Fig 1: A schematic diagram showing the underlying picture of the multifluid method and the MOGLI model. The box on the right shows an example grid showing three different cold gas structures in blue. The volume-filling fractions of cold gas is parametrised as α. The zoomed-in view in the middle shows the model’s assumption of the underlying cold gas structure as numerous spheres. Zooming in further, the left panel shows the different interactions in the MOGLI model, in particular drag, mixing and growth, along with other contributing variables, like the local turbulent velocity (vturb,local). © Hitesh Kishore Das, MPA
The space around galaxies might not glow brightly in telescopes, but it is, in fact, filled with gases at vastly different temperatures. From plasma at a million degrees Celsius to much colder, tiny, cold clouds at temperatures that can be found on Earth. Understanding how these gases interact is key to explaining how galaxies grow, form stars, and evolve. But the vast temperature difference has proved to be a significant challenge for simulations, as it also results in a big difference in densities. A team of scientists from MPA and AIP (Potsdam) has now developed a new model, MOGLI, that can track these interactions in unprecedented detail. By treating hot and cold gas as two coupled components that exchange material and energy, a multifluid approach, developed in engineering circles for numerous terrestrial applications, allows large cosmological simulations to capture the hidden life of cold gas.
The big bound systems of billions of stars called galaxies, of which our own Milky Way galaxy is one, sit inside an invisible halo of dark matter. The brighter part of the galaxy that is usually seen through a telescope sits close to the centre of this halo. The rest is filled with sparser gas, which is difficult to observe, but plays a very crucial role in shaping the galaxy and moulding its evolution. It acts as a gas reservoir that feeds the galaxy for future star formation, and receives gas when the existing stars explode as supernovae. This reservoir surrounding the galaxy is aptly called the circumgalactic medium.
This surrounding gas is far from uniform. Observations and simulations alike show that it exists in many different forms, hot, thin plasma which fills most of the volume, and small, cold clouds that contain a big chunk of the mass. These cold clouds are essential for feeding star formation, yet they are fragile, constantly buffeted by the hot wind around them. Whether they survive or evaporate determines how galaxies evolve over billions of years.
The cold clouds are so small compared to the galaxy that it is infeasible to simulate the galaxies while resolving the clouds. The key challenge is the vast difference in densities. In the best cases, it is like studying a mixture of air and water; at worst, it is like looking at the mixing of stone in air. Even accounting for Moore's law of advancing computational technology, a simulation with a sizable number of galaxies which also captures these interactions would take at least a century. So, astrophysicists usually have to compromise on accurately modelling these small-scale effects to simulate whole galaxies in practice. This leads to a limited prediction power with non-trivial dependency on assumptions made for such interactions. One workaround for accounting for small-scale effects that can not be simulated is to rely on “subgrid” recipes. A true subgrid model self-consistently keeps track of processes happening at the small scales without having to resolve those scales, removing the earlier restriction.
To address this, a joint team of researchers from MPA and AIP (Potsdam) developed MOGLI (Model for Multiphase Gas using Multifluid Hydrodynamics), a new framework that represents hot and cold gas as two distinct but interacting fluids. This model borrows the multifluid model implemented in AREPO, an in-house moving-mesh magnetohydrodynamic code at MPA, which is widely used in large-scale simulations. This multifluid method was originally devised to simulate more practical scenarios, similar to studying air bubbles in water. But the same method can also be used to simulate the hot and cold gas interaction, where instead of treating the mixture as a single fluid, the model keeps track of both components and how they exchange momentum, heat, and mass.
In MOGLI, the two components, namely the hot, thin gas and the cold, dense gas, influence each other through three key processes:
- Drag, which describes how the hot gas pushes or pulls on cold clouds;
- Turbulent mixing, where chaotic motions blend the two phases, and
- Growth, where cold gas forms as hot gas cools and condenses.
These interactions depend on the local properties of the astrophysical gas. For example, one important component is the local turbulence, much like how smoke disperses differently in a gentle breeze compared to a storm. By linking the model to measurable flow properties, MOGLI can adapt naturally to a wide range of environments, from the calm outskirts of galaxies to violent galactic winds.
A good model has to pass through a process of verification where it is tested against some benchmark scenarios, to show that a simulation with the subgrid model is indeed equivalent to a simulation that simulates the small-scale effects. In this case, more expensive, high-resolution turbulent box simulations were carried out very similar to the ones in a previous monthly highlight, and were chosen as the benchmarks. Low-resolution simulations with MOGLI were able to mimic the cold gas behaviour with respect to destruction rates, survival criteria and spatial dispersion from the benchmarks across a very wide range of simulation parameters. This breakthrough means that large-scale simulations can now include a more faithful representation of multiphase gas dynamics, bridging the gap between what telescopes observe and what computers can model. It opens the door to exploring long-standing questions: how does the cold gas reach so far from galaxies? How do outflows from galaxies recycle their gas? And what determines the mix of hot and cold gas we see in the halos of galaxies like our own Milky Way?
MOGLI is flexible by design. Future extensions could include magnetic fields, denser gas or even colder gas, which are thought to further influence how gas mixes and cools. For now, MOGLI provides a major step toward a more realistic picture of the turbulent, ever-changing environment that surrounds galaxies, giving a peek into the evolution of the tiny, cold gas which can persist and shape the life cycle of galaxies. This will hopefully lead to a new generation of subgrid models and large-scale simulations with stronger predictive capabilities to test our understanding of the Universe.
A good model has to pass through a process of verification where it is tested against some benchmark scenarios, to show that a simulation with the subgrid model is indeed equivalent to a simulation that simulates the small-scale effects. In this case, more expensive, high-resolution turbulent box simulations were carried out very similar to the ones in a previous monthly highlight, and were chosen as the benchmarks. Low-resolution simulations with MOGLI were able to mimic the cold gas behaviour with respect to destruction rates, survival criteria and spatial dispersion from the benchmarks across a very wide range of simulation parameters. This breakthrough means that large-scale simulations can now include a more faithful representation of multiphase gas dynamics, bridging the gap between what telescopes observe and what computers can model. It opens the door to exploring long-standing questions: how does the cold gas reach so far from galaxies? How do outflows from galaxies recycle their gas? And what determines the mix of hot and cold gas we see in the halos of galaxies like our own Milky Way?
MOGLI is flexible by design. Future extensions could include magnetic fields, denser gas or even colder gas, which are thought to further influence how gas mixes and cools. For now, MOGLI provides a major step toward a more realistic picture of the turbulent, ever-changing environment that surrounds galaxies, giving a peek into the evolution of the tiny, cold gas which can persist and shape the life cycle of galaxies. This will hopefully lead to a new generation of subgrid models and large-scale simulations with stronger predictive capabilities to test our understanding of the Universe.
A rendering of the evolution of 100 unresolved cold gas clouds with a radius 𝐿box/256,
where 𝐿box is the box size with 643 cells with MOGLI. Due to fast cooling, all the unresolved clouds grow. Without a subgrid model like MOGLI, it would require a hefty 30003 cells to run an analogous simulation.
Author:
Hitesh Kishore Das
PhD student
Multiphase gas dynamics
Tel: 2239
hitesh@mpa-garching.mpg.de
Original publication:
Hitesh Kishore Das, Max Gronke, Rainer Weinberger
MOGLI: Model for Multiphase Gas using Multifluid hydrodynamics
accepted by MNRAS
Source | DOI
