Millisecond radio transients are important both as tracers of intense, coherent radiative processes and as probes of ionized material in the Universe. The best known class of fast radio transient is the pulsar, a rotating neutron star that is detected through brief intense pulses of light \citep{1968Natur.217..709H,1982Natur.300..615B}. Active stars \citep{2008ApJ...674.1078O} and substellar objects \citep{2008ApJ...684..644H,2007A&A...475..359G} have distinct radio transient behavior that is rich with physical information. Radio light is modified by the ionized gas found throughout our Galaxy and between galaxies through the processes of dispersion, scattering, scintillation, and polarization. These processes can be used to measure distances, sizes, densities, and magnetic fields, and study Galactic structure \citep{2002astro.ph..7156C,2013ApJ...769..130C}.

[20]r0.72 \label{localization}A single-dish telescope localizes dispersed transients to a precision of \(\sim\)10\({}^{\prime}\), while interferometers produce images with arcsecond localization, orders of magnitude better. This image shows how we used the VLA for the first blind interferometric localization of a millisecond transient with a single detection —\citep2012ApJ…760..124L—.

Interferometers expand the science potential of fast transients and pulsars by their ability to localize. Interferometry improves on single-dish techniques through unique multiwavelength associations, unbiased measurements of source flux and spectra, and robust rejection of bad data. Figure \ref{localization} compares the single-dish and interferometer data streams, showing how we used a single pulse to localize a transient neutron star, J0628+0909, two orders of magnitude better than the previous single-dish localization. The combination of this new approach with a commensal observing mode will dramatically expand the discovery potential of the VLA and open access to a range of science topics.

The discovery of more than ten isolated millisecond pulses by the Parkes, Arecibo, and Greenbank Observatories has revealed a new population of radio transients, the FRB \citep2007Sci…318..777L,2013Sci…341…53T,2014ApJ…792…19B,Masui+submitted. FRBs are highly dispersed (frequency-dependent arrival time) with the magnitude of this dispersion, the “dispersion measure” (DM), up to \(\sim\)1600 pc cm\({}^{-3}\). This is more than an order of magnitude higher than expected from the Galaxy, which suggests that they come from cosmological distances \citep{2014ApJ...780L..33M}. These mysterious transients were discovered in large pulsar surveys at 1.4 GHz, which steadily continue to discover them \citep{2015MNRAS.447..246P,2015ApJ...799L...5R}.

What are FRBs and how can we use them? Models for a cosmological origin must account for isotropic radio luminosities of order \(10^{43}\) erg s\({}^{-1}\), far beyond that of the most common source of millisecond radio transients, Galactic neutron stars \citep[\(\sim 10^{34}\) erg s\({}^{-1}\);][]2003ApJ…596..982M. Yet, with an estimated rate of roughly 10\({}^{-3}\) galaxy\({}^{-1}\) yr\({}^{-1}\), these events are roughly as common as core-collapse supernovae \citep{2006Natur.439...45D}.

The number of models of FRB origin now exceeds the number of known FRBs, reflecting the excitement of the field and the urgent need for new observations. Origin models have invoked the giant pulses of relatively ordinary pulsars \citep{2015arXiv150505535C}, the merger of white dwarfs and neutron stars \citep{2013ApJ...776L..39K,2013PASJ...65L..12T,2014MNRAS.441.2433R}, the collapse of magnetars into black holes \citep{2014A&A...562A.137F,2014ApJ...780L..21Z,2014MNRAS.437.1821M}, and ordinary stars in the Milky Way \citep{2015MNRAS.454.2183M}. Models predict that FRBs could be associated with either type Ia supernovae, gamma-ray bursts \citep[potentially seen as kilonovae;][]{2013Natur.500..547T}, the first gravitational wave sources \citep{2014arXiv1408.0013D}, or have no associated multi-wavelength transient \citep{2015arXiv150100753C}.

[21]l0.35 \label{igmdm}Fast radio transients are dispersed as they propagate through the IGM or ISM. For radio sources associated to a distance, the DM directly measures the electron (and baryon) density along the line of sight. Measuring IGM DM as a function of distance would constrain the dark energy equation of state.

If cosmological, radio transients offer a unique and powerful way to study the intergalactic medium (IGM). Dispersion and scattering induced by propagation are easy to measure \citep{2003ApJ...596..982M} and constrain the environment near FRBs \citep{2013ApJ...776..125M,2014ApJ...785L..26L}. For FRBs at known distances (e.g., associated with host galaxy or afterglow), the dispersion and distance will directly measure the IGM electron density (Figure \ref{igmdm}); hundreds of such detections at high redshift will constrain the dark energy equation of state \citep{2014PhRvD..89j7303Z}. Induced polarization could be used to measure the magnetic field strength in the IGM, which is hard to measure by other means \citep{2014arXiv1406.2526F}. Scintillation induces spectral structure to study the propagation medium and constrain the size of the source \citep[e.g.,][]{2014MNRAS.442.3338P}.

