IAN SMAIL

Home Research Current Diary Publications Gallery Morphs Group

CV


RESEARCH INTERESTS

This page gives brief summaries of the main areas covered in my research. More details about some of my current projects can be found on this page.

My research interests fall into three main areas: studying the most active phases of galaxy formation and evolution using obscured galaxies at high redshifts, understanding the growth of groups and clusters of galaxies structure through mapping their dark matter with gravitational lensing, and investigating the influence of environment on the evolution of galaxies by contrasting the properties and evolutionary histories of galaxies in high-density regions such as clusters with those in the low-density field.


Dusty Galaxies

Vigorously star forming galaxies are characterised by large quantities of dust [1], this absorbs light from stars and reprocesses it into the far-infrared (50-200um). Thus the most active galaxies in the local Universe have very high far-infrared/optical ratios, compared to more `normal' galaxies such as the Milky Way and are a major constituent of space-based surveys in the far-infrared (these are impossible from the ground as the atmosphere is opaque at these wavelengths). However, it was only with the advent of the SCUBA submm camera on the James Clerk Maxwell Telescope on Hawaii, that we had the capability to search for similar systems at higher redshifts, where the restframe far-infrared emission is redshifted into the submm waveband. Such dusty, luminous galaxies were expected to be a major component of the galaxy population in the early Universe as young galaxies should be actively forming stars.

I was lucky to be a member of a small team of young researchers who had the opportunity to obtain the first deep submm maps with SCUBA [2] when the detector we were scheduled to use on the JCMT failed. We obtained maps of the core regions of two massive clusters of galaxies and detected a total of five sources, this survey was subsequently expanded to cover seven clusters and yielded a sample of 15 background submm galaxies [3]. The most recent cluster lens field to be observed has yielded a striking giant arc of an intrinsically faint submm galaxy, underlining the usefulness of lensing as a tool for such surveys [34].

Initial follow-up of these fields using deep, archival HST optical imaging [4] and Keck spectroscopy [5] suggested that many of these sources had optically bright counterparts. However, subsequent work use very sensitive radio maps [6], deep near-infrared imaging [7, 35] and millimeter interferometry [8] has revised this view to the point where we believe over half of the population are so faint in the optical that they are difficult to detect in even the deepest images, although 25% of these may appear as Extremely Red Objects (EROs) in the near-infrared. The ERO population has long been thought to contain examples of dusty starburst [9,10] and AGN [11].

While many of the submm galaxies are faint in the optical, prompting suggestions of new techniques to learn more about their properties [12], a few examples are sufficiently bright that they can be studied using large ground-based telescopes. These detailed studies have yield important insights into the nature of the faint submm population [13,14,15,16,17,18], including the source of the immense luminosities of these systems, their masses and gas fractions [19,20]. These observations suggest that the submm galaxies are massive, gas-rich systems which are forming stars at a rate of 1000's Mo/yr over a region 10-kpc in extent [37]. Although a small fraction may be artificially boosted by foreground galaxy-scale lenses [21,33], or represent less-luminous, but colder galaxies at lower redshifts [22].

We have has worked to interpret the properties of these galaxies and place them within the framework of models for galaxy formation and evolution [23,24]. This work suggests that the submm population may contain the bulk of the massive star formation at high redshifts, and that this is almost completely hidden in the optical and UV [25,26]. We have also worked on the relation between the submm population and other classes of high-redshift sources such as X-ray selected AGN [27] or mid-infrared sources [38 ,39 ,40], as well as the observational opportunities of future surveys [28], with second-generation instruments such as SCUBA-2.

My most recent work on the submm population has concentrated in part on targetted surveys of regions around known high-redshift sources, Quasars or luminous radio galaxies [42], to search for overdensities of submm sources which may mark the cores of proto-clusters in the process of collapse [29,30,43]. This project has proved successful and we are actively expanding the survey to provide a more representative view of the highest density regions in the early Universe [44], and the relationship between submm galaxies and other classes of high-redshift galaxy and AGN [45].

The other focus of my recent work has been to exploit the precise positions for a large fraction of the brightest SCUBA galaxies available from deep radio maps [31], combined with the massive light-grasp of 10-m class telescopes, to undertake the first large spectroscopic surveys of this population [32]. This has facilitated the first reliable measurement of the spatial clustering of these galaxies, confirming that they are strongly clustered, as expected if they are truly massive galaxies [41, 36].


Gravitational Lensing

My PhD thesis covered various aspects of gravitational lensing by massive clusters of galaxies. The first project involved spectroscopic observations of a giant arc in Abell 963, these confirmed that the arc represented a highly distorted (and magnified) image of a background star-forming galaxy at z=0.77 [1]. I followed that up with an analysis of archival images of distant clusters from the Danish 1.5-m telescope at La Silla (taken for an early high-redshift supernova survey) to search for gravitationally-lensed arcs and arclets [2]. This turned out to be only moderately successful, in part because the clusters surveyed were not particularly massive and X-ray luminous, but mostly because of the limited spatial resolution available from ground-based imaging - which makes identification of lensed features difficult, although that isn't to say I've given up trying [3].

However, it is the high-resolution imaging available from Hubble Space Telescope which is essential for successful lens surveys [4,10,15]. HST is also invaluable for studying the properties of the arcs, such as their morphologies, internal structure, colours and most recently dynamics, in much more detail [5,6,24] than is possible from the ground [7]. HST also provides unique opportunities to exploit gravitational lensing to undertake sensitive surveys for high-redshift sources, such as supernova [8].

