I mostly work at the intersection between theory and observations.
I try to bridge models with data by interpreting observations
in the context of theoretical predictions, using analytical as well as
numerical methods.
I am interested in a variety of topics in astrophysics.
The main project of my Ph.D. revolves around understanding
high-redshift quasars
using large-volume cosmological simulations.
Specifically, I am building a theoretical framework to
interpret observations of quasars and galaxies across cosmic time,
with the aim of answering fundamental questions such as: Where does
quasar activity takes place? How do the properties of quasars relate
to the ones of their host halos/galaxies? What is the timescale of
quasar activity and how does this relate to the growth of
supermassive black holes?
In this context, I first studied the implications of the
demographic and clustering properties of quasars at \(z\approx2-4\).
I then shifted my focus to JWST observations of
the quasar-galaxy
cross-correlation function at \(z\approx6-7\), and proposed a model
to jointly constrain the properties of quasars and galaxies.
Currently, I am focusing on the role played by quasar obscuration
across cosmic time, and I am working on extending quasar population models
to incorporate single quasar lightcurves.
I am also interested in the formation and evolution of galaxies.
For instance,
I studied models of
galactic outflows in the context
of observations of
extended [CII] halos in high redshift galaxies.
In the past, I also worked on parameter inference for
overlapping gravitational wave signals
and on using radiative transfer simulations to gauge
the morphology of protoplanetary discs.
Studying the extreme clustering of high-z quasars with large-volume N-body simulations
Quasar clustering measurements provide a powerful probe of the
physical processes that govern the growth of supermassive black
holes and the coevolution with their host galaxies.
Observations from wide-field quasar
surveys indicate that the quasar auto-correlation length increases
dramatically from z~2 to z~4. The clustering
amplitude measured at z~4 is so extreme that it has been difficult to
reproduce theoretically, even when assuming
that quasars are hosted by the most massive dark matter
halos residing in the most extreme environments at that redshift.
As the first project of my Ph.D., I focused on this decade-long
problem and revisited it thanks to the new, large-volume, N-body
cosmological
simulations that have been developed in recent years by multiple
teams. I created a model that
simultaneously reproduces both the observed quasar clustering
and quasar demographics. The model builds on a novel method
to compute the halo mass and halo cross-correlation functions by
combining multiple large-volume dark-matter-only cosmological
simulations with different box sizes and resolutions. Armed with
these halo properties, the model exploits the conditional luminosity
function framework to describe the stochastic relationship between
quasar luminosity, L, and halo mass, M, and it is able to quickly
predict observables such as the quasar auto-correlation function
and the quasar luminosity function as well as other relevant quantities
such as the quasar luminosity-halo mass relation, the
quasar host mass function, and the quasar duty cycle.
We applied this model to observations of the quasar clustering and
demographics at z~2 and z~4, and found that we can reproduce all
the data very well. The very large clustering amplitude at z~4 implies
that the relation between quasar luminosity and halo mass is very steep
and has very little intrinsic scatter. This has profound
consequences in terms of accretion physics and the growth of supermassive
black holes at high redshift.
(figure from Pizzati et al. 2024a)
A unified model for the clustering of quasars and galaxies at z>6
Since its launch at the end of 2021, the James Webb Space Telescope (JWST)
is rapidly revolutionizing our understanding
of the high-redshift Universe. One of the most powerful JWST instrument
for studying the early Universe is the NIRCam Wide Field Slitless Spectroscopy mode,
which allows for the
spectroscopic identification of high-redshift quasars and galaxies
over large comoving volumes. Exploiting NIRCam/WFSS observations
of several bright high-z quasars fields, the EIGER survey has identified
many [OIII]-emitting galaxies that are physically associated
with the targeted quasars, revealing a variety of large-scale
enviroments around the quasars, included one of the most spectacular
overdensities revealed in the Epoch of Reionization. By putting together
different quasar fields, in Eilers et al. 2024 we have measured
for the first time the quasar-galaxy cross-correlation function at z~6,
together with the galaxy auto-correlation function at the same redshift.
These measurements hold the key to understanding the properties of quasars
and galaxies at high redshift and their relation with the host dark matter halos.
However, the interpretation of these data is not straightforward, and
requires the development of a theoretical framework that can simultaneously
reproduce the observed properties of quasars and galaxies at z~6.
By building upon the empirical quasar population model described above
(see Pizzati et al. 2024a), we have worked on developing such
framework. In this context, we made use of a new,
large-volume N-body simulation with more than
a trillion particles, FLAMINGO-10k, that is able to model quasars and galaxies
simultaneously. We successfully reproduced observations
of z~6 quasars and galaxies (i.e., their clustering properties and
luminosity functions), and used them to infer key quantities such as their
luminosity-halo mass relation, the mass function of their host halos,
and their duty cycle/occupation fraction. By focusing on the inferred
properties of quasars, we presented for the first time a homogeneous analysis of
their evolution up to z>6. The picture that emerges reveals
a strong evolution of the host halo mass and duty cycle of quasars
at z~2-6, and calls for new investigations of the role of
quasar activity across cosmic time
(figure from Pizzati et al. 2024b)
Massive black holes and quasars in the FLAMINGO simulation
Cosmological hydrodynamical simulations follow the non-linear
evolution of structures in the Universe by modeling a large
variety of processes that are important to the physics of galaxies,
stars, and black holes. In the last decade, these simulations have
become capable of reproducing a large number of observed galaxy
properties. As a consequence, they have started to play a key role
in shaping our understanding of the formation and evolution of
galaxies in the universe, as well as the relation between galaxies
and the supermassive black holes (SMBHs) that are harbored at their
center. While much effort has been devoted to the study of the
connection between SMBHs and their hosting galaxies in terms of
gas fueling and AGN feedback, only few studies have focused on
the global properties of SMBHs and their evolution over cosmic
time. Specifically, SMBHs are visible at cosmological distances
when they turn into bright quasars. Quasars are, however, extremely
rare objects, and hence their statistical properties (such as their
luminosity and spatial distributions) can only be studied
theoretically when large simulated volumes are available.
In a project that I am co-supervising, we use the cosmological simulations from
the FLAMINGO suite (which are the largest ever run) to
reproduce two basic observational probes of quasar activity:
the quasar luminosity function and the quasar autocorrelation
function. We study how well FLAMINGO can match observational
constraints on these quantities at different redshifts. We also
investigate which black holes contribute to the bright quasar
population, and split their relative contribution in terms of
black hole mass, host halo mass, and Eddington ratio. We are also
interested in the relation between massive black holes and galaxies
in FLAMINGO across cosmic time. Thanks to the state-of-the-art
capabilities of FLAMINGO, for the first time we can probe
the ability of large-volume hydrodynamical
simulations to capture the evolution of massive black holes and
quasar activity across cosmic time
(figure from Schaye et al. 2023).
Modeling outflows and [CII] halos in high-redshift galaxies
Investigating the complex environments of galaxies during the Epoch of Reionization
is one of the most pressing research goals of modern astrophysics.
As shown by cosmological simulations, galaxies are already at these epochs,
and they present different properties with respect to the ones
in the local Universe. Unraveling how galaxies formed and evolved during these
remote epochs (i.e., a few hundred million years after the Big Bang)
is at the heart of our understanding of cosmic evolution.
In this context, the advent of telescopes such as ALMA and NOEMA have opened a new
window on the primordial Universe, allowing us to shed light on
the obscured star formation and ISM line emission at rest-frame FIR
wavelengths up to z~7.
One of the most compelling findings made by ALMA is that a significant
fraction of z>4 galaxies is surrounded by extended (10-15 kpc)
[C II]-emitting haloes that are not predicted by even the most
advanced zoom-in simulations.
This discovery poses a series of thorny theoretical questions,
involving their formation mechanisms, evolution, and impact on the
enshrouded galaxies as well as on the external regions where
the intergalactic medium (IGM) resides.
As part of my Master Thesis, I worked with Andrea Ferrara, Andrea Pallottini,
and the cosmology group at SNS to take on the problem of a finding plausible
mechanism to explain the formation of these halos. In particular,
we focused on the hypothesis that these halos result from the remnants of past –
or ongoing – outflow activity. We explored this idea by using a semi-analytical model for
an outflow that undergoes catastrophic cooling in the inner region of the halo. Computing
the abundance of singly ionized carbon and simulating the resulting [C II] emission, we
compared it directly with observational evidence coming from recent ALMA data,
and we conclude that outflows represent a promising answer to explain the origin of the
observed [C II] halos. In a series of two papers, we analyzed
the properties of [CII] halos in stacked and individual observations
(from the ALMA ALPINE survey) and showed that the outflows that are needed
to generate such halos have mass loading factors in the range ~4-7,
and scale with stellar mass in a way that is
consistent with the momentum-driven hypothesis. Our model
points to the presence of extended [CII] emission at high redshifts
as a tangible sign of star-formation-driven feedback mechanisms
being already in place well into the Epoch of Reionization.
