Postdoc opportunity - work with me

I’d like to tell you about a new postdoc opportunity; measuring galaxy morphology in Euclid images using deep learning.

Key points

  • Starting Q4 2023 (flexible)
  • 2 year postdoc contract
  • Salary approx. €40k EUR per year (approx. €30k EUR after tax)
  • Working alongside ESA’s data scientists at ESAC in Madrid, Spain
  • Specific travel funding to visit NASA’s Space Telescope Science Institute (STSci) in Baltimore - and me in Toronto!
  • Application deadline July 31st

The core science question is whether galaxies in dense cluster environments have different morphologies to those in field environments. We’d like to measure the morphology of cluster and field galaxies, see how they differ, and interpret what physics might be driving this difference.

There’s two reasons that now is the time to do this.

  • Euclid is going to increase the number of z ~ 1 galaxies with resolved detailed features by orders of magnitude. We can develop the method on the Hubble archive and immediately apply it to Euclid. Plus, of course, JWST!
  • We can now accurately and automatically measure detailed morphology with Zoobot, deep learning models trained to predict what Galaxy Zoo volunteers would say. Read more here.

You would be working at ESAC side-by-side with ESA’s data scientists and using their new platform, ESA Datalabs, where researchers can attach Jupyter-like notebooks directly to ESA’s archives (inc. Hubble, JWST, and soon Euclid). We’re hoping this project will both showcase what’s possible with Datalabs and provide a template for other astronomers to apply deep learning models to Datalabs images.

The project relies on using deep learning models to measure morphology. We would be working closely together and I can help you get up-to-speed with the models we have so far. If you’re interested in pushing computer science boundaries with new model ideas, I’d love to collaborate on this. But you do not need to be a deep learning expert - this is an astronomy project first-and-foremost. The deep learning models are a (neat) tool to do new science. To help with that science, we’ve put together an all-star panel of international experts on cluster environments and on Hubble/JWST/Euclid.

ESA are funding this project via a contract with Centro de Astrobiologia (CAB), who in turn are hiring for this position and would be your formal employer. The CAB job advert is here

CAB are responsible for the hiring decision. My advice to them is to look for a candidate with:

  • A keen interest in galaxy morphology science. You should care about the answers!
  • Fluency in scientific Python (a.k.a. the pydata stack), demonstrated by previous projects
  • Some experience with machine learning is preferred; deep learning not required.
  • Comfortable collaborating in a team environment. Iconoclasts would miss out on the resources we’ve put in place to support you.
  • Spanish language skills are not required at ESAC, an international facility with members from every ESA country.

It goes without saying that CAB will welcome individuals from any background. One of my favourite things about being an astronomer is the diversity of the people I meet. And Madrid is lovely - I hope to visit you lots, especially during the Canadian winter!

CAB’s application system is in Spanish and not the most intuitive. To guide you through the process, begin applying by emailing J. Miguel Mas Hesse and cc. Jan Reerink (, They will then help you make your application.

Please drop me a message if you have any questions about the project. Good luck!