While single-dish telescopes have pioneered the study of fast radio transients, interferometers are poised to transform the field. Interferometers form “synthetic” apertures many kilometers in diameter, which allows them to expand on every limitation of single-dish telescopes:
Precise localization: Interferometers image with arcsecond precision, as shown in the image of a pulsar pulse shown in Figure \ref{candplot}.
High survey speed: Interferometers have large fields of view and are powerful survey machines.
Robust calibration and interference rejection: Interferometers can measure fluxes more accurately and reject interference that complicates single-dish fast transient searches.
Interferometers are technically more demanding than single-dish telescopes because their fundamental measurement is the correlation of pairs of antennas. Thus, where a single-dish telescope has a single data stream (or a few, if using a multi-beam receiver), a comparable interferemeter like the VLA has 27 antennas and thus 351 data streams. An efficient algorithm for extracting transients from this massive data stream could revolutionize the study of fast transients by uniquely associating radio transients with multiwavelength counterparts (e.g., FRB host galaxies, RRAT NS hosts, stellar/planetary associations).
We have commissioned the Jansky Very Large Array (VLA) to observe with millisecond integrations and data rates of 1 TB hour\(^{-1}\) \cite{2012ApJ...760..124L}. We have also developed an extensive, parallelized software system to search visibility data for dispersed transients1. The pipeline is written in Python/Cython and run within the NRAO software package CASA 2.
Portions of the code base are available at http://github.com/caseyjlaw/tpipe.↩