EPSRC logo

Details of Grant 

EPSRC Reference: EP/W005654/2
Title: Novel Non-linear Techniques for Cosmic Large Scale Structure
Principal Investigator: Bose, Dr B
Other Investigators:
Researcher Co-Investigators:
Project Partners:
Department: Sch of Physics and Astronomy
Organisation: University of Edinburgh
Scheme: EPSRC Fellowship
Starts: 01 April 2022 Ends: 16 February 2026 Value (£): 424,963
EPSRC Research Topic Classifications:
Cosmology Extra-Galactic Astron.&Cosmol.
EPSRC Industrial Sector Classifications:
Related Grants:
Panel History:
Panel DatePanel NameOutcome
27 Jul 2021 Stephen Hawking Fellowship - R2 Interviews Announced
Summary on Grant Application Form
Our current understanding of the Cosmos is far from perfect. It can explain a large number of observations and experimental results, but to do so it needs to invoke a form of energy that is both non-luminous and ubiquitous. So ubiquitous in fact that if the energy of the Universe were a football team, we'd only be able to see the goalkeeper. This is obviously a (big) problem. So let's assess the standard model. At its core, the model is a combination of three assumptions:

1) The Universe is roughly the same wherever you go and wherever you look.

2) Space is expanding.

3) Einstein's theory of gravity is universal, i.e. it applies everywhere, and the majority of the dark energy is constant in space and time.

The first assumption is well confirmed by what telescopes see in the sky - everywhere we look things are roughly the same. And everywhere we go ... well we can't go everywhere, but there are strong philosophical arguments to support this assumption!

The second assumption is also very well confirmed by observations of distant galaxies, which are all moving away from us at a speed that is proportional to their distance from us.

The jury is still out on the third assumption. This is where this project comes in. My goal to develop a theoretical model which can map fundamental theories of gravity and cosmology to the positions, shapes, sizes and velocities of the galaxies we see in the sky. By doing this, we can test assumption (3) and hopefully gain clues about the true Nature of the Universe and/or its dark side. In particular, I aim to produce predictions for how galaxy shapes and sizes change when their light passes through the Universe to our telescopes and also predictions for how galaxies congregate together under gravity. These two phenomena will be observed under a metaphorical microscope in the next 5-10 years by very powerful space and ground based telescopes, leaving cosmological clues nowhere to hide.

To make the best use of these observations, the map between the theory of gravity & dark energy and real predictions for galaxy observations needs to accommodate a large range of physical scales, over a trillion trillion metres, and it must do that accurately. To boot, this map needs to be general, in the sense, it must be able to bridge this gap for strong alternatives to assumption (3). To boot again, this map needs to be very very quick to traverse - to properly test any alternatives statistically, hundreds of thousands of predictions need to be made in a reasonable time, roughly a few days, on a super computer.

To summarise, we need both a comprehensive map between fundamental theory and predictions, and a very fast car to travel the road with. The map is something I have been part of developing and will complete during the project. The metaphorical car, in this case a computer code, is something I've led development of, but it is still a prototype and requires a lot of work to take us all the way to what we see in the sky.

Finally, I am also proposing a new means of testing assumption (3). Neural networks are used extensively nowadays in applications such as recognising human faces in photos. This involves showing the computer a large number of photos with and without faces, and over this training it 'learns' the characteristics of a face. What I propose here is training a neural network on 'photos' of the galaxy distribution of universes with and without assumption (3) included, so that it can learn what the 'face' of assumption (3) looks like. These training universes can be simulated with the map and code I will develop. When presented with the real Universe, our neural network can then tell us if it 'sees' assumption (3) present or not, and in doing so act as a compass in our navigation towards the fundamental Nature of the Universe.

The map, the code and the network will reach completion just in time for the largest galaxy survey to date - the Euclid mission.

Key Findings
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
Potential use in non-academic contexts
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
Impacts
Description This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
Summary
Date Materialised
Sectors submitted by the Researcher
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
Project URL:  
Further Information:  
Organisation Website: http://www.ed.ac.uk