VLBI data analysis breakthrough: fully automated pipelines

The calibration of radio Very Long Baseline Interferometry (VLBI) data is no longer confined to a small community of experts. The advent of the fully automated pipelines VIPCALs and AVICA has made the calibration of continuum VLBI data accessible to the broader astronomical community.
 

Cartoon representation of the workflow transforming VLBA observations of a radio-loud active galaxy into final calibrated data through a fully automated process enabled, for the first time, by VIPCALs and AVICA.

 

They are based on the two most widely used software packages for VLBI data reduction, AIPS (VIPCALs) and CASA (AVICA), and require only minimal user input: a complete data calibration can be performed by providing only the target and file names.  
The two pipelines were developed by Diego Álvarez-Ortega and Avinash Kumar, PhD students at the University of Crete (UoC) and IA-FORTH, in the context of the SMILE project, under the supervision of Dr. Carolina Casadio.


Both softwares have been validated on a representative test sample of 1000 radio-loud sources from the SMILE project covering a broad range of flux densities (50 - 1000 mJy), observing times (1 - 60 min), and frequency setups (4 - 43 GHz) from archival Very Long Baseline Array (VLBA) projects. The pipelines’ performance have also been compared against manually calibrated datasets from the Astrogeo database.  While manual calibration of VLBI data may commonly requires from several hours to several days of user-time, VIPCALs and AVICA calibrated the tested sample with average calibration times of 10 and 30 minutes per dataset, respectively.  


In summary, VIPCALs and AVICA enable robust, reproducible, and, for the first time, fully automated calibration of VLBI continuum data, significantly lowering the entry barrier for VLBI science and making large-scale projects such as SMILE feasible.

Signal-to-noise ratio comparison of three datasets calibrated manually (Astrogeo) and with the two automated pipelines, VIPCALs and AVICA. The number of detections is calculated per scan and per baseline.

 

Articles:

 

Repositories:

https://github.com/dalvarezo/VIPCALs
https://github.com/avikhagol/avica