The discovery of FRBs has been confused with the simultaneous discovery of “perytons”, a type of terrestrial interference at the Parkes observatory \citep{2011ApJ...727...18B}. Recently, \citet{2015MNRAS.451.3933P} conclusively showed that these perytons originate from a microwave oven at the local visitor center. While this discovery highlights the risk of confusing interference and astrophysical transients, it has strengthened the argument for an astrophysical origin for FRBs by providing a physical model for peryton origin. Beyond the fact that perytons have non-physical spectral (do not follow plasma dispersion or scattering laws) and spatial properties (always spatially extended), they are always correlated with an on-site interference monitor and only detected when the visitor center is open and the telescope is pointed toward it. Using these rules to define perytons shows that all Parkes-detected FRBs are astrophysical. This, combined with detections at Arecibo \citep{2014ApJ...790..101S} and the Green Bank Telescope (Sievers, Masui, Chang et al., submitted), show that FRBs are an astrophysical phenomenon.

Our proposed system will detect FRBs brighter than \(\sim 100\) mJy with 3 ms integrations with a precision better than 3\({}^{\prime\prime}\). At this sensitivity, the VLA will find one FRB for every two hundred hours of L-band observing (every two months; §\ref{tos}). Observations above 2 GHz (e.g., the VLA sky survey) would have finer localization and lighter computing burden, which would make the VLA a powerful platform to study FRBs in a new spectral domain. Each VLA detection will be localized well enough for unique multiwavelength associations and measurements of FRB flux, spectrum, and polarization that can be corrected for primary beam effects (unlike single-dish detections). Models predicting a Galactic stellar counterpart will be easily tested with optical/IR imaging of modest depth. Deeper optical imaging will find unique galaxy associations for FRBs out to z\(\sim\)1. We will make rapid follow-up observations and co-observing possible, which may be as critical to understanding FRB origin as it has been for understanding GRBs \citep{1997Natur.387..878M,2013Natur.500..547T}.

Compact objects occupy a special role in the galactic nuclei and star clusters. Dense stellar environments favor different channels for compact object formation, so discovering such objects provides novel insight. Furthermore, fast transients allow us to measure dispersion and map the electron density and mass distribution in these regions. Interferometric fast imaging is ideal for studying these regions, because they search large fields of view and associate radio sources with optical or X-ray counterparts.

The “rotating radio transient” (RRAT) is a neutron star transient that pulses irregularly, but with a pulsar-like period of a few seconds \citep{2006Natur.439..817M}. RRATs (much like FRBs) were discovered by searching large, single-dish radio surveys for individual millisecond pulses. But only about 30% of the roughly 90 known RRATs¹¹A catalog of known RRATs is available at http://astro.phys.wvu.edu/rratalog. have timing solutions sufficient to break the degeneracy between the RRAT’s position and spin-down parameter, which needs to be measured to infer magnetic field strength. For that reason, it is not clear whether RRATs are part of an ordinary class of neutron star \citep[exhibiting the “giant pulse” or “nulling” phenomena;][]2006ApJ…645L.149W,2010MNRAS.402..855B or the magnetar class \citep{2007ApJ...670.1307M,2009MNRAS.400.1439L}.

The proposed system would localize RRATs to better than arcsecond precision with a single detection to unambiguously identify (or rule out) association with X-ray counterparts expected for a magnetar origin \citep{2006ApJ...639L..71R}. We demonstrated this technique through fast imaging with of RRAT J0628+0909 with the VLA \citep[Figure \ref{localization} and][]2012ApJ…760..124L. Given the density of known RRATs, an observation in the Galactic plane will find one in \(\sim 60\) hours of L-band observing, which suggests that several will be found by chance in a typical observing semester (see §\ref{tos}).

In the several decades since their serendipitous discovery, radio pulsars have proven to be incredibly useful tools for exploring and constraining a variety of topics in physics and astrophysics. With ever-increasing instrumental and computing capabilities, the ability of new pulsar discoveries to continue to advance our understanding of physics shows no sign of slowing; rather, pulsar science is a key motivation for proposed future large-area radio telescopes such as the SKA \citep[e.g.,][]{2004NewAR..48..993K,2015aska.confE..36K}. While the next generation of pulsar searches is expected to be done with interferometric instruments, this approach has been relatively unexplored in the real world – nearly all known pulsars have been discovered using large single dishes \citep[e.g.,][]{2001MNRAS.328...17M,2006ApJ...637..446C,2014ApJ...791...67S}. In addition to the single-pulse searches already described, our proposed system for high time resolution imaging with the VLA offers an excellent opportunity to gain practical experience searching for periodic signals in this type of data.

