Resolving the decades-long transient FIRST J141918.9+394036: an orphan long gamma-ray burst or a young magnetar nebula?
B. Marcote, K. Nimmo, O. S. Salafia, Z. Paragi, J. W. T. Hessels, E. Petroff, R. Karuppusamy
DDraft version April 22, 2019
Typeset using L A TEX twocolumn style in AASTeX62
Resolving the decades-long transient FIRST J141918.9+394036: an orphan long gamma-ray burst or a youngmagnetar nebula?
B. Marcote, K. Nimmo,
2, