Closing the Loop From Astronomical Data Exploration To Scientific Discovery Usin... M. López-Caniego

Closing the Loop From Astronomical Data Exploration To Scientific Discovery Using the Jupyterverse - Marcos López-Caniego, Aurora Technology for the European Space Agency A big fraction of the new discoveries in astronomy come from the reanalysis of archival data. The European Space Agency’s astronomy science archives are using the Jupyter ecosystem to connect the queries from the users in their archives to ESA Datalabs, the science platform sitting next to the data at the European Space Astronomy Centre in Madrid, Spain. This is particularly important for missions like ESA’s Euclid cosmology mission that will generate tens of petabytes of data in the next few years. Other applications like ESASky and the European James Webb Space Telescope science archive already generate notebooks in real time based on the user’s queries that with the click of a button can be executed in the science platform, closing the loop from data discovery and data analysis to scientific discovery. Moreover, we are also exploring new ways to facilitate data discovery using multi-agent systems and generative AI to translate natural language to astronomy SQL-like queries, generating jupyter notebooks with the necessary code to query the data bases and discover data in large astronomical data repositories where the Jupyter ecosystem is playing a crucial role. Co-Authors:Javier Espinosa (Starion for ESA), Maria Arévalo (Starion for ESA), Miguel Doctor (Telespazio for ESA), Deborah Baines (ESA), Rachana Batawdekar (ESA), Bruno Merin (ESA), Vicente Navarro (ESA), and Sandor Kruk (ESA).