My thesis dealt with two aspects of gravitational lensing: i) the use of lenses to estimate the redshift distribution of distant field galaxies and ii) weak shear reconstructions of the mass distributions in cluster lenses.

Lenses can be used to estimate the redshift distribution of the background field population by comparing the strength of the distortion imprinted on the galaxy shapes by cluster lenses with similar masses at different redshifts. At its simplest, if a lens at redshift z and is beyond the bulk of the faint field population then the average distortion induced in the field population will be lower than for an equivalent mass lens at lower z, [9]. Unfortunately, the strength of the lensing signal will also appear lower if the more distant cluster is less massive than the lower redshift one - a not unreasonable expectation given that clusters are believed to grow significantly over the redshift range populated by faint field galaxies, z=0-1.

This technique has also benefited from the use of HST, which along with detailed mass models of cluster lenses (constrained by the geometry of multiply-imaged arcs identified in the core of the clusters), allow us to not only estimate the statistical redshift distribution for the field population, a factor of 10× fainter than is possible from classical spectroscopy with even 10-m telescope [10], but also predict the redshifts for individual arclets to 10-30% precision [11]. We subsequently confirmed the accuracy of this new technique for the brighter arclets which are within the reach of classical spectroscopy [12,13].

The other field which I have tackled with lensing is measuring the mass of clusters and mapping the distribution of dark matter within them [14]. Again HST has been a great benefit for this work, allowing me to study a wider range of cluster masses [15], across a wider field of view [25] at higher redshifts [17,18] and with better resolution than is possible from the ground [16,26]. The mass distributions derived in this way can be compared and contrasted to those assumed for X-ray analyses of cluster mass distributions [26,27]. Testing the assumptions used by the X-ray analyses is particularly important when these are used to infer constraints on cosmological parameters [28]

Finally, weak lensing studies can be extended beyond rich clusters to study the mass distribution on the largest scales in the Universe [19], as well as providing unique insights into the nature of dark matter [20] as well as the properties of the dark matter halos of galaxies and their variation with environment, epoch and morphology [21,22,23].


Cluster Galaxies

Deep lensing surveys through massive clusters also yield high-quality information on the galaxies residing within the clusters [1,2]. This makes surveys of rich clusters a particularly fruitful area of research, especially when HST imaging is obtained as part of the programme [3,4]. These data, combined with spectroscopic information, can be analysed to investigate the effects of environment on galaxy characteristics and their evolution [6,7,28].

One project which has exploited deep HST of rich clusters to investigate the properties of the cluster galaxies is the MORPHS project [5,6]. The MORPHS group obtained HST imaging of 10 clusters at z > 0.37 and analysed these to provide morphological information on roughly 2000 galaxies in these fields. This influential project has produced unique constraints on a variety of cluster-related issues, including the photometric, spectroscopic and morphological evolution of cluster galaxies [7,8].

The most striking discovery from the MORPHS project deal with the evolution of the early-type galaxies which dominate the cluster population. Morphologically classified samples show a steep decline in the proportion of lenticular (S0, galaxies with a large bulge component and a weak disk) to elliptical galaxies (systems consisting almost entirely of bulge stars) out to z=0.5 [9]. Taken together with the expectation that elliptical galaxies are old, high-evolved systems [7,8,27], this suggests that the lenticular galaxies were formed relatively recently (or more likely transformed from a different morphological class). This suggestion has provoked a lively debate [12,13,14], and various interpretations [15]. The existence of morphological evolution in rich clusters is in little doubt, however the nature of the processes which drives it is still an open question.

My most recent work on the properties and evolution of galaxies in high density environments has focused on testing the evolution of galaxies outside the highest density (core) regions probed by HST surveys [16,17]. By following the changing properties of galaxies (colour, star-formation rate or morphology) from the core out to the edge of the cluster (where it blends into the overall large-scale structure) we can test different environmental processes which have been proposed [18,19,20,21,26]. To undertake these tests however we require spectroscopic confirmation of membership, which can be observationally demanding to obtain [22,23].

Finally, I have started to investigate the role that dust plays in altering our perception of the nature of cluster galaxies at high redshifts. Dust is a by-product of massive star formation and at the same time obscures it from sight. Thus it is possible to hide relatively large amounts of star formation in otherwise quiescent looking cluster galaxies [24,25].


Field Galaxies

My work on the properties of lensed galaxies sparked an interest in the (unlensed) field galaxy population - as studied through the deepest imaging in optical [1,2] and near-infrared passbands [3]. I have also studied the clustering properties of the faint galaxies detected in these deep images [4,5], which has illustrated the ambiguities inherent in such approaches. My more recent work on the faint field population has concentrated on a small subset of galaxies which exhibit extremely red optical-near-infrared colors, these Extremely Red Objects (EROs) comprise passive, evolved galaxies at high redshifts whose colors are dominated by old, red stars [6,7], and more actively star-forming galaxies which are reddened by the presence of large amounts of interstellar dust [8]. The morphologies of the different population can also provide a useful guide to their relative abundances [9]. Once again, the magnifying power of gravitational lenses has proved a very useful tool for studying these faint galaxies in great detail [10].


Miscellaneous

And then we are left with the research which doesn't fit into the tidy categories described above. These are many random things which you accrete as you go through life - due to being at telescope at wrong time [1,2] or too often [3], listening to wrong seminars [4] or just being on vacation [5].