Radio Transients Surveys


Introduction - Key Science Goals, Context, and The Fast/Slow Transient Split

The paper will present an overview of the science, techniques, and state-of-the-art of surveys for radio transients. The paper is aimed to non-experts such as early graduate students who are interested in working in the field, but will also provide detailed information for researchers currently working in the field. Progress in the field is rapid so there will be a heavy emphasis on long-term science goals and technqiues. I will discuss important developing topics such as fast radio bursts, where there is still considerable uncertainty about the significance of current discoveries.

The science and techniques of radio transient surveys are divided into domains: fast and slow. The dividing line is a characteristic time scale of 1 second. On short time scales, coherent emission processes such as the pulsar mechanism dominate. On longer time scales, incoherent synchrotron processes dominate. Technically, slow transient searches are in the domain of interferometric imaging. Fast transient searches have been traditionally the domain of high bandwidth single dish timing. Interferometric techniques, however, are becoming more important for fast transient searches. The next generation of large telescopes are primarily large-N interferometers. As these new telescopes are commissioned, interferometric techniques must be expoited in order for fast transient searches to advance scientifically.

The science of radio transients has primarily been driven by follow-up of optical and high energy events such as gamma-ray bursts and supernovae. Surveys have the opportunity to discover new classes of transient that are not known or predicted by multi-wavelength information. Serendipitous discovery has demonstrated that radio transient surveys can be critical for discovery of dust-obscured objects and coherent emitters that would not have been otherwise discovered. There is the potential for new radio transient surveys to flip the dominant paradigm in which radio discovery leads to optical and high energy follow-up.

The paper is timely because of the launch of major surveys with the Very Large Array and LOFAR and the planned launch of surveys with ASKAP and MeerKAT. Pre-cursor surveys with the VLA, the Allen Telescope Array, Parkes, Arecibo, and the GBT provide an important foundation for planning. The coming of Advanced LIGO GW interferometer results at the end of the decade will heighten interest in disovery of radio transient counterparts to GW events.

Fast Transient Science


Pulsars are the most common and most intensively studied of all radio transient sources. Since discovery of pulsars in 1967 (Hewish et al., 1969) and millisecond pulsars in 1982 (Backer et al., 1982), the study of pulsars has provided deep insights into stellar evolution, the structure of nuclear material, general relativity, and the interstellar medium. The extensive achievements of pulsar science have effectively created a separate and distinct sub-field of radio transient science. The study of pulsars has led to two Nobel Prizes in Physics (Hewish, for the discovery of pulsars; Hulse and Taylor, 1993, for the discovery of gravitational waves from a binary pulsar system). The power of pulsars to probe such a wide range of physics is the ultimate promise of radio transient discovery.

Pulsar phenomena and observational techniques have been reviewed extensively and recently (e.g., Lorimer et al., 2004; Kramer, 2008; Lorimer et al., 2010; Manchester, 2013). Briefly, pulsars are spinning neutron stars with strong magnetic fields that are misaligned to the spin axis. Radio emission is created through particle acceleration at the magnetic poles. This emission appears as time variable emission as the radio beam sweeps past the observer. Pulsars are primarily detected through this periodic emission with some notable exceptions discussed below.

More than 2400 pulsars are known, primarily as the result of large area surveys conducted with large single dish radio telescopes Approximately, 10% of pulsars have periods \({\stackrel{\scriptstyle <}{\scriptstyle \sim}}10\) milliseconds and are known as millisecond pulsars (MSPs). Pulsar periods span from 1.4 milliseconds to 8.5 seconds (Hessels et al., 2006; Young et al., 1999). Pulsar spectra are typically steep with indices \(\alpha\approx -1.7\) , with some notable exceptions of flat spectra \citep{}. The large majority of known pulsars are in the galactic plane with notable populations in globular clusters \citep{} and the Galactic Center (Johnston et al., 2006; Deneva et al., 2009; Shannon et al., 2013; Eatough et al., 2013). Extragalactic pulsars have only been detected in the Magellanic Clouds (e.g., Ridley et al., 2013). More discussion of properties? luminosity range, DM?, polarization properties, spectra, lifetimes

Intermittency of the pulsed radio emission is an important property for radio transient science. Many pulsars exhibit time variable emission properties, including cases where the periodic emission turns off entirely for extended periods of time, either as the result of intrinsic physics or interactions with a binary companion \citep{}. This can lead to time variable sources that resemble transients in conventional slow transient imaging.

Detection of single pulses is central to two populations, rotating radio transients (RRATs) and giant pulses. RRAT pulses appear in phase with the pulsar period but only at random intervals with duty cycles \({\stackrel{\scriptstyle <}{\scriptstyle \sim}}10^{-5}\) (McLaughlin et al., 2006; Keane et al., 2011). The periods of RRATs are typically longer than that of the average pulsar suggesting a connection with anomalous X-ray pulsars and magnetars. RRATs are estimated to be as numerous as field pulsars. The behavior of RRATs may represent an extreme end of the intermittency spectrum of radio pulsars (Young et al., 2014).

Giant pulses are defined as individual pulses that exceed the mean pulse by at least an order of magnitude. The Crab pulsar was discovered through its giant pulses (Staelin et al., 1968), which can exceed the mean flux density by a factor greater than 1000 and occur with a power-law distribution (Mickaliger et al., 2012). Giant pulses primarily appear simultaneously with the main pulse and interpulse. High time resolution observations have found that some giant pulses are unresolved with a time resolution of 1 nanosecond (Hankins et al., 2003). Giant pulses have also been detected from the MSP PSR B1937+21 (Cognard et al., 1996). The very high flux density and short duration of giant pulses implies extraordinarily high brightness temperatures, i.e., \(T_B \approx 10^{37}\) K. This extreme brightness temperature makes giant pulses appealing targets for extragalactic single pulse searches (McLaughlin et al., 2003).

**May be useful to have discussion here or more generally about how classes are defined. Generally, new classes are defined observationall (RRATs, GPs, FRBs), but they develop more physical definitions as a theory is developed. GPs now have some theory, so they are not exclusively defined by 10x the mean brightness. RRATs are in this transition period and may in fact be purely observational phenomenon (Keane et al). FRBs are still pre-theory.**

Fast radio bursts and perytons

Fast radio bursts (FRBs) and perytons are both single pulse events that are phenomenologically closely related to RRATs and giant pulses. They are distinguished by their apparently large dispersion measures (DMs, discussed in Section \ref{sec:fastprop}) and their very low repetition rate. FRBs have been interpreted as cosmological sources, although numerous alternative explanations have been put forward. Perytons, on the other hand, appear to be terrestrial in orign, and possibly anthropogoenic.

FRBs, also known colloquially as “Lorimer bursts,” were first reported by Lorimer et al. (2007). The first event, FRB 010824, identified by the event epoch, shares the primary properties of the FRB class. A single, very bright dispersed pulse was discovered at a frequency of 1.4 GHz in a Parkes multi-beam survey at high galactic latitude. The pulse dispersion with frequency was well-fit with a \(\nu^-2\) dependence and the observed DM was an order of magnitude in excess of that predicted by galactic models. In addition, the pulse width appeared to include an intrinsic width \({\stackrel{\scriptstyle <}{\scriptstyle \sim}}5\) milliseconds and a frequency-dependent width that scaled as \(\nu^-4.8\), roughly consistent with the expectations of scatter-broadening. There was no obvious association with an energetic time variable source such as a gamma-ray burst and the relatively poor localization of the single dish telescope (\(\sim 7\)) prevent association with any known objects. Repeated observations of the same field after discovery did not find any other events. Thus, the defining characteristics of the FRB class are bright, single pulse events demonstrating plasma propagation properties consistent with very large electron column densities.

Approximately 10 FRBs have now been reported in the literature, all discovered with single dish telescopes and thus, poorly localized, and all but one discovered with the Parkes telescope (Keane et al., 2011; Thornton et al., 2013; Burke-Spolaor et al., 2014; Spitler et al., 2014; Petroff et al., 2014; Ravi et al., 2014). Several surveys have failed to discover FRBs, possibly due to a lower event rate, frequency-dependent effects or propagation effects (Deneva et al., 2009a; Siemion et al., 2012; Coenen et al., 2014; Petroff et al., 2014a; (citation not found: Law2015). The event rate for FRBs is uncertain but is \(\sim 10^4\) sky\(^{-1}\) day\(^{-1}\) above a fluence of \(\sim 1\) Jy \(\times\) msec. DMs as large as 1100 pc cm\(^{-3}\) have been detected.

A cosmic origin for FRBs has been called into question by the discovery of perytons (Burke-Spolaor et al., 2011). Also discovered as single pulse events, perytons exhibited large apparent DMs and narrow pulse widths. Dispersion, however, was not always purely parabolic and pulse widths sometimes exceeded 10 milliseconds. Further, perytons had the remarkable characteristic of appearing simultaneously in multiple independent beams of the Parkes multi-beam receiver, most likely indicating an origin in the far sidelobes of the antenna, where receiver sensitivity is isotropic. Further, the clustering of events in time appeared not to be stochastic and possibly related to the integer second (Kocz et al., 2012). These facts together indicated an atmospheric or anthropogenic origin for perytons. Perytons and FRBs have been considered as a unified class with an atmospheric origin (Kulkarni et al., 2014).

