Leveraging Artificial intelligence for Astrophysical Source parameter Estimation and Rapid electromagnetic follow-up with the future space-based Gravitational Waves Observatory

The Laser Interferometer Space Antenna (LISA), a milli-Hertz gravitational wave (GW) observatory, has extraordinary scientific potential. It is expected to revolutionize multimessenger astrophysics in the mid-2030s through synergies between LISA and both space- and Earth-based electromagnetic (EM) observatories. Mergers of supermassive black hole binary systems will be primary sources for these multimessenger studies. LISA will detect GWs emitted during the pre-merger phase and pinpoint the source location, allowing EM observatories with narrow fields of view to search for potential counterparts. For this synergy to be effective, rapid alerts of sky positions and coalescence timing are crucial—not only at the moment of merger but also in the hours leading up to it, when precursor signals may occur. This raises a vital question: what strategies or technologies will ensure that EM observatories are prepared to respond in nearly real time?My project addresses this urgent need by developing a low-latency alert pipeline (LLAP). Unlike detailed astrophysical analyses, which can be adjusted post-event, the LLAP must be operational from the start of LISA’s mission to enable prompt identification, analysis, and follow-up of transient events. This pipeline will leverage deep learning algorithms to process LISA’s raw phasemeter data and generate posterior parameter distributions for binary coalescences. This novel approach bypasses standard initial data processing stages, aiming to reduce alert latency for EM observatories. Moreover, it integrates data analysis tools and hardware, as the LLAP will be validated using experimental data from a Field-Programmable Gate Array-based LISA Delay Hardware Simulation developed at AEI Hannover. This experiment simulates onboard measurement chains and allows for real-time injection of GW signals during hardware testing. LASERGWO will pave the way for future multimessenger observations.