Three-dimensional radiative MHD simulations of near-surface convection in main sequence cool stars

 Andrea Perdomo García has recently completed her PhD thesis, Three-dimensional radiative MHD simulations of near-surface convection in main sequence cool stars, supervised by Manuel Collados and myself.

Andrea’s thesis explored the atmospheres of cool main-sequence stars through realistic three-dimensional radiative MHD simulations with the MANCHA code. A major part of her work was devoted to one of the central challenges of such simulations: treating radiative transfer accurately while keeping it computationally feasible. She developed and tested opacity-binning strategies for stars ranging from F to M spectral types, investigated the increasingly important role of molecular opacity toward cooler stars, and studied how different opacity treatments affect atmospheric structure and radiative energy exchange.

She then applied these methods to 3D simulations of G2V, K0V, and M2V stars, exploring their convection and photospheric structure and extending the calculations to magnetic simulations in which fields generated through the Biermann battery were amplified by small-scale dynamo action.

It has been a true pleasure to work with Andrea over the years and to follow her development as a researcher. I wish her all the best in her future career and in her new position at the Max Planck Institute in Heidelberg. I am sure there are many interesting problems, simulations, and probably quite a few opacity tables still ahead!

Hydrodynamic simulations of cool stellar atmospheres with MANCHA

 

 

Perdomo García, A., Vitas, N., Khomenko, E., Collados, M., A&A, 688, A27 (2024)

arxiv

Three-dimensional simulations do much more than reproduce what we observe on stellar surfaces. They give us a unique laboratory in which we can experiment with stellar atmospheres: change the physical ingredients, follow their interaction, and understand how convection, radiation and atmospheric structure are connected.

In this paper, we used the MANCHA code to simulate the atmospheres of three cool main-sequence stars, G2V, K0V and M2V, and explored one particularly important ingredient: radiative transfer and opacity. We tested how different approximations to the enormously complex stellar opacity affect radiative energy exchange and, ultimately, the atmospheric structure.

The results show how increasingly important a realistic treatment of opacity becomes as we move toward cooler stars, where molecules start to dominate the radiative properties of the atmosphere.

Another important piece of Andrea’s PhD work, and another step toward realistic 3D modelling across the cool-star sequence. 

Moving to the dark side!

 

I have some big news to share. From March 1, I’ll be starting a new chapter and moving to the dark side of astrophysics: the substellar world of ultracool dwarfs!

This seems like a natural moment to make the move. With the Euclid ESA mission now opening an enormous new window on the faint and cool populations of the Milky Way, we are entering an era in which the number of known ultracool and substellar objects will increase substantially. Turning this wealth of new observations into physical understanding will require careful work on the modelling side.

In some ways, though, I am not moving very far at all. My main interests will remain what they have been for years: developing numerical agoriths for radiative transfer, equations of state, opacities, and numerical modelling of atmospheres. The difference is that I will now be applying them much further down the temperature scale. At the temperatures of ultracool dwarfs, molecules dominate the opacity, chemistry becomes increasingly tricky, condensates and clouds appear, and the coupling between chemistry, radiation and atmospheric structure gets the central stage in modelling problem. 

I’ll be joining the SUBSTELLAR project led by Eduardo Martín, one of the co-discoverers of Teide 1, the first confirmed brown dwarf, announced in 1995. The work is supported by the European Research Council through the ERC Advanced Grant SUBSTELLAR, devoted to pushing the frontier of substellar science with the Euclid mission.

A new wavelength regime, a new class of objects, and plenty of new physics — but, fortunately, still lots of opacities, radiative transfer and numerical challenges. I hope that this move will also allow me to bring some of my experience in stellar astrophysics across the boundary between these fields. I’m looking forward to seeing where it leads.

May the (gravity and Lorentz) Force be with me!

Opacity for realistic 3D MHD simulations of cool stellar atmospheres

The first paper of Andrea Perdomo Garcia is just submitted for publication in Astronomy & Astrophysics, and out on arxiv.org/abs/2306.03744. The paper is all about computing the opacities for realistic modelling of cool stellar atmospheres. It is divided in three unities. First (Section 3) it describes the computation of detailed monochromatic opacity including millions of atomic and molecular spectral lines and millions of wavelength points. For this the code SYNSPEC (Hubeny and Lanz, 2011, 2017a, b) is used. Then (Section 4) the monochromatic opacities are used to construct opacity distribution function which reduces the number of wavelength points from millions to thousands. The results are compared in detail with ones produced by Kurucz. Some striking similarities and some warning differences are found. Finally (Section 5), the opacity distribution function to construct opacity bins. This method, originally proposed by Nordlund (1982) is the key ingredient for realistically simulating stellar atmospheres in 3D as it reduced the problem further, from thousands of wavelength points to only a few. However, the method depends on a choice of some free parameters. In our paper the possible choices are carefully analyzed and some interesting conclusions are offered. 

In Sect.3 there are two figures (Figs.2 and 3) that I find very useful and illustrative. The monochromatic opacity (Fig.2) and the radiative heating rate (Fig.3) are shown as 2D functions of wavelength (X-axis) and height in the atmosphere (Y-axis) for four different cool stars (all with solar metalicity). Optical depths in the continuum and continuum+lines are overplotted.

(Andrea is the final year PhD student at Instituto de Astrofisica de Canarias and Univeridad de La Laguna, supervised by Manolo Collados Vera and myself. Stay tuned, more cool stuff is coming out from her research this year.)

Charles Hermite (1822 - 1901)

unday morning in Paris offered an opportunity to walk to the Montparnasse Cemetery and pay my respects to some of my personal heroes buried there.

While the graves of Beckett, Cortázar and Poincaré attract plenty of attention and visitors, it is perhaps less well known that the great French mathematician Charles Hermite is buried there as well.

Hermite was not only a mentor to Henri Poincaré, Henri Padé, Thomas Stieltjes and Mihajlo Petrović Alas. His work on interpolation and function approximation lies at the very foundations of many modern numerical methods used in computational fluid dynamics and radiative transfer, even if that connection is not always obvious, or properly acknowledged.

His 1878 paper, Sur la formule d'interpolation de Lagrange, is still worth reading for anyone interested in function approximation, almost 150 years later.

Today, even the name on his gravestone is barely readable.

Robert Jelle Rutten (1942 - 2022)

So long, my teacher and friend. You will be hugely missed.

1D Solar Atmosphere Models in IDL: Standard Solar Model (from "The Sun", by Stix)

Stix, in his book ("The Sun: An Introduction", Springer, 2002) in Sect.2.4 introduces a standard solar model from the core to the surface defined as $\tau = 2/3$. His Table 2.4 (on p.56) lists the values of various quantities of this model versus the column mass or the height. Check the book for the details of the model. The table in ascii format is here: stix_tab.2.4.txt The columns are: $m/m_\odot$, $r/r_\odot$, $p \mathrm[Pa]$, $T \mathrm[K]$, $\rho \mathrm[kg/m^3]$, $L/L_\odot$, $X$, $\mu$, $\Gamma_1$.