Alice Townsend

I am Alice Townsend, a postdoctoral research fellow at the University of Birmingham. My research centres on explosive cosmic transients — in particular Type Ia supernovae (SNe Ia) — and how we can use them to better understand the Universe. I am especially interested in strongly gravitationally lensed supernovae and their potential as powerful tools for cosmology.

I completed my PhD at Humboldt University of Berlin, Germany, in February 2026. My doctoral research used data from the Zwicky Transient Facility (ZTF) to identify and characterise strongly lensed SNe Ia in archival data. I also developed a photometric classification framework for supernovae detected by ZTF.

Prior to my PhD, I obtained an MPhys in 'Physics with Astrophysics' from the University of Manchester in 2021, where my master's thesis focused on transitional millisecond pulsars.

Strongly lensed supernovae

Strong gravitational lensing is a consequence of general relativity, where light from a bright background source is deflected and magnified by an intervening massive object. Strong lensed supernovae, in particlar Type Ia supernove, allow us to calculate cosmological parameters, distributions of dark matter, as well as the astrophysics of the supernovae and the intervening lensing galaxies themselves. My work involves searching for lensed supernovae in archival and live data from ground-based telescopes such as ZTF.

Photometric classification of supernoave with ML

We are in an era of astronomy where we have an unprecedented amount of photometric data (i.e. light curves) of transients, but not enough spectroscopic resources to follow up and classify every object. In this case, photometric classification with machine learning is neccessary to type each supernova we detect. This is important for improving statistics in Type Ia cosmology, as well as for rate studies of each class. My research is applying photometric classification methods to live and archival ZTF extragalactic transients.

SN Ia cosmology

Type Ia supernovae are bright, standardisable candles that play a key role in constructing the cosmological distance ladder. One exciting application of my photometric classification work is the upcoming ability to classify the entire ZTF extragalactic transient sample photometrically. This will yield over 50,000 Type Ia supernovae — an unprecedented dataset for cosmological studies. With it, we will be able to measure fundamental parameters such as the Hubble constant, the matter density of the Universe (Ω_M), and the growth of structure in the early Universe (σ₈).

Photometric classification of supernovae detected by the Zwicky Transient Facility using noise augmentation (2026)

We developed a photometric classification framework tailored to ZTF supernovae. By training on real spectroscopic data and mitigating selection bias through augmentation, we achieved an average SN Ia recall of 98.1%, and this remains above 96% for events as faint as an apparent magnitude of 20.5.

Candidate strongly lensed type Ia supernovae in the Zwicky Transient Facility archive (2025)

We conducted a systematic search through three and a half years of the ZTF archive and found two promising lensed Type Ia supernova candidates. This required an efficient analysis pipeline that removed possible contaminants (such as SLSNe), whilst retaining objects that displayed the characteristics of lensed supernovae.