With few-ms time resolution, our system will be limited to detecting pulsars with spin period \(\gtrsim\)10 ms. While this excludes the possibility of detecting fully-recycled millisecond pulsars, interesting sources in the 10–100 ms range include both fast-spinning young neutron stars (e.g., Crab, Vela) and mildly recycled double-NS binary systems. The latter enable precise NS mass measurements and direct tests of general relativity \citep{1989ApJ...345..434T,2006Sci...314...97K}. Furthermore, a pulsar-black hole binary system is a highly anticipated discovery that would provide an unprecedented gravitational laboratory, precise measurements of black hole mass and spin, and tests key components of general relativity such as the “cosmic censorship” conjecture and “no-hair” theorem \citep{1996PhRvD..54.1474D,2004NewAR..48..993K,2015aska.confE..40K,2014MNRAS.445.3115L}. Detection of such sources may be challenging due to the high expected acceleration in the binary \citep{2014MNRAS.445.3115L}, this factor motivates additional work on optimized algorithms for interferometric searches.

In addition to improved survey speed due to the wider field of view, interferomtric searches offer several other improvements over traditional single-dish pulsar searches. One is immediate localization on the sky: For pulsars, astrometry is traditionally derived from timing observations – in some cases, such as for MSPs, to extremely high levels of precision comparable (or better) to that obtained with VLBI \citep{2008ApJ...679..675V}. However, this method is based on Earth motion, so it requires at minimum one, and ideally three or more, years of monitoring to provide accurate results. For shorter monitoring periods, the measured source position is highly covariant with spin-down rate and other long-term timing parameters. In modern large pulsar surveys, it is often impractical to devote this much follow-up time to every newly-discovered source. Only those pulsars with an indication of new physics are pursued fully, but that selection process may be biased and incomplete. Arcsecond localization corresponds to timing precisions on the order of milliseconds (\(\delta t\sim 1\mathrm{AU}\times\delta\theta\)), so VLA imaging would be an ideal complement to timing of slow pulsars. This “extra” information will help measure spin-down (and in turn magnetic field), and identify any other interesting timing fluctuations deserving of more intense followup, on much shorter timescales and with less investment of telescope time than a traditional timing campaign.

Finally, our search would be sensitive to very long-period pulsars (\(P\gtrsim 5\) s), which are hard to find with single dishes due to “\(1/f\)” receiver gain fluctuations and man-made interference. This will help to fill in a currently poorly-constrained region of the pulsar population parameter space.

A series of recent observations of late-type dwarfs and substellar objects (“ultracool dwarfs”) demonstrates that they are capable of generating coherent, periodic bursts \citep{2002ApJ...572..503B,2006ApJ...653..690H,2013ApJ...767L..30W}. The emission mechanism for the radio emission from these objects is clearly tied to their ability to generate a large-scale magnetic field \citep{2008ApJ...684..644H}, though the mechanism is poorly understood \citep[and references within]{r12}. Beyond testing fundamental magnetic dynamo theory, late-type dwarf stars are important because they dominate the stellar population of the Galaxy and thus represent the most likely place to find extrasolar planets in the solar neighborhood \citep{2007AsBio...7...30T}. Intense, magnetospheric activity is correlated with intense UV emission and massive ejections of plasma that could affect the habitability of the nearest extrasolar planets \citep{2014MNRAS.439.3225L}.

Our proposed system will enable blind searches for radio emission from nearby ultracool dwarfs. Previous searches have targeted specific objects for limited durations and may be subject to selection biases related to the duty cycle of the magnetically-generated emissions. By contrast, our proposed system will be able to build up a local “exposure map” for ultracool dwarfs, to either enable their discovery or place constraints on their activity. Our search will find objects in which magnetic dynamos can be studied, which is critical input for theories of the interiors of planets and ultracool dwarfs. Further, our radio observations alone will study the “space weather” of nearby ultracool dwarfs and the implications for habitability. Such a survey will be particularly timely in the context of forthcoming observations at visible wavelengths with both the LSST and the Transiting Exoplanet Survey Satellite \citep[TESS,][]2014ebi..conf.3.10R.

The transient detection system proposed will find transients on timescales as long as \(\sim 1\) min. This could be used to trigger new VLA observation in response to a short-lived stellar flare, such as a broad-band continuum observations to understand an emission mechanism \citep{2013ApJ...767L..30W}. In the case of our own Sun, astronomers could use the system to self-trigger data recording if/when the Sun goes into outburst during a multiwavelength campaign \citep{2013ApJ...763L..21C}.