This page describes the research I carried out for my PhD thesis. If you prefer, you can jump straight to the publications list.

Observations of distant galaxies have revealed a population quite unlike that found in the local Universe; at high redshift, galaxies form stars more rapidly than they do locally, and their morphologies are often dominated by large clumps. While large-scale statistical studies can reveal these trends in overall properties, a detailed understanding of the processes involved in galaxy evolution in the early Universe requires high-resolution, spatially-resolved data.

Even with 8-10m telescopes and advances in adaptive optics, current facilities limit the spatial resolution achievable at z ~ 2 to around 1-1.5kpc. This is an order of magnitude larger than the scales of typical star-forming clumps locally, making direct comparison difficult.

My PhD research at Durham has focused on overcoming this problem by the use of gravitational lensing. By targeting distant galaxies that lie behind massive foreground clusters, we are able to benefit from lensing magnification factors of up to 50x, resulting in spatial resolution of order ~100pc. Thus, we can isolate individual clumps or HII regions in high-redshift galaxies with 'normal' luminosities and star formation rates for their redshift, and study them in unprecedented detail.

Previous work (see e.g. Swinbank et al., 2009; Jones et al., 2010) has shown that these high-redshift clumps follow similar scaling relations in size and luminosity to HII regions in local spiral galaxies, but offset by a factor of ~100x in luminosity; i.e. they are forming stars 100x faster in a given area than local HII regions. This makes them similar to the most intense star-forming regions found in interacting galaxies in the local Universe (such as the Antennae at the top of this page), but are ubiquitous in apparently non-interacting systems at high redshift.

In an effort to understand this result further, we obtained narrowband imaging of a sample of eight gravitationally-lensed galaxies at z=1-1.5. These were selected so that their Hα emission lines fell within one of the available narrowband filters. Using the closest broadband filter for continuum subtraction, we could generate Hα excess maps of the galaxies and thus measure the sizes and luminosities of any star-forming clumps. We found 57 clumps in the sample, and found them to be remarkably well-matched in surface brightness, and to exhibit clear evolution between local HII regions and the z>2 star-forming clumps of previous work.

Star formation rate (SFR) vs size for star-forming clumps at different redshifts. The dashed lines indicate the median surface brightness for each sample (from Livermore et al. (2012).

The apparent redshift evolution between 'normal' HII regions in the local Universe and the intense star-forming clumps at high-z indicates that the latter may simply be scaled-up equivalents of the former. This is further suggested by the evolution in the HII region luminosity function, the break of which shifts to higher luminosity at higher redshift (see Livermore et al. (2012) for details).

These trends can be partially explained by increasing gas fractions; higher gas fractions in high redshift galaxies would lead to collapse on larger scales, resulting in larger and more intense star-forming regions.

Observations of cold molecular gas, the fuel of star formation, is difficult at high-redshift galaxies, and the statistics are correspondingly small. Nonetheless, several studies have found trends of higher gas fractions at high redshift (see e.g. Tacconi et al., 2010; Daddi et al., 2010; Geach et al., 2011).

In 2010, we obtained time on the IRAM Plateau de Bure Interferometer to search for cold molecular gas in a representative galaxy at z=4.9, which is magnified by gravitational lensing. The detection was tentative, but is indicative of a high gas fraction of ~0.6, consistent with the trends observed at lower redshifts (see Livermore et al. (2012b) for full details).

While high gas fractions in high-redshift galaxies go some way towards explaining the intense star-forming clumps, taken alone they overpredict the star formation rates. In Livermore et al. (2012), we therefore suggested that there must be a compensating factor from higher rotation frequencies within high-redshift galaxy disks.

Accordingly, my current work is focussing on the dynamics of high-redshift galaxy disks and their star-forming clumps. To this end, I am using Integral Field Spectroscopy on instruments such as SINFONI on the Very Large Telescope (VLT), NIFS on Gemini and OSIRIS on Keck. We have constructed a sample of 10 gravitationally-lensed galaxies at z ~ 1.5 - 4, which we are using in conjunction with the previous samples mentioned above to glean a better understanding of the detailed dynamics of high-z galaxy disks.


  • A. Adamo, G Östlin, N. Bastian, E. Zackrisson, R.C. Livermore and L. Guaita 2013 ApJ 766 105: "High-resolution study of the cluster complexes in a lensed spiral and redshift 1.5: constraints on the bulge formation and disk evolution" ADS link

  • R.C. Livermore, T. Jones, J. Richard, R.G. Bower, R.S. Ellis, A.M. Swinbank, J.R. Rigby, Ian Smail, H. Ebeling and R.A. Crain 2012 MNRAS 427 688: "Hubble Space Telescope H-alpha imaging of star-forming galaxies at z=1-1.5: evolution in the size and luminosity of giant HII regions" ADS link

  • R.C. Livermore, A.M. Swinbank, Ian Smail, R.G. Bower, K.E.K. Coppin, R.A. Crain, A.C. Edge, J.E. Geach and J. Richard 2012 ApJL 758 35: "Observational limits on the gas mass of a z=4.9 galaxy" ADS link

  • J. Richard, T. Jones, R.S. Ellis, D.P. Start, R.C. Livermore and A.M. Swinbank 2011 MNRAS 413 643: "The emission line properties of gravitationally lensed 1.5 < z < 5 galaxies" ADS link

  • T. Yuan, L.J. Kewley, A.M. Swinbank, J. Richard and R.C. Livermore 2011 ApJL 732 14: "Metallicity gradient of a lensed face-on spiral galaxy at redshift 1.49" ADS link