Research Projects

  • image

    The THESAN project

    Reionization meets galaxy assembly

    The THESAN project is a suite of large volume radiation hydrodynamic simulations that self-consistently model the reionization process and the galaxies responsible for it with unprecedented physical fidelity. THESAN provides higher resolution than previous simulations of comparable volume and employs galaxy formation models known to produce physical properties in concordance with observations down to the present-day Universe. Such an approach is ambitious but essential to push the frontier in our understanding of the intergalactic medium and its connection to galaxies during the first billion years after the Big Bang.

    Visit the project website for more details.

  • image

    Building realistic ISM models

    Effect of radiation fields, molecular thermochemistry and dust

    We have developed a novel framework to self-consistently model the effects of radiation fields, dust physics and molecular chemistry (H$_2$) in the interstellar medium (ISM) of galaxies. The model combines a state-of-the-art radiation hydrodynamics module with a non-equilibrium thermochemistry module that accounts for H$_2$ coupled to a realistic dust formation and destruction model, all integrated into the new stellar feedback framework SMUGGLE. We test this model on high-resolution isolated Milky-Way (MW) simulations. We show that photoheating from young stars makes stellar feedback more efficient, but this effect is quite modest in low gas surface density galaxies like the MW. The multi-phase structure of the ISM, however, is highly dependent on the strength of the interstellar radiation field. We are also able to predict the distribution of H$_2$, that allow us to match the molecular Kennicutt-Schmidt (KS) relation, without calibrating for it. We show that the dust distribution is a complex function of density, temperature and ionization state of the gas which cannot be reproduced by simple scaling relations often used in the literature. Our model is only able to match the observed dust temperature distribution if radiation from the old stellar population is considered, implying that these stars have a non-negligible contribution to dust heating in the ISM. Our state-of-the-art model is well-suited for performing next generation cosmological galaxy formation simulations, which will be able to predict a wide range of resolved ($\sim 10$ pc) properties of galaxy.

    As an example we show the movie of the MW simulation showing a false color RYB movie of the disc as seen face on. The RYB image is constructed from the infrared (red), optical (yellow) and ionizing UV radiation (blue) fields generated self-consistantly from the simulations.

  • image

    Radiation feedback on giant molecular cloud scales

    Understanding the role of photoheating and radiation pressure feedback in dispersing giant molecular clouds

    We have developed a model for the interaction between dust grains and radition fields. This model is then used to test the effectiveness of photoheating, stellar winds and both single (UV) and multi-scattered (IR) radiation pressure in dispersing giant molecular clouds on short timescales. One of the most exciting new results is that the dust grains segregate increasing the dust-to-gas ratio in high density regions, which in turn increases the effectiveness of radiation pressure feedback mechanism.

    The movie shows the dispersal of a massive GMC due to radiation pressure feedback from the newly formed massive stars.

  • image

    Early stellar feedback

    Role of photoheating and radiation pressure magnifying SNe driven outflows

    Recent works have shown that SNe feedback alone is inefficient in driving galactic scale outflows. As the stars form in the highest density peaks, injection of SNe energy into these high density regions results in a very short Sedov-Taylor phase thereby generating low amounts of momentum input. However, if photons from massive stars can preprocess the star formation sites through photoheating and radiation pressure then the SNe can go off in highe temperature low density medium instead therby increasing its impact and driving large scale outflows. To test this scenario I am currently running small scale simulations of a patch of the ISM (1 x 1 x 10 kpc), with the aim to resolve the Stromgren radius around the stars. These simulations will be the first to self-consistantly model the effect of photoheating and radiation pressure in the ISM. Initial results suggests that photoheating and radiation oressure can ditrupt the disk more efficiently and can lauch larger velocity outflows compared to the run with only SNe.

  • image

    AGN-ICM interaction

    Impact of thermal conduction in clusters

    The most common mechanism invoked to explain the quenching and continued quiescence of massive galaxies is feedback from the central supermassive blackhole. However, the exact details of how AGN feedback couples to the surrounding gas in galaxies is still not properly understood, so the modeling efforts have been necessarily crude. There has been a recent push towards kinetic AGN feedback which has had general success in achieving self-regulation. However, it has been known for some time that purely hydrodynamic jets form low-density channels through which the jet can freely flow out, leading to very inefficient heat and metal mixing within the core. This means that an uncomfortably large amount of blackhole energy is needed in order to offset cooling loses in the core.

