Star and Planet Formation

Our research is focused on understanding how stars and planets form.

Star formation research lies at the heart of almost all of present-day astronomy and astrophysics.

We use star-forming regions to probe the expansion of the Universe and to test our models of cosmology. We then use these same regions to study how galaxies form, grow, and mature.

We also use star formation to understand the origin of planets. Our own solar system probably assembled while the Sun itself was still being formed, and so star and planet formation are closely linked. So ultimately we wish to use our knowledge of star formation to better understand they way the Universe evolved to have galaxies such as the Milky Way, and planets such as the Earth.

The challenges

There are three key challenges in star formation research in that it:

  • spans a huge range of scales, from whole galaxies, containing billions of stars, down to the individual planets around stars like the Sun
  • involves a very broad range of physical processes, including radiation-matter interaction, fluid dynamics, magnetic field interactions, chemistry, turbulent theory, and even solid-state processes
  • is a process which can take millions of years, and so astronomers are only able to catch the briefest glimpse into a very long, dynamic, and complicated process.


Our ultimate aim is to predict the collective properties of stars, in particular:

  • the distribution of stellar masses (or Initial Mass Function)
  • the statistics of binary systems and higher multiples
  • the structure of star clusters
  • the interaction of newly-formed stars with their surroundings
  • the external appearance of star formation regions - and how, if at all, these quantities depend on environment, metallicity and epoch.

Our research projects

To address our research challenges we employ several different methodologies, combining observations from the world’s top observatories with state-of-the-art computer simulations of star formation. We also work closely with the Astronomy Instrumentation Group to develop new cameras for ground and space-based observatories, to help obtain better information on the star formation process.

Prestellar cores and the origins of the Intial Mass Function (IMF)

Low-mass prestellar cores

A typical star has a mass four or five times less than the Sun, its luminosity is about ten times lower than the Sun, and it lives longer than the current age of the Universe. These stars are formed in low-mass prestellar cores, of which many are nearby, so they can be observed in detail.

Our research has shown that by using the iram 30m telescope and HERSCHEL these cores tend to be concentrated in filaments, and that their mass distribution appears to be very similar in shape to the stellar Initial Mass Function. Using theoretical arguments, we have also demonstrated that each low-mass prestellar core is likely to spawn between four and five stars.

Modelling core collapse

To predict what form of stars a core collapse might form, we have to resort to numerical simulations. The observed intensity is a projection on the plane of the sky, and places very limited constraints on the distribution of matter along the line of sight, whilst the Doppler shifts of emission lines constrain only the radial component of the velocity of the emitting or absorbing matter and the tangential components are normally unknown. Consequently, we have to solve an awkward inverse problem to obtain the three-dimensional density- and velocity-fields that are required to set up initial conditions for such simulations.

Our research has recently shown how, with a minimum of credible assumptions one can generate 3D initial conditions (density- and velocity-fields) for simulations of core evolution, which match — in a statistical sense — those observed in a particular star formation region. In this way we have attempted to predict the basic properties of the stars that will have formed in Ophiuchus a million years from now.

Stellar multiplicity

Although single stars like the Sun are not uncommon, most stars are born in binary systems (two stars orbiting their mutual centre of mass) or higher multiples (i.e. triples, quadruples, etc.). Some of these systems are quickly disrupted by external tidal forces, including impulsive interactions with other stars, but many survive for ever. Higher multiples are normally only stable if they are hierarchical; for example, a quadruple might comprise two close binary systems (say, each with separation S ≤ 10 AU), with the two binary systems in orbit about each other in a much wider orbit (say, S ≥ 100 AU). A key goal of star formation is to establish the statistics of binary systems, and to explain how they are produced. Indeed, many perople see reproducing binary statistics as being the acid test of star formaton theory.

Our research simulations of the collapse and fragmentation of cores in Ophiuchus reproduce some of the statistical properties of intermediate-mass binary systems very well. The simulations also form quite a large number of hierarchical multiples, including a stable hierarchical sextuple system.

Disc fragmentation

Most of the matter that is pulled towards a forming star by its gravity has too much angular momentum to accrete directly onto the star, and instead it collects in a disc around the star. Here, a variety of processes act to redistribute the angular momentum, allowing some of the material to spiral inwards and accrete onto the central star. However, there are two other possibilities. The first possibility is that the disc becomes sufficiently massive and cold that it fragments gravitationally to produce low-mass hydrogen-burning stars and brown dwarfs.

Our research has shown that the properties of the stars formed in this way match very well with the available observations. The critical issue here is that the luminosity generated by accretion onto the central star may heat the surrounding disc, thereby rendering it stable against fragmentation. We have recently shown that, since accretion onto the central star is often episodic, there can be periods between bursts of accretion when the disc cools down sufficiently to fragment; disc fragmentation only requires ~ 104 years, and is certainly one of the most promising ways of forming low-mass stars and brown dwarfs.