Turbulence and morphology of protoplanetary discs
Constraining the strength of gas turbulence in protoplanetary discs is an open problem
that has very relevant implications for the physics of gas accretion and planet formation.
A promising method to gauge the level of gas turbulence in discs is to measure the vertical
scale height of the dust component - which is expected to be coupled to the gas component
through the gas-dust coupling. This has become possible in the last few years
thanks to the very high-resolution observations provided by the ALMA telescope.
These observations uncovered a large amount of features in the 2d images
of protoplanetary discs, including dark gaps and emission rings.
As shown by Pinte et al. 2016,
it is possible to exploit these features to uncover the 3-d morphology of discs.
The idea is very simple: due to projection effects, a gap in a disc emission
profiles will be partly filled by the emission coming from the
neighbouring regions. This effect is stronger along the minor axis
of the disc, whereas the major axis is only marginally affected.
For this reason, one can compare the gap contrast along the major and
minor axes to infer the amount of "filling" that has taken place.
The extent of this gap filling is directly dependent on the vertical structure of the
disc. Hence, a measurement of the gap contrast
along the major and minor axes can be used to infer the vertical scale
height of the disc, and to indirectly gauge the amount of gas turbulence in discs.
In a work done with Giovanni Rosotti
and Benoit Tabone, we applied this method to the sample of discs
observed by the DSHARP collaboration, for which high resolution images
are available. We built a radiative transfer model to reproduce the
observed gap constrast for different values of the dust scale heights.
and found that the scale heights that yield a better agreement with data are
generally low, implying
low levels of gas turbulence in discs. This represents an important
step towards understanding the physical processes that govern the
formation of planets in discs.
We also investigated the properties of the sources
for which our method yields no significant constraints on the disc
vertical structure, and concluded
that this is because these discs have either a low inclination or gaps
that are not deep enough. Based on our analysis, we provided an empirical
criterion to assess whether a given disc is suitable to measure the
vertical scale height (figure from Pizzati et al. 2023).
Overlapping gravitational waves signals in the next generation of detectors
With the first direct detection of gravitational waves in 2015,
the era of gravitational-wave astronomy has begun. This first decade
of observations has already provided a wealth of information about
compact binary mergers, neutron stars, and black holes. The future
of this research field looks even brighter, with the next generation of detectors
like the Cosmic Explorer and the Einstein Telescope expected to
observe hundreds of thousands of binary coalescence events each year.
With huge leaps in sensitivity, however, new challenges arise.
One of the most pressing issues for the next generation of detectors
is that signals from coalescing binaries will be so frequent that
they will start to overlap with each other.
When two signals overlap, the standard data-analysis pipelines
used to detect and infer the properties of these signals
are not guaranteed to work. While the detection
pipelines have been shown to work well even in the presence of
overlapping signals, the parameter inference algorithms used to
estimate the properties of these signals represent a more compelling
problem, as overlap of multiple signals in the time/frequency
domains may result in biases in the final parameter estimates.
In a work done with Bangalore Sathyaprakash, Surabhi
Sachdev, Anuradha Gupta, and the LIGO group at Penn State,
we were among the first to quantify the biases arising by using
the current parameter inference pipelines to constrain the
parameters of a compact binary coalescence (CBC) signal
in the presence of multiple overlapping ones. We showed that, since
the current detection pipelines provide the coalescence time of
each signal with an accuracy of ~10 ms, one can set a prior
on the coalescence time exploiting this information. In this way,
it is possible to correctly infer the properties of
multiple overlapping signals even
with the current data-analysis infrastructure. By studying different
configurations of overlapping signals, we find that the parameter
inference is robust provided
that the coalescence times of the signals in the detector frame
are more than ~1-2 seconds apart. This implies that for the vast
majority of the signals that will be detected by future detectors,
our current framework for parameter inference will work just fine.
Signals whose coalescence epochs lie
within ~0.5 seconds of each other, however, suffer from significant
biases in parameter inference, and new strategies and algorithms
will be required in the future to overcome such biases
(figure from Pizzati et al. 2022).