Research

From spectra to atmospheres

My research turns starlight into chemistry. I develop the software and methods that let us characterize exoplanet atmospheres across resolutions and instruments, and I use them on real data, all the way from the telescope to the posterior distribution.

High-resolution spectroscopy

Resolving individual molecular lines (and atomic lines in the visible) to detect species via cross-correlation, using high-resolution spectrographs such as GIANO-B at the Telescopio Nazionale Galileo (TNG), HARPS-N, ESPRESSO, IGRINS and, soon, ANDES at the Extremely Large Telescope (ELT).

Low-resolution spectroscopy

Capturing the broad shape of the spectrum with space telescopes (e.g. the James Webb Space Telescope, JWST, and the Hubble Space Telescope, HST; and, in the future, the Ariel mission) to constrain overall abundances and clouds.

Bayesian retrievals

Differential-evolution MCMC over large model grids, at both low and high resolution, to infer temperatures, pressures and abundances with related uncertainties.

Multi-resolution synergy

Combining low-resolution space data (e.g. JWST, HST; and, in the future, Ariel) with ground-based high-resolution spectra to break parameter degeneracies.

GUIBRUSHR®

Graphic User Interface for Bayesian Retrieval Using Spectroscopy at High Resolution

GUIBRUSHR is the flagship pipeline I develop with the exoplanet group at INAF–OATo. Its scientific aim is to measure what exoplanet atmospheres are made of, and to connect that composition to how planets form and evolve. To get there it brings together, in one tool, analysis steps that are usually scattered across separate codes, and it is built to make a normally complex, error-prone process straightforward: an intuitive interface replaces long configuration files, and streamlined data management handles large astronomical datasets.

I led the rewrite of the original IDL codebase into Python, adding CPU parallelization that sharply cut execution times, and I'm now extending it with GPU parallelization of the radiative-transfer step. The software generates transmission and emission models for clear and cloudy atmospheres using radiative-transfer engines such as petitRADTRANS and PyratBay, and explores parameter space with differential-evolution MCMC retrievals and real-time visualization.

GitLab repository Data Reduction tool

Key modules. GUIBRUSHR is organized into modules, each handling one step on the path from raw starlight to a planet's atmospheric composition, tied together by a local database that keeps every target, observing night and result reproducible.

Reading the spectra

Data Reduction: turns raw high-resolution spectra into calibrated, wavelength-aligned data ready for analysis. Now also a standalone tool supporting GIANO-B, HARPS-N, CRIRES+ and ESPRESSO.
Telluric Removal: Earth's atmosphere and the host star imprint strong lines that bury the planet's far weaker signal. This module removes them order by order with a PCA-based approach, while preserving the slightly Doppler-shifted planetary lines.
Cross-Correlation: co-adds the thousands of faint molecular lines into a single, statistically significant detection, confirming which species are present from the planet's orbital motion.

Modelling the atmosphere

Forward Model: builds physical transmission and emission spectra for the expected molecules, the temperature structure and clouds (with petitRADTRANS or PyratBay), used both as cross-correlation templates and as the models compared with the data.
Bayesian Retrieval: a differential-evolution MCMC infers the atmosphere's molecular abundances, temperature structure, clouds and velocities, with proper uncertainties, running on high-resolution, low-resolution or both jointly.
Synthetic data & database: generates realistic mock observations to validate the full chain, while a local SQLite3 database stores targets, observing nights and results for reproducible science.

GUIBRUSHR has already contributed to peer-reviewed work, including Basilicata et al. (2025).

pyExoRaMa

An interactive Python tool for the radius–mass diagram of exoplanets

pyExoRaMa is a Python adaptation of software originally written in Mathematica by Dr. Li Zeng. It lets you visualize and analyze exoplanet and host-star data across many parameters at once. As the exoplanet catalog keeps growing, it supports statistical studies that reveal parameter interdependencies and compare observations against theoretical models.

GitHub Zenodo ASCL RNAAS

Atmospheres I'm characterizing

Putting the methods to work on real planets, all the way from data reduction to the posterior distribution.

WASP-107b paper in prep

The warm super-Neptune WASP-107b is my main science case. I combine ground-based high-resolution spectroscopy with low-resolution data from space telescopes in a single joint retrieval with GUIBRUSHR, constraining its atmospheric composition and thermal structure far more tightly than either dataset could alone. This was the subject of my MSc thesis and is now being written up for publication, with implications for what future ANDES@ELT observations of smaller planets could achieve.

WASP-77Ab paper in prep

WASP-77Ab is a benchmark hot Jupiter with well-characterized high- and low-resolution observations, which makes it the ideal target for validating GUIBRUSHR. I use its combined datasets to confirm that the pipeline recovers the expected abundances and temperature profile, demonstrating that the joint high- plus low-resolution approach is robust before turning it on less-explored worlds.

Observations

I'm a primary observer for the GAPS programme (Global Architecture of Planetary Systems) at the Telescopio Nazionale Galileo in La Palma, working with both the HARPS-N and GIANO-B spectrographs, the same high-resolution data that feed my atmospheric models. I also observe at the Asiago Astrophysical Observatory with the University of Padua.

Earlier work: solar physics & space instrumentation

Before specializing in exoplanets, I built software for ESA and INAF space missions.

For several years I developed and validated flight and ground software for solar coronagraphy and space-weather missions. The work ranged from formation-flying metrology and instrument command testing to balloon-borne coronagraphy and operational forecasting of coronal mass ejections, and it gave me the mission-grade software practices I now bring to exoplanet pipelines.

PROBA-3 / ASPIICS

End-to-end MATLAB simulator for the Shadow Position Sensor (SPS) metrology subsystem; later managed the in-flight SPS software.

Metis / Solar Orbiter

Python GUI converting PDOR/IOR command sequences into TCL files to test the METIS coronagraph on the ground.

CorMag / HEMERA

Pointing, thermal and communication software for a stratospheric-balloon coronagraph, supported on launch campaigns in Timmins, Canada.

Space weather (SWESNET)

Automatic detection of CMEs from coronagraph images and drag-based prediction of their arrival time at Earth.

See the full CV & publications →