Numerical tools

Our principal modelling tools are Smoothed Particle Hydrodynamics (sph) and radiation transport. The sph code we use currently is seren developed by Hubble et al. These codes include robust treatments of the energy equation, the propagation of ionising radiation, and the creation of sink particles (to represent protostars). They also include the option to follow the long-term dynamics of protostars that form in simulations, using an accurate and fast N-body integrator. The radiation transport code we use is phaethon, which was developed by Stamatellos et al.and is widely used to generate spectral energy distributions and isophotal maps for comparison with dust continuum observations of interstellar clouds and protostars, like those made with herschel and spitzer.

In addition, we are working on developing and applying statistical descriptors of star-forming clouds and star clusters; for example, we developed the Q parameter, which is widely used to characterise young star clusters.


As a core contracts and fragments to form stars, the density rises rapidly, but the temperature remains low; then, when the stars start to form, the density in the envelope remains high and the temperature starts to rise. Under these circumstances a rich network of chemical reactions comes into play, with both gas-phase processes and processes on the surfaces of grains playing important roles.

The result is a great diversity of molecules, which we observe using the apex telescope, the hifi instrument on Herschel, the sofia airborne observatory, and the alma interferometer in the Atacama desert of Chile. These observations not only yield important constraints on the chemical and physical conditions in star-forming matter, but also information on chemical reactions under conditions that cannot be reproduced in the laboratory, and insights into the origin of the complex organic molecules required for life and the formation of water. In addition, there are certain reactions, like those that slowly replace hydrogen with deuterium in many molecules (e.g. H2O → HDO) whose progress can be used as a clock to estimate the age of a core.

Molecular clouds and feedback from massive stars

Formation and evolution of molecular clouds

In galaxies like the Milky Way, stars tend to form in the spiral arms, where interstellar matter is collected into cold, dense molecular clouds. The densest parts of a molecular cloud are called prestellar cores. These are the places where the matter is cold and quiescent enough to be held together by self-gravity, and is starting to condense into new stars.

Our research team has recently observed the most massive prestellar core ever identified, and still growing by accretion along a system of converging filaments. This core seems destined to form a massive star (i.e. a star that is at least ten times more massive than the Sun, at least ten thousand time more luminous, lives for at most thirty million years, and ends its life in a supernova explosion).

Feedback from massive stars

Although massive stars do not form often (and when they are also short-lived) they do have a profound effect on their surroundings. Their high luminosities and high surface temperatures mean that they emit a lot of energetic photons, which ionise and heat the surrounding gas, suddenly increasing its pressure by a factor of more than a thousand. Powerful stellar winds are generated before the star goes supernova.

We are using supercomputer simulations to explore how the radiation from a massive star affects star formation in its neighbourhood. The increased pressure compresses pre-existing cores, causing them to collapse quickly and form new stars. However these stars would have formed anyway, even without the radiation from the massive star - they would just have taken longer to do so.

Our research has found that the lasting effect from the massive star is actually to reduce the total amount of star formation, and the mean masses of the stars formed, by blowing away a large part of the original cloud.

Telescopes and instruments

During the early stages of star formation, when the matter destined to form stars is still quite extended and hence resolvable, it is very cold, typically T ≤ 10 K, and consequently it emits mainly at long wavelengths, 0.05 to 2 mm. At these long wavelengths the Earth's atmosphere is very opaque; there are a few bands where observations can be made from the tops of mountains, if the atmosphere is very dry, but otherwise it is necessary to send up telescopes on planes, or balloons, or satellites.

Matt Griffin lead the team that built the SPIRE instrument on the Herschel space telescope, which has increased the number of known prestellar cores by over an order of magnitude, enabling us to tie down their statistical and individual properties in unprecedented detail, showing that most cores are embedded in, and grow from filamentary structures.

Academic staff

Dr Paul Clark

+44 (0)29 2087 5107

Dr Jane Greaves


+44 (0)29 2087 5999

Dr Pantelis Papadopoulos

STFC Ernest Rutherford Fellow

+44 (0)29 2087 6649

Dr Nicolas Peretto

+44 (0)29 2087 4649

Dr Sarah Ragan

Marie Skłodowska-Curie Fellow

+44 (0)29 2087 4289

Professor Anthony P Whitworth

+44 (0)29 2087 4798

Associated staff

Dr Emily Drabek-Maunder

Research Associate

+44 (0)29 2087 6997

Dr Ana Duarte Cabral Peretto

Research Fellow

+44 (0)29 2082 0199

Dr Zeinab Khorrami

+44 (0)29 2087 0115

Dr Oliver David Lomax

Research Associate

+44 (0)29 2082 0199

Dr Kenneth Marsh

+44 (0)29 2082 0199

Dr Andrew Rigby

+44 (0)29 2087 0115