perform larger and more sophisticated galaxy formation simulations is fiercely competitive. Governato and his team1 make a clever choice in simulating dwarf galaxies, rather than larger structures in the Universe. Simulations are severely limited by computation time, and thus the smaller size of a dwarf galaxy leaves more computational resources for resolving smaller physical scales. This is key to Governato and colleagues' success. Their simulations follow the wider environment in which the galaxy forms, while resolving the smallest physical scales relevant to the process of star formation.

The authors' models1 reveal that stars and gas have an active role in shaping a dwarf galaxy. In the normal course of star formation, massive stars are produced that have very short lifetimes, ending in spectacular supernova explosions. Strong winds from these explosions remove gas from the region of star formation. If star formation takes place in the centre of a dwarf galaxy, supernova winds preferentially remove the low-angular-momentum gas that has sunk to the galaxy's centre. The dark matter must also react to gas removal in order to maintain dynamical equilibrium — by expanding outwards.

The mark of a satisfying astrophysical solution is that it solves multiple problems with a single physical process. Governato and colleagues claim that supernova winds explain both the shallow, cored density profiles observed in dwarf galaxies and their lack of a central bulge, simultaneously overcoming two major problems in CDM galaxy formation. Still, the team has so far fully simulated only two dwarf galaxies. Simulations over a wider range of masses and environments are required to verify their bold claim that supernova winds solve the core–cusp problem. However, a strong hint that this team is heading in the right direction is visually evident — images of their simulated galaxies are nearly indistinguishable from the real thing (Fig. 1).

Figure 1: Which is the real galaxy?

Governato and colleagues' numerical simulations1 produce galaxies that seem identical to images of real galaxies. (Real galaxy (right) and background image courtesy of the Sloan Digital Sky Survey Collaboration (; simulated galaxy (left) and composite image courtesy of C. Brook, F. Governato and P. Jonsson.)

Governato and colleagues show1 that supernova winds are sufficient to reshape low-mass dwarf galaxies, reconciling the galaxies' properties with the CDM theory of galaxy formation. What has been a long-standing problem may now quite literally be gone with the wind.


  1. 1

    Governato, F. et al. Nature 463, 203–206 (2010).

    ADS  CAS  Article  Google Scholar 

  2. 2

    Davis, M. et al. Astrophys. J. 292, 371–394 (1985).

    ADS  CAS  Article  Google Scholar 

  3. 3

    Blanton, M. R. et al. Astrophys. J. 631, 208–230 (2005).

    ADS  CAS  Article  Google Scholar 

  4. 4

    Navarro, J. et al. Astrophys. J. 462, 563–575 (1996).

    ADS  MathSciNet  CAS  Article  Google Scholar 

  5. 5

    Moore, B., Governato, F., Quinn, T., Stadel, J. & Lake, G. Astrophys. J. 499, L5–L8 (1998).

    ADS  Article  Google Scholar 

  6. 6

    Simon, J. D., Bolatto, A. D., Leroy, A., Blitz, L. & Gates, E. L. Astrophys. J. 621, 757–776 (2005).

    ADS  CAS  Article  Google Scholar 

  7. 7

    Swaters, R. A., Madore, B. F., van den Bosch, F. C. & Balcells, M. Astrophys. J. 583, 732–751 (2003).

    ADS  CAS  Article  Google Scholar 

  8. 8

    de Blok, W. J. G. Adv. Astron. 2010, doi:10.1155/2010/789293 (2010).

Download references

Author information



Rights and permissions

Reprints and Permissions

About this article

Cite this article

Geha, M. Gone with the wind?. Nature 463, 167–168 (2010).

Download citation


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing