Enabling Multi Messenger Astronomy with a low-latency LISA data pipeline

Project Information

EMILIA

Grant agreement ID: 101065596

DOI

10.3030/101065596

Project closed

EC signature date 4 July 2022

Start date 1 January 2023

End date 31 December 2024

Funded under

• Marie Skłodowska-Curie Actions (MSCA)

Total cost

No data

EU contribution

€ 169 326,72

Coordinated by

ARISTOTELIO PANEPISTIMIO THESSALONIKIS
Greece

Periodic Reporting for period 1 - EMILIA (Enabling Multi Messenger Astronomy with a low-latency LISA data pipeline)

Reporting period: 2023-01-01 to 2024-12-31

Summary of the context and overall objectives of the project

In recent years, following the first detection of Gravitational Waves (GWs), we have witnessed the birth of GW Astronomy. So far, there have been a plethora of events recorded, providing us with invaluable information about the very nature of the merging binaries. An exceptional case is the event GW170817, a Neutron Star merger, which was observed with both gravitational and electromagnetic (EM) waves. From a single event alone, by combining both ways of observation, we were able to vastly improve our understanding of such cataclysmic events. In the near future, in particular, in the early 2030s, the ESA Laser Interferometer Space Antenna (LISA) is going to be launched. LISA is a space-borne Gravitational-Wave observatory that, in contrast to the present ground-based detectors, is going to be signal-dominated. The LISA data will give us the unique opportunity to observe different types of GW events, from the merger of supermassive black hole binary systems to Double White Dwarfs. Some of those signals could be potentially combined with the EM observations emitted by the same event. This complementarity of information will enable us to push our knowledge boundaries in astronomy, astrophysics, and cosmology. With EMILIA, we have built a framework to enable multi-messenger astronomy with LISA, by developing a suit of data analysis tools based on novel techniques.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

LISA is expected to measure a huge number of overlapping signals simultaneously, which means that we need to design an analysis strategy that will recover all intertwined and unknown types of waveforms from the data. This, rather complicated scheme, is referred to as a Global Fit. Our main objectives with EMILIA were to build a data analysis pipeline following a Global Fit scheme, that approaches the analysis of LISA data by combining “traditional analysis methods” and Machine Learning techniques. We have been focusing on Searching for individual transient signals from different types of sources, with the aim of characterizing the residual confusion foreground signal plus instrumental noise, and producing fast solutions for chirping and quasi-monochromatic signals. For reaching our two main objectives, we have created a data analysis pipeline that utilizes the Snakes On A Plane (SOAP) algorithm in order to recover weak continuous GW signals buried in the noise. We have found that this algorithm is more effective when analyzing residual data, e.g. on data where the brightest signals have been already found and subtracted. However, we have discovered that even residual data might have non-gaussian statistical properties, which challenges any algorithm that searches for weak continuous signals. This unexpected but impactful discovery was studied in detail, and a robust statistical framework tailored for searching for GWs in data with outliers was designed. Finally, we have developed a standalone Global Fit pipeline based on the combination of “traditional” Bayesian techniques, and Machine Learning methods. This complementarity of methods, together with hardware acceleration based on Graphical Processing Units, allowed us to greatly improve the efficiency of our Bayesian Global Fit scheme.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

Our LISA data analysis pipeline is capable of performing full analyses on the data and producing astrophysical catalogues in a fraction of computational time compared to existing state-of-the-art methods. It is designed to be modular and it's open source. This means that any member of the scientific community can use and modify the software. This allows for easier prototyping and thus enabling us to unlock the full scientific potential of the future LISA mission. At the same time, it promotes inclusion by giving the opportunity of isolated members of our scientific community (geographically or socially) to join the effort. Finally, we have developed statistical methods that are robust to data outliers. The introduced techniques will be useful not only in the field of Gravitational Wave Astronomy, but in other fields as well.