Home   News   Research Teaching   Popular Science   Personal

Cosmological All-wavelength Radiative Transfer

I study the formation, evolution, and multi-wavelength properties of galaxies and quasars from cosmic dawn to present day, by combining multi-scale cosmological simulations with multi-wavelength radiative transfer, which I have developed over the past few years. The combination of these two powerful tools has a wide range of applications for various topics.

In 2006, I have developed a physically motivated method, MIM (Merger-tree In Multi-scale), for galaxy formation and evolution (Li et al. 2007, ApJ 665, 187). This technique employs a set of multi-scale simulations, which includs large-scale cosmological N-body simulations to identify halos or regions of interest, then zoom in onto the region with higher resolution in order to extract the merging history, and then re-simulate the merger tree again on galactic scale hydrodynamically with gas, star formation, and black hole accretion. MIM method ensures galaxy formation in a large-scale cosmological context, and resolves small-scale physics with high resolutions.

In 2007, I have developed a 3-D Monte Carlo radiative transfer code, ART2 -- All-wavelength Radiative Transfer with Adaptive Refinement Tree (Li et al. 2008, ApJ 678, 41), to calculate the multi-wavelength properties of galaxies and quasars. ART2 incorporates the radiative equilibrium algorithm developed by Bjorkman & Wood (2001), which treats dust emission self-consistently; an adaptive grid similar to Jonsson (2006), which efficiently covers a large dynamic range in spatial scale and captures inhomogeneous density distributions (especially in galaxy mergers); a multi-phase ISM model which accounts for the observed scaling relations in molecular clouds; and a super-novae origin dust model which can explain the existence of dust in cosmologically young objects. ART2 self-consistently and efficiently produces SEDs, images, and properties in a full spectrum from x-tay to millimeter.

These essential developments pave the way for studying a wide range of exciting topics. By combining cosmological simulations with radiative transfer, we can follow structure formation dynamically and photometrically, from the first galaxies and quasars in the epoch of reionization (z >6), to the epoch of peak growth (z~2-1), and to the present day.

Galaxies & quasars at redshifts z ≥ 6, an exciting frontier in observational cosmology

The discovery of luminous quasars and starburst galaxies at z~6 indicate the presence of supermassive black holes of ~109 Msun, supermassive galaxies of ~1012 Msun, and ultra-luminous infrared galaxies with LFIR ≥ 1012 Lsun when the Universe was less than one billion years old. These frontier observations challenge theoretical models to explain the formation and evolution of these massive objects in the early Universe, and their multi-wavelength properties.

Using the MIM method, together with a self-regulated black hole growth model developed by Di Matteo, Springel & Hernquist (2005), I produced a luminous quasar at z~6.5, whose properties resemble those of the most distant Sloan quasar SDSS J1148+5251 detected at z=6.42. (Li et al. 2007, ApJ, 665, 187). We find that luminous quasars at z~6 can form in massive halos in the LCDM cosmology, and they grow through hierarchical mergers of gas-rich galaxy progenitors.

Furthermore, I applied ART2 to the above hydrodynamic simulations, and reproduced the dust properties, and the multi-wavelength SED and other properties of SDSS J1148+5251 (Li et al., 2008, ApJ 678, 41, astro-ph/0706.3706)

These are the first efforts to model the formation and evolution of quasars and galaxies at z ≥ 6, and their multi-wavelength properties. By combining large-scale cosmological simulations with ART2, we'll be able to unveil the nature of the galaxies and quasars in the epoch of reionization, and their contributions to reionization.

model quasar

A brief formation history of a luminous quasar at z~6.5, formed through hierarchical galaxy mergers following the halo merger tree derived from cosmological simulations. The images are projected stellar density, color coded by specific star formation rate (blue indicates strong SF, red indicates little or no SF). This system evolves from a starburst (z~14 - 7.5) to a quasar (z~7.5 - 6), and to a "red and dead" elliptical at z~5, and it transforms from a cold to a warm ULIRG. During the peak quasar phase (z~6.5), it has a number of properties (BH mass, bolometric luminosity...) similar to the most distant Sloan quasar SDSS J1148+5251. (Credit: Li et al. 2007, ApJ 665, 187)

Galaxies and quasars at lower redshifts

At intermediate redshifts epcoh of peak growth, most star formation takes place in dusty, heavily obscured galaxies. A large population of infrared luminous submillimeter galaxies (SMGs) has been unveiled recently, which may account for 50% of the cosmic SFR at z~2, as well as a large population of LIRGs or ULIRGs. These galaxies appear to be massive starburst systems with SFR~101-3 Msun/year, Mstar~1012 Msun, and large CO reservoirs. Many show disturbed morphologies or multiple components indicating gas-rich mergers. Moreover, AGNs are detected in some of these galaxies. ART2 is well suited to study these populations. When applied to a large suite of galaxy simulations, we produce a wide range of multi-band properties as observed, and gain insights on the nature of these objects, and their contributions to the luminosity functions (Li et al, in preparation).

Nature of dark matter and dark energy

Probe the nature of dark energy using the Baryonic Acoustic Oscillations in large-scale cosmological simulations, and study dark matter annihilation by looking for gamma-ray emission from the Milky Way-like galaxies.

PhD Thesis

I obtained my Ph.D. from the Department of Astronomy at Columbia University in 2005. For my thesis under supervision of Dr. Mordecai-Mark Mac Low (AMNH), I studied star formation in different environments, from molecular clouds to isolated galaxies, and to galaxy mergers.