    However, I show in cosmological simulations (Kannan et al., Submitted) that AGN-ICM coupling can be drastically improved in the presence of anisotropic thermal conduction. It is considerably easier to mix thermal plasma in the presence of conduction, since the plasma is formally always buoyantly unstable and thus already prone to mixing! The buoyancy instabilities lower the restoring buoyancy forces leading to efficient mixing of the thermal plasma even with low levels of external turbulent driving. This efficient turbulent mixing readily isotropizes the injected AGN feedback energy, thereby quenching the clusters more efficiently. These simulations show that it is important to simulate thermal conduction in order to accurately model cluster physics.

  • image

    Anisotropic diffusion in AREPO

    Implementation of an extremum preserving anisotropic diffusion solver

    The numerical modeling of the anisotropic diffusion equation is problematic and non-trivial. Widely used discretization approaches violate the second law of thermodynamics i.e., heat can flow from lower to higher temperatures. This accentuates temperature extrema causing numerical instabilities which can trigger unphysical temperature oscillations.

    I implemented an extremum preserving anisotropic diffusion solver for the unstructured meshes of the moving mesh code AREPO (Kannan et al. 2016b). It relies on splitting the one sided facet fluxes into normal and oblique components. This is achieved by decomposing the gradient of temperature in the co-ordinate system defined by the cell center and its appropriate neighbors. The neighbors (2 in 2D and 3 in 3D) that form the new co-ordinate system are chosen such that the components of the vector in this system are all positive. The flux along the face normal will always be along the temperature gradient but the same cannot be said about the other oblique components. Therefore, the oblique flux is then non-linearly limited in such a way that the total flux is both locally conservative and also extremum preserving. The extremum preserving property of the scheme ensures that the second law of thermodynamics is not violated.

    The values of the relevant variables at the cell faces are extrapolated from the mesh generating points using a a very simple yet robust interpolation scheme that works well for strong heterogeneous and highly anisotropic problems. The required discretization stencil is essentially small, consisting of just the point and its Delaunay connections. The numerical diffusivity was shown to be as low as ~1%, even at really low resolution and decreases at almost second order with increase in resolution. This ensures that our scheme has negligible numerical diffusivity in all practical applications.

    We have also implemented a semi-implicit algorithm where, the non-linear flux terms that depend on the internal energy are integrated explicitly, while the other terms are integrated implicitly. This linear system is solved using HYPRE, which is a library for solving large, sparse linear systems of equations on massively parallel computers. This method is extremely fast as it requires only the solution of a linear system per timestep and is faster than the simple explicit scheme by a factor of 20.

  • image

    Cluster Cosmology

    Quantifying hydrostatic mass bias

    As the most massive gravitationally bound structures in the Universe, galaxy clusters represent an important tool for cosmological studies. However, this requires an accurate knowledge of their mass. The inference of cluster masses from X-ray and SZ measurements typically relies on the assumption of hydrostatic equilibrium. However, simulations have demonstrated that this assumption can be violated (by about ∼ 20%) by bulk motions in the gas or by non-thermal sources of pressure.

    However, recent results from Planck, have shown that the mass bias required to bring the cluster counts and CMB into full agreement is larger (about ∼ 40%). One exciting possibility if the effect of anisotropic thermal conduction in the outskirts of clusters. It has been theorized that Magneto-Thermal Instability in an anisotropically conducting plasma can tap into the radially decreasing temperature gradient to drive rigorous convection regardless of the background entropy gradient. This can increase turbulent pressure support in the outskirts of galaxies. Thus understanding the effect of this instability in the whole cluster mass range is extremely important and might account for the difference in the predicted and observed hydrostatic mass bias values.

    I am running cosmological simulations with ansiotropic thermal conduction that are ideally suited to explore how SZ- and X-ray scaling relations, and the mass bias of clusters come into existence and how they are affected by the detailed micro-physics of the intracluster medium and the exact model used for AGN feedback. Moreover, these simulations will help improve our understanding of the systematics involved in theoretical modeling and can inform current and future cosmology surveys such as Planck, SPT, ACT EUCLID, LSST and WFIRST.