If the large DM associated with FRBs is in fact from a cosmic origin, then FRBs may lie at cosmological distances. The Galaxy, the intergalactic medium, the host galaxy, and the host object may all contribute to the total DM. The Galactic contribution is small by definition, while the host galaxy contribution could be quite large if events are seen through an edge-on galaxy. The latter is unlikely to be dominant, however, because there are not apparent non-repeating single pulse events with Galactic DMs. Host object components to the DM have been proposed but are limited by the free-free absorption that would occur (Loeb et al., 2014; Kulkarni et al., 2014).

Thus, the IGM object component appears to be substantial. The IGM plasma density is small \citep[$\sim 10^{-7}$ cm$^{-3}$;][]{2004MNRAS.348..999I} but path lengths are very large. The population of FRBs is projected to extend out to \(z\sim 1\). This provides a powerful tool for probing and characterizing the diffuse ionized plasma of the IGM, which is otherwise difficult to detect (Ioka, 2003; Inoue, 2004; Macquart et al., 2013; Zheng et al., 2014). Localization of FRBs to host galaxies with a measured red shift provides the necessary counterpart to measured plasma effects to translate observational results into cosmological constraints.

The very large distances and large rates for FRBs present a challenge to models for the intrinsic source. Luminosities must exceed those of known Galactic fast transients by orders of magnitude. The rates above current detection thresholds already exceed those of gamma-ray bursts and the supernova population within a Gpc\(^3\) (Siemion et al., 2012). Numerous models with origins from compact objects and other sources have been developed, including predictions of evaporating black holes (Rees, 1977), merging neutron stars (Hansen et al., 2001), other neutron star magnetospheric activity (e.g., Falcke et al., 2014), and cosmic strings (Cai et al., 2012).

Stellar and exoplanet coherent emission

The Sun exhibits steady quiescent emission as well as variable emission on timescales from the very short to the very long with a variety of physical processes including both coherent and incoherent mechanisms (Güdel, 2002; Bastian, 2004; Benz, 2008). The full scope of solar activity is beyond the bounds of this review, however, that activity exists on a continuum of stellar activity that stretches over 8 orders of magnitude in luminosity.

Coherent emission processes in sub-stellar and stellar objects generate fast transients and have primarily been observed in brown dwarfs and highly magnetized low mass stars (Berger, 2002; McLean et al., 2012) although there are important cases with other stellar types (e.g., Slee et al., 2008). Among the most exciting of results is the detection of periodic emission from brown dwarfs (Berger et al., 2005; Hallinan et al., 2006; Hallinan et al., 2007). Short bursts of emission are detected with periods of hours, which have been determined to be the rotational period of the brown dwarf. The emission is frequently highly circularly polarized and has structure on sub-minute timescales. The emission is likely generated by the electron cyclotron maser process from regions with magnetic fields that are \({\stackrel{\scriptstyle >}{\scriptstyle \sim}}1\) kG. The radio emission exceeds by orders of magnitude what is predicted by the radio-X-ray luminosity correlation seen over a wide range of stellar types and luminosities.

Transient planetary radio emission was discovered by Burke et al. (1955) with observations of Jupiter at a frequency of 22 MHz. This decametric (DAM) radiation had an upper frequency cutoff of 40 MHz. The spectrum is dynamic and includes multiple components that are strongly circularly polarized, highly beamed, drift rapidly in frequency, and have timescales that range from milliseconds (S bursts) to hours (Zarka, 1998). The emission is many orders of magnitude more luminous than thermal emission or steady synchrotron emission from the magnetosphere. Similar features are observed from Saturn, Uranus, and Neptune.

DAM emission originates from an interaction in the Jovian magnetosphere with energetic electrons, either injected from volcanic activity on the moon Io or from other sources (non-Io DAM). The upper cutoff is set by the cyclotron frequency for the magnetic field: \[\nu_{max} = 2.8 B {\rm\ MHz},\] where \(B\) is the magnetic field in Gauss. The inferred magnetic field strength for Jupiter is 14 G.

If moved to distance of a few parsecs, Jovian bursts are undetectable by even the largest of existing or planned low frequency radio telescopes. However, larger magnetic field strengths and higher electron fluxes could lead to detectable bursts from exoplanets. Scaling laws for magnetospheric strength as a function of radius, suggest that hot Jupiters, i.e., gas giants close to their host star, may have detectable emission (Zarka et al., 2001; Lazio et al., 2004). A limited number of searches have been carried out for exoplanet radio emission without success (Bastian et al., 2000; Lazio et al., 2007;