Speaker
Description
With baselines up to 2000 kilometers, the International LOFAR Telescope has a unique high-resolution view on the low-frequency radio sky. Unlike more typical VLBI arrays, however, with its 76 operational stations it forms a sizable network with dense uv coverage across angular scales from approximately a degree down to a few tenths of an arcsecond. Due to its low frequency nature, the ILT's field of view is a substantial 6.25 square degrees on the sky. Unfortunately, successful science exploitation of the longest baselines across such an area requires data to be stored at high time and frequency resolution, making typical ILT observation around 5 TB in size. Needing around double the observations to cover the entire northern sky compared to the LOFAR Two-Metre Sky Survey done with only the Dutch stations, a data processing strategy that does processing at scale is needed.
In this talk I will present an overview of the LOFAR HBA VLBI pipeline for data between 120 and 168 MHz from two sides. First I will discuss how we tackle ILT calibration in general. Its main challenges such as the ionosphere will be discussed, as well as the differences with more traditional VLBI. Secondly, I will cover how we orchestrate the data reduction at scale by using the Common Workflow Language to interface with managed clusters like Slurm. Current efforts are focussing on delivering a first end-to-end pipeline to go from raw data to a widefield sub-arcsecond resolution image without human intervention, to be prepared for when LOFAR2.0 comes online.
