In radio astronomy, meeting the data challenge of the Square Kilometre Array : launch of the ASTRA project
Data volumes so large that they can no longer be processed and analysed on local machines, established processing methods that are difficult to adapt to distributed architectures, and the challenge of making these resources accessible to astronomers who are not specialists in radio astronomy : these are some of the challenges faced by the scientific community with the arrival of instruments such as the Square Kilometre Array Observatory (SKAO).
France recently became the 14th member country of the SKAO, marking an important step in its involvement in this new era of radio astronomy.
![<multi> [fr]Les membres de la collaboration ASTRA réunis à l'Observatoire de Paris-PSL lors de la réunion de lancement du projet. [en]The ASTRA collaboration meets at the Observatoire de Paris </multi>](IMG/png/kick-off_astra_20260617.png)
The ASTRA project
The ANR-NumPEx ASTRA (Alliance for Scalable Techniques in Radio Astronomy) project, whose kick-off meeting took place from 15 to 17 June at the Observatoire de Paris-PSL, aims to address these challenges in preparation for the implementation of the French node of the European SKA Regional Center (SRC).
To achieve this goal, ASTRA will develop generic, unified and extensible platforms specifically designed to meet the challenges of modern radio astronomy.
The project will deploy four radio astronomy processing pipelines already used with SKA precursor instruments such as NenuFAR, ASKAP and LOFAR, on a distributed computing architecture involving major French high-performance computing centres.
These four pipelines correspond to four major scientific programmes of the SKA :
- pulsar observations ;
- the epoch of reionisation of the Universe ;
- wide-field radio imaging ;
- mapping of neutral hydrogen in the Magellanic Clouds.
![<multi> [fr]Légende : Les quatre thématiques scientifiques étudiées dans le cadre du projet ASTRA. En haut à gauche : représentation de la technique de Pulsar Timing Array (EPTA). En haut à droite : le Petit Nuage de Magellan observé à 21 cm (Pingel et al., 2022). En bas à gauche : la source radio B3 0157+405A issue du relevé LoTSS (crédit Maya Horton et l'équipe *surveys* de LOFAR). En bas à droite : simulation de l'époque de la réionisation (Alvarez et al., 2009). [en]Illustration of the four science cases of ASTRA. Top-left: Pulsar Timing Array representation from EPTA (https://www.epta.eu.org/news/ipta-dr2-news.html). Top-right: The Small Magellanic Cloud at 21 cm (Pingel et al. 2022). Bottom-left: The radio source B3 0157+405A from the LoTSS wide-area survey (credit Maya Horton and the LOFAR surveys team). Bottom-right: Simulation of the Epoch of Reionisation (Alvarez et al. 2009). </multi>](IMG/png/astra2.png)
En haut à gauche : représentation de la technique de Pulsar Timing Array (EPTA).
En haut à droite : le Petit Nuage de Magellan observé à 21 cm (Pingel et al., 2022).
En bas à gauche : la source radio B3 0157+405A issue du relevé LoTSS (crédit Maya Horton et l’équipe *surveys* de LOFAR).
En bas à droite : simulation de l’époque de la réionisation (Alvarez et al., 2009).
An alliance of skills
The ASTRA consortium, led by Marc-Antoine Miville-Deschênes (LPENS), brings together a wide range of complementary expertise in astrophysics, data processing and high-performance computing.
The project builds on developments carried out within the joint laboratories (LabCom) ECLAT and AERIAL, as well as within the European projects EXTRACT and SPECTRUM, and the NumPEx Exsa-AToW project.
The ASTRA collaboration brings together LPENS (ENS-PSL), IRISA (University of Rennes), CEA/IRFU, LUX and LIRA at the Observatoire de Paris-PSL, as well as APC, LPC2E, LAB and Lagrange.
As a true alliance between the French actors involved in radio astronomy data processing, ASTRA will strengthen the links between the astronomical community and the high-performance computing community. The project will run for four years.
Contact
LUX : Florent Mertens
florent.mertens at obspm.fr
