Azib Norazman

Research

Exocomets are the analogous term for comets in our Solar System - icy bodies that have perturbed inwards from Kuiper belt analogues (debris discs) by dynamical interaction within the planetary system. In recent years, there have been a number of stars that have shown evidence of exocometary activity due to the unique shape of their transit lightcurves. The most convincing proof of this is the young system Beta Pictoris (Zieba et al. 2019). Similar systems such as KIC 3542116 and KIC 11084727 (Rappaport et al. 2018) have also detected cometary activity in their lightcurves. The significance of exocomets allows us to further understand the morphology dynamics of extrasolar systems, and understand whether other planetary systems can be compared to our own.

An automated method to detect comet-like transits was developed in Kennedy et al. 2019, which used Kepler data in the main analysis. In my first project, I developed this method and applied the search towards the newer TESS data, as the mission provides a much larger, all-sky dataset, significantly reducing the bias of the star sample compared to previous surveys.

In my second project, I developed a new exocomet detection method using machine learning (Convolutional Neural Networks (CNN)). The motivation was that, given that we had very few detections from both the Kepler and TESS work, could other methods serve as a better, alternative technique to search for exocomets. CNNs are known for their pattern recognition capabilities, and so this appeared to be an ideal challenge to both detect exocomets but also separate them from false positives such as exoplanets or binary stars.

In my final project, I searched through the K2 data to complete the consensus of searching for exocomets in space-based photometric surveys. K2 serves as a uniquely suitable mission for exocomet science, given its Kepler-like precision and observing young star clusters such as Upper-Sco and Rho-Oph, and is a potential gold mine for exocomet detections.

My other research interests include the search for faint binary systems with spectroscopic data. This research comes from my MSc thesis on using PolyChord nested sampling towards the development of the Gaia DR4 pipeline with Dr George Seabroke and Dr Pablo Lemos.

Other Work

I have a keen interest in Earth Observation (EO). Before my PhD, I was an EO Data Scientist with the University of Leicester and Sylvera where I used machine learning to quantify aboveground biomass to monitor the performance of carbon offsets. The wider goal is to build confidence and transparency in the carbon market as we work towards Net Zero. I also held a part-time position in a start-up company specialising in cost-effective satellites to explore their next avenue of satellites in EO.

Publications

Roles

Affiliations

Conferences