  • image

    Radiation fields in galaxies

    Impact of local radiation fields on gas cooling rate in galaxies

    I performed a cosmological hydrodynamical simulation of a representative volume of the Universe, as part of the Making Galaxies in a Cosmological Context (MaGICC) project. This simulation testes the effectiveness of UV photoheating feedback (Stinson et al., 2012), in a variety of morphologies, masses and environments. The simulated sample compared well with a wide range of observed relations of galaxies like the stellar-halo mass relation, the galaxy stellar mass function, the observed number density evolution of low mass galaxies etc. (Kannan et al., 2014a). These results showed the importance of pre-supernova feedback, in regulating the properties of low mass galaxies.

    I also developed a novel method to include the impact of local radiation field from young and old stars and hot ICM on gas cooling rates in galaxies. This model was implemented on the fly in the hydro codes GASOLINE and AREPO, using simple approximations. I have shown that the hot, diffuse circum-galactic gas is adversely affected by the local radiation field reducing the cooling rate and raising the equilibrium temperature of the gas. This reduced the gas accretion rate onto the disk, which in turn reduced the star formation rates (Kannan et al., 2014b). I have, in collaboration with Joshua Suresh, recently proposed local radiation fields as a possible explanation for abundances of various ionization states of elements like OVI (Suresh et al., 2015) in the COS-Halos survey. Other recent works have also tried to explain these observations along the same lines (Werk et al., 2016). I have also shown that radiation fields from the hot ICM potentially plays an important role in quenching galaxies in cluster environments. They were shown to drastically reduce the gas cooling rates in satellite galaxies of clusters bringing the quenching timescales closer to the observational constraints (Kannan et al., 2016a).

    I am now improving upon the modeling by implementing an efficient and accurate radiative transfer algorithm. In the diffusion approximation, the radiative transfer equation can be written as a sum of anisotropic diffusion and advection terms. The diffusion part is solved with my ansiotropic thermal conduction algorithm and the advection part is solved using a standard advection solver. This method is extremely fast and efficient and has been seen to produce accurate shadows (a very difficult problem for previous diffusion based radiative transfer modules, eg. Petkova & Springel 2009; Sales et al. 2014).

  • image

    Mergers and morphology

    Impact of mergers on the morphology of galaxies

    I studied the effect of mergers on the morphology of galaxies by means of the simulated merger tree approach. This method combines N-body cosmological simulations and semi-analytic techniques to extract realistic initial conditions for galaxy mergers. I show that the satellite mass accretion is not as effective as previously thought, as there is substantial stellar stripping before the final merger. The fraction of stellar disc mass transferred to the bulge is quite low, even in the case of a major merger, mainly due to the dispersion of part of the stellar disc mass into the halo. The inclusion of the hot gas reservoir in the galaxy model contributes to reducing the efficiency of bulge formation. Overall, our findings suggest that mergers are not as efficient as previously thought in transforming discs into bulges (Kannan et al., 2015) and recent works have confirmed my findings (Sparre & Springel, 2016). In fact observations of the local Universe prefer low bulge formation efficiencies further giving credence to my result that mergers are inefficient in transforming disks to bulges (Fontanot et al., 2015).

  • image

    Measuring the spin of Sgr A*

    Impact of framedragging on the kinematics of galactic center stars

    During my undergraduate, I worked on the fascinating strong field prediction of general relativity, that the rotation of a massive object would distort the spacetime metric, making the orbit of a nearby test particle precess (Frame dragging effect). If this effect could be observed for Galactic-center (GC) stars, then in principle the spin of the Massive Black Hole at the center can be measured. A low-velocity perturbative expansion of the Kerr metric is done to obtain a simplified metric and then the resulting geodesic equations were numerically integrated in order to quantify this effect. The kinematic effect at pericenter passage due to frame dragging was calculated and was seen to be of the order of order 10 m/s for known galactic center stars (Kannan & Saha, 2009). If observed, this would provide an accurate, direct and a model independent measurement of the spin of the black hole.