PI: Florian Beutler

In 2024, we launched a project focused on advancing the data analysis of the Dark Energy Spectroscopic Instrument (DESI), with the goal of incorporating higher-order clustering statistics into the core DESI cosmological analysis. DESI is collecting the distances of about 40 million galaxies. The 3D clustering of these galaxies contains a wealth of information on key cosmological parameters, like e.g. the amount of dark matter and dark energy.

An essential first milestone was the implementation and validation of the window function, which is critical for accurately interpreting the clustering signal in a survey with complex geometry. This work, led by Mike Wang (PDRA, University of Edinburgh), culminated in a peer-reviewed publication.

We are now addressing the key observational systematics in the DESI data, with a particular focus on fibre collisions, which is an instrumental effect that impacts small-scale clustering measurements. Initial progress on this was made through an MPhys project led by Aaditya Sarma (student, University of Edinburgh), who will continue developing this work over the summer. This effort is expected to result in a dedicated publication later this year. Looking ahead, the overarching goal is to take a leading role in the analysis of the DESI Year 3 dataset, scheduled for 2026. The DESI dataset represents a tenfold increase in volume over previous spectroscopic surveys and offers an unprecedented opportunity for precision cosmology. Caroline Guandalin, Mike Wang, and Samuel Brieden (all PDRAs based in Edinburgh) are currently developing multiple analysis components and authoring papers in preparation for this flagship analysis. The resources provided through this supercomputing grant have been essential to enabling this work, especially in the testing and validation of analysis tools at the scale required for DESI.

Figure: This figure shows the likelihood distributions for key cosmological parameters using two DESI samples, the high redshift Quasar (QSO) sample and the low redshift Luminous Red Galaxy (LRG) sample. This was the first demonstration of our analysis pipeline in early 2025.