Ashley Chrimes
Gamma-ray bursts mark the spectacular explosions of massive, rapidly spinning stars. The strongly beamed jets produced are visible to us as flashes of high energy electromagnetic radiation when Earth happens to lie within the beaming angle. They are among the most powerful cosmic explosions known, visible across much of the observable Universe. As the jet decelerates upon colliding with material around the progenitor star, the opening angle increases, the energy scale drops, and we observe an ‘afterglow’. The light curve and spectral behaviour of GRB afterglows can be used to infer the properties of the circumstellar medium. Specifically, these afterglows typically exhibit behaviour consistent with moving through a wind-like density profile, expected close to a progenitor star, or a constant-density medium, normally interpreted as the interstellar medium (ISM).
However, early stellar wind modelling struggled to produce transition radii (where the stellar wind transitions to the ISM) close enough to the star to explain the observed proportions of wind and ISM-like bursts. These studies have also been limited the lack of large samples of well-studied afterglows that have been modelled in detail to yield key values such as the emission radius, wind parameter A and ISM density n.
Recent advances in stellar evolution and hydrodynamic wind modelling, and over a decade of GRB afterglow data, mean that we can now re-address these questions. We describe our work using GRB progenitors identified in the BPASS stellar evolution models, and an expanded and re-analysed afterglow sample. The models have been selected for their agreement with GRB rates, galactic environments and other properties. We aim to determine if the range of wind radii, A and n required by observations can be reproduced, and if so, under what conditions. Our aims, methodology and some preliminary results are presented.
As the jet decelerates, the peak emission frequency drops, and we observe a fading afterglow which can be seen from X-rays to radio waves. The density profile of the circumstellar medium can be inferred from afterglow observations. GRB circumstellar media can be constant density, like the interstellar medium, or wind-like, with a power-law density profile, e.g. Chevalier et al. 2004 and Li et al. 2015.
Simulations of progenitor winds have struggled to produce constant-density environments close to the progenitor, the variety of wind and ISM densities observed, and predictions for the distributions of these parameters. E.g. Eldridge et al. 2006, van Marle et al. 2006.
To model progenitor winds, we use the BPASS models, detailed in Eldridge et al. 2017, which incorporate binary interactions. The model outputs include mass loss rate, but not wind speed. We employ the wind speeds of Lamers et al. 1995, revised down for high temperatures based on recent work, e.g. Vink et al. 2018. Wolf-Rayet wind speeds from Lamers & Nugis 2000 are used.
Selecting models that satisfy GRB requirements, we can compare to the observed rate and produce other distributions such as metallicity. Chrimes et al. (2020) showed a two-pathway model of tidally—spun fast rotators, and accretion—spun quasi-homogenously evolving stars, could fit the rates and Z distribution. These models are used for our wind study.
There are three parameters of interest—the termination shock radius, the wind density for wind-like afterglows, and the constant-density equivalent. We calculate the termination shock radius analytically for every GRB model, using the three-wind model of Garcia-Segura & Mac Low 1995. The phases are main sequence, supergiant, and Wolf-Rayet.
Analytic solutions are valid for supersonic shocks and cold interstellar media. Hydrodynamical simulations are too computationally expensive to run for every model, but include thermal pressure, cooling and instabilities. For a model subset, we calculate the shock radius using the PLUTO code, quantifying uncertainties in the analytic approach. The wind is initialised with the density parameter, proportional to mass-loss rate divided by wind speed.
Trial runs confirmed that PLUTO produces the same results as the analytic approach when the analytic assumptions are met. The shock radius for a 60 solar mass, solar metallicity star is shown as a function of time. The hydrosimulation equivalent, at t=1.5 Megayear, is also shown. The shock radius is approximately 10 parsec in each case.
The analytic calculations were applied to every GRB model. Bimodality in the shock radius distribution corresponds to QHE models, which have low Z, weak winds and longer lifetimes, and tidal models which can occur with higher Z and stronger winds. The wind density distribution has two tidal peaks, corresponding to different spectral types of progenitor.
Collaborator Ben Gompertz is re-analysing and expanding the afterglow sample of Gompertz et al. 2018. A plot from this paper shows the wide density range, from 10 the minus 6 to 10 the power of 2. This spread is expected to reduce when the afterglows are fit with larger datasets, allowing more parameters to be constrained.
Further work includes finishing the grid of hydrosimulations, and completing the afterglow sample analysis. Can the complex shock structure seen in hydrosimulations explain the range of ISM-like densities seen, or is a large variety of environmental densities required? And what further insights can be gained into GRB progenitors and environments?