The next generation uGMRT interferometric data processing pipeline: exploring data archives for new science

The Upgraded Giant Metrewave Radio Telescope (uGMRT) is a sensitive low frequency radio interferometer that operates in the frequency range of 100 -1400 MHz.With 30 antennas and 400 MHz instantaneous bandwidth, it produces data at rates that require automated data processing pipelines running on high performance computing machines.We have developed an automated pipeline, CAsa [2] Pipeline-cum-toolkit for Upgraded GMRT data REduction (CAPTURE) that produces continuum images from raw visibilities [1].CAPTURE is publicly available (https://github.com/ruta-k/CAPTURE-CASA6) and is in active use, both, as a stand alone pipeline and as a toolkit to build customized analysis pipelines [e.g. 3, 4].Our next goal is to improve CAPTURE to explore archival GMRT data for science topics of growing interest such as long period transients.We are optimising the pipeline to bring it in a new avatar which will i) make efficient use of parallel computing, improving the throughput; ii) utilise machine learning based methods for radio frequency interference (RFI) flagging; iii) produce science-ready data products and iv) have new capabilities for searching for long period transient sources.The development of this pipeline will be described and the results from the first tests presented.We will discuss the results in the context of the upcoming Square Kilometre Array Observatory science and data processing.