Research Interests

Main topics of interest

Probing the interstellar medium with GRBs

Broad band GRB host galaxy properties

Cosmic chemical enrichment


The Gamma-Ray Burst Phenomenon
GRBs are broadly divided into long and short duration bursts, and it is the former variety that is associated with cosmological probes. The link between long GRBs and the death of massive stars has now been indisputably established, as has their cosmological origin, which stretches out to the epoch or reionization, a time when the first galaxies and stars were forming. Their extremely luminous and unobscured high-energy emission pinpoints regions of star formation irrespective of host galaxy luminosity, and as such makes GRBs potentially unique and effective tracers of the cosmic star formation rate (SFR). Their initial high-energy emission is followed by a mutliwavelength, luminous afterglow that shines through their host galaxy and provides an almost perfect background light source; they have extremely simple intrinsic spectra (simple power-law or single broken power-law), they are immensely bright at early times, and unlike QSOs, they fade to leave a clear, uncontaminated view of the host galaxy. This combination of properties make GRBs unlike any other probes of the high redshift Universe.


Probing the Interstellar medium with GRBs
The mostly unattenuated high-energy emission that makes GRBs such unique tracers of distant-star formation is also its shortfall, since it leaves no trace of the host galaxy and intervening medium that it travelled through. It wasn't until the discovery of the much longer lived, multi-wavelength GRB afterglow that we were able to measure the distance (redshift or z) of the GRB explosion, probe the chemical composition of its host galaxy, and localise the position of the GRB with arcsecond accuracy, enabling further investigation of the properties of the host galaxy long after the GRB even had faded.

Figure 1: Keck/LRIS spectra of the optical afterglow of GRB 080607 at z=3.06 (Prochaska et al., 2009). The red dotted line is the best-fit model of the intrinsic afterglow spectrum reddened by dust in the host galaxy, with a broad extinction feature centred at ~8800Å (2175Å restfame). A deep Lyman-α absorption trough identified with neutral hydrogen (HI) within the host galaxy is seen at λ≈4900Å corresponding to an HI column density NHI=1022.70±0.15cm-2. The model (cyan solid line) includes absorption from H2 in a GRB spectrum. The line opacity at λ>5500Å is dominated by metal-line transitions from gas in the host galaxy and includes bandheads of the CO molecule.


Gas Absorption
The light released by GRBs is attenuated by material within the host galaxy and other intervening systems along the line-of-sight that leave their unique finger print on the observed GRB afterglow spectrum, consisting of discrete absorption lines that correspond to the chemical make-up of each absorbing galaxy. Due to the stretching of light as it travels through an expanding Universe, these absorption lines are shifted to longer wavelengths (i.e. redshifted), and provided that the intrinsic, pre-redshifted energy of the absorption features is known, the distance that the light has travelled can be measured. Furthermore, the depth, shape and number of absorption features provide an encoding of the neutral hydrogen contents, metal fraction or metallicity, ionisation state, and kinematics of the gas in distant galaxies (see Fig.1).

Figure 2: Observed GRB Swift and GROND afterglow SED (data points and dashed lines). The best-fit intrinsic power-law spectrum (solid line) is also shown. Such an example GROND/Swift SED is routinely measured for ~40 GRBs per year.


Dust Extinction
Another source of attenuation is dust, which absorbs and scatters light in and out of the observer's line-of-sight, collectively referred to as dust extinction. The resulting attenuation is, however, relatively featureless in comparison to the absorption and emission lines produced by neutral gas and metals, and as such, a broad wavelength coverage is needed to accurately measure the dust extinction effect. The simple power-law or broken power-law GRB afterglow spectrum allows the broad, smooth shape cut-out by dust-extinction to be measured with relative ease, in large contrast to the complex spectra of QSOs. This is illustrated in Fig.2, which shows an example GRB afterglow spectral energy distribution (SED) (black solid line) that has been eaten away by intervening dust (red shaded area) and gas (blue shaded area), leaving behind the observed spectrum (data points and dashed line). The frequency range covered by the UV/optical and X-ray Telescopes onboard the space-mission Swift, and by GROND are also indicated to highlight the importance of the wavelength range covered by these instruments to measure the imprint left by dust extinction, and gas and metal absorption. The depth of the imprint left on the afterglow SED by dust extinction provides a measure of the abundance of extinguishing dust along the line-of-sight, typically quantified by the amount of visual extinction, Av.

Figure 3: Stellar mass vs. redshift of GRB hosts and field galaxies (Perley et al., 2013). Diamonds and circles are optically-selected GRB host galaxies from Savaglio et al. (2009) and Laskar et al. (2011), respectively. Hollow symbols indicate K-band faint hosts (K>23 Vega mag). Red squares correspond to galaxies from Perley et al. (2013) hosting GRBs with afterglows extinguished by Av >1 mag. Light gray dots correspond to field galaxies from MODS (Kajisawa et al. 2010), where the size of the data point is scaled by the galaxy SFR. The curves indicate the quartile boundaries in the expected distribution for K<23galaxies.


Broad band GRB host galaxy properties
GRB host galaxies have been predominantly investigated from ultraviolet to mid-infrared wavelengths, and our understanding of their emission properties at longer wavelengths have been impeded by the difficulty in detected these distant galaxies at submillimetre and radio wavelengths. Many historic GRB host galaxy follow-up campaigns at longer wavelengths selected targets based on the GRB afterglow and host galaxy properties, which omitted those heavily dust-extinguished GRB afterglows and host galaxies. Recent progress in detecting those more significantly dust extguished optical afterglows have resulting in a much larger fraction of detected dusty, and massive GRB host galaxies, with bright far-infrared and submillimetre emission. This is illustrated in Fig.3, which shows the distribution of GRB host galaxy stellar mass as a function of redshift for a sample of optically-selected host galaxies (blue data points; typically low-mass, dust-poor galaxies), and a sample of galaxies selected for hosting GRBs with heavily dust-extinguished afterglows (red data points; typically higher-mass and redder galaxies).






Figure 4: Herschel/PACS 100micron and 160micron images of the host galaxy of GRB070306 (Schady et al., 2014). The images are 2'x2' and 3'x3', respectively, centred on the GRB afterglow position (red cross), and ahve been smoothed using a Gaussian with sigma two times the pixel scale. The image is displayed with a linear greyscale ranging from -1 mJy pix-1 (black) to +1 mJy pix-1 (white), and contour levels from 1sigma to 6sigma are over plotted in black. The spatial scale is indicated on the top left.

The observed diversity observed in the properties of GRB host galaxies, from low-mass chemically-unevolved galaxies, to large, dusty galaxies, brings complexity in studying the dominant factors that contribute to the production of GRBs. To investigate this further we used the far-infrared Herschel space observatory to observe a sample of five GRB host galaxies, selected on the bases of strong evidence of dust along the GRB line of sight. These observations probed the galaxy dust-emission peak, enabling us to measure the galaxy dust temperature and mass, as well as the dust-obscured SFR. Surprisingly, we only detected thermal dust emission from one of the five GRB host galaxies in our sample (see Fig.4). We combined our sample with a further sample of 22 GRB host galaxies also observed with the Herschel observatory, and found that, although stellar mass is known to be related to the dust mass of galaxies, the Herschel detection rate of the combined sample of 23 GRB host galaxies suggested that the GRB afterglow dust-extinction is a better tracer of dust-rich host galaxies than the galaxy stellar mass. Only 35% of galaxies in the sample were detected, whereas this detection rate almost doubled when we only considered those galaxies that hosted significantly dust-extinguished GRBs. When considering only those galaxies with stellar mass M*>1010Msolar, the detection rate was still only ~40%. The high Herschel detection rate of GRB host galaxies suggests that the dominant dust component extinguishing the GRB afterglow lies within the galaxy interstellar medium, rather than in discrete dense clumps.


Scaling relations in GRB host galaxies

Figure 5: The M*-SFR-g parameter space for a sample of galaxies collected by Mannucci et al. (2010). Galaxies at z<2.5 were found to lie close to a place (the FMR) illustrated by the small, black dots.

The Fudamental Metallicity Relation The fundamental metallicity relation (FMR) is believed to represent the chemical evolution of galaxies from high to low redshift, by providing a close relation between their stellar mass, M*, SFR, and gas-phase metallicity (Zg). If verified, this would present a powerful tool for understanding the physical processes that drive the assembly and chemical evolution of galaxies. However, verification of the FMR is severely hampered at high z due to the inadequacy of standard emission-line Zg diagnostics and the difficulty in obtaining spectra for faint, low-M* galaixes. GRBs offer the perfect solution to this problem, by providing bright afterglows with high-quality spectra from which Zg can be measured in absorption, even in low-M* systems. Our group is currently is pursueing an observational campaign in order to acquire the host galaxy observations required to fully investigate this possibility.