I have moved! New homepage https://homepages.dias.ie/jmackey/

From 1. January 2016 I have moved to the Dublin Institute for Advanced Studies. Please go to my new homepage for up-to-date information.


The sections below describe some of my recent work. Some projects have links to pages with more information and/or images and animations. The AIfA stellar physics group's ISM project page has more information about my previous research project as it relates to the interstellar medium. I also have a page describing the simulation code pion that I wrote and maintain.

Introduction to my research

Massive stars, despite being few in number and short-lived, are the main drivers of the evolution of gas in galaxies. They have strong stellar winds and emit ionising radiation, both of which heat the surrounding gas and drive strong outgoing shocks. When they reach the end of their lives, they explode as supernovae and/or gamma-ray bursts, heating the surroundings, driving strong shocks through nearby gas, enriching their environment with heavy elements, and their supernova remnants are important sites for acceleration of high-energy cosmic rays.

Quantitative modelling of most of these processes requires multi-dimensional simulations, so most of my work is numerical. By comparing our models to observed systems, we can learn about the evolutionary history of stars, thereby verifying and constraining stellar evolution models. We can then use these models to calculate the observable and dynamical consequences of interactions of supernovae with their surroundings. In addition, we can quantify the efficiency with which the energy emitted by massive stars gets transferred to kinetic energy in the interstellar medium.

NGC 6357 HST
Hubble Space Telescope image of Pismis 24 and NGC 6357, where winds and radiation from young massive stars are evaporating the cloud of gas they were born in.
[Image Credit: NASA, ESA, and J. Maíz Apellániz (Instituto de Astrofísica de Andalucía, Spain)]

Interacting supernovae from photoionization-confined shells around red supergiant stars

Authors: Jonathan Mackey, Shazrene Mohamed, Vasilii Gvaramadze, Rubina Kotak, Norbert Langer, Dominique Meyer, Takashi Moriya, Hilding Neilson.

Published in Nature, (August 2014), vol. 512, pp. 282-285.

Our paper addresses two separate problems that have so far no convincing explanation. First, the red supergiant star Betelgeuse is surrounded by a static shell of dense gas. This gas is made of Betelgeuse's stellar wind, driven from the star's surface, but has somehow been decelerated and piled up into a shell. Second, about 10% of supernovae (exploding stars) show that the stellar debris crashes into dense gas surrounding the exploded star, and it is difficult to understand how this dense gas comes to be there, so close to the star. We discovered a new mechanism by which winds from red supergiants can be decelerated and piled up into a shell around the star.

We realised that interstellar radiation (UV light that is almost everywhere in space) can ionize and heat up the outer regions of the cool wind ejected by the star, and this abrupt change in temperature drives a shockwave bask towards the star. This shockwave decelerates the stellar wind and piles it up in a shell around the star, which can trap up to 35% of all of the stellar wind. When we apply our model to the wind of Betelgeuse, we find that it matches the properties of the shell very well. This is the first convincing explanation of how this shell of gas has formed.

An approximate analogy is a hydraulic wave ("stopper") in a river, usually below rapids or a small waterfall. It is a wave that breaks upstream but never goes anywhere because it is fighting against the current of the river. Another analogy is a "tidal bore", where the incoming tide moves up the estuary of a river and drives a wave that breaks upstream. One of the best-known examples is the Severn Bore.

When Betelgeuse explodes as a supernova, the star's debris will crash into the shell after a year or more, and the force of this collision will light up the shell. It should be even more spectacular than the rings of light seen around supernova 1987A, and will certainly be much brighter because Betelgeuse is 300 times closer. The collision may even be visible with the naked eye (if we are still here when it happens!).


Betelgeuse Infrared

Herschel image of Betelgeuse's circumstellar medium at 70 microns (credit: image by Vasilii Gvaramadze, using data from ESA/Herschel/PACS/L. Decin et al.). The star is the bright central spot, surrounded by the inner shell. Betelgeuse is flying through space (here towards the top left corner) at faster than the speed of sound, so it drives a bow shock ahead of it.



Schematic diagram of Betelgeuse's wind

Schematic diagram of the circumstellar structures.
We explain the photoionization-confined shell that separates the neutral inner wind from the ionized outer part of the wind, and predicts its properties.
(credit: image by Shazrene Mohamed; a modified version of this figure can be found in our paper on the Nature website).


Runaway red supergiants: Betelgeuse

Authors: Jonathan Mackey, Shazrene Mohamed, Hilding Neilson, Norbert Langer, Dominique Meyer.

2012, ApJ Letters, 751, L10. (DOI, arXiv, ADS) Download a preprint here
JPEG figures: Fig.1 | Fig.2a | Fig.2b | Fig.3 | Fig.4 | Fig.5

Motivated by Shazrene's simulations of Betelgeuse's bow shock (ADS, A&A), we wanted to find an explanation for the low mass of Betelgeuse's bow shock, and to explore whether the mysterious upstream bar could be part of the explanation (image to the right).

We ran some 2D hydrodynamic simulations of winds from runaway massive stars as they evolved from blue supergiants to red supergiants, and at certain points in the evolution our model can explain the low mass of the bow shock and also has an upstream feature (a second bow shock) which may be a counterpart to the bar.

If we compare our model to observations, it implies that Betelgeuse may have about 20 thousand years until it explodes as a supernova at the end of its life (although there are obviously many uncertainties associated with this!).

Movies of the simulation can be found here:

Abstract from the paper:

A significant fraction of massive stars are moving supersonically through the interstellar medium (ISM), either due to disruption of a binary system or ejection from their parent star cluster. The interaction of their wind with the ISM produces a bow shock. In late evolutionary stages these stars may undergo rapid transitions from red to blue and vice versa on the Hertzsprung-Russell diagram, with accompanying rapid changes to their stellar winds and bow shocks.

Recent 3D simulations of the bow shock produced by the nearby runaway red supergiant (RSG) Betelgeuse, under the assumption of a constant wind, indicate that the bow shock is very young (less than 30 thousand years old), hence Betelgeuse may have only recently become a RSG. To test this possibility, we have calculated stellar evolution models for single stars which match the observed properties of Betelgeuse in the RSG phase. The resulting evolving stellar wind is incorporated into 2D hydrodynamic simulations in which we model a runaway blue supergiant (BSG) as it undergoes the transition to a RSG near the end of its life.

We find that the collapsing BSG wind bubble induces a bow shock-shaped inner shell around the RSG wind that resembles Betelgeuse's bow shock, and has a similar mass. Surrounding this is the larger-scale retreating bow shock generated by the now defunct BSG wind's interaction with the ISM. We suggest that this outer shell could explain the bar feature located (at least in projection) just in front of Betelgeuse's bow shock.

Simulation Snapshot

IRAS image of Betelgeuse at 60 microns (credit: Vasilii Gvaramadze). The star is moving towards the top left. The bright star produces image-processing artefacts.

Simulation Snapshot

Snapshot of a simulation showing gas number density (above) and temperature (below) at a time when the outer blue supergiant bow shock is collapsing onto the newly-expanding red supergiant wind. Here the star is moving from left to right. See movies and paper for more details.


Effects of magnetic fields on photoionised pillars and globules

Authors: Jonathan Mackey, Andrew Lim.

This is a follow-up project to the "Formation of Elephant Trunks" project described below. In the previous project interstellar magnetic fields were not considered, so here we performed the same simulations, but this time with magnetic fields of various strengths and initial orientations. It is very difficult to measure magnetic field strengths in the interstellar medium, so we wanted to see if they can have an observable effect on pillars. If so, then perhaps observations of pillars could be used to constrain the magnetic field strength.

To the right is an image from the Hubble Space Telescope of the star cluster NGC 3603 with two massive pillars nearby, and two evolved massive stars with shells/rings of circumstellar material. Below that are simulated observations of pillars produced in our simulations, with weak (top), medium strength (middle), and strong (bottom) magnetic fields. The magnetic field can confine ionised gas in dense sheets and filaments which are very bright in optical images, similar to what is seen in NGC 3603 and NGC 6357 (above). More work is needed to see if winds from the nearby star cluster could also produce such structures, but our work shows that magnetic fields can have strong effects.

Our work is complementary to similar calculations by Henney et al. (2009) and models with more realistic initial conditions by Arthur et al. (2011). Movies from their simulations are on YouTube.

NGC 3603 Simulation results

Top: HST image in H-alpha of NGC 3603 and its two massive pillars. The strong H-alpha emission in ionised gas closer to the cluster could be confined by magnetic fields as in our simulations, or it could be due to ram pressure from the stellar winds of the cluster stars.


Astrophysical Fluid Dynamics Code with Photo-ionisation

I have developed a grid-based fluid-dynamics/ray-tracing code as part of my PhD thesis. A description of the code together with results of test problems can be found here. The fluid dynamics module can solve the equations of hydrodynamics and ideal MHD on 1D/2D slab-symmetric, 2D axisymmetric, and 3D Cartesian grids. Photoionisation is modelled using the on-the-spot approximation and a ray-tracing module for either a point source or parallel rays.

Photoevaporating Clumps

Formation of "Elephant Trunks" in H II Regions

Authors: Jonathan Mackey, Andrew Lim.

Many HII regions are observed to contain massive parsec-scale pillars of gas and dust, often called elephant trunks (for example the Elephant Trunk nebula in IC 1396, or the famous pillars in the Eagle Nebula, M16). Their lifetimes and formation mechanisms are still poorly understood; to study this we set up a numerical models in which randomly placed dense clumps of gas were exposed to ionising radiation from a point source and allowed to evolve. We found pillars occasionally develop when different clumps partially shadow each other and merge into a column of dense neutral gas. This shadowing mechanism was further investigated with simulations of specific clump configurations using different thermal physics models, where we found that more realistic cooling models produced longer-lived and denser pillar-like structures.
See this page for more information.

Pillar Formation


Valid HTML 4.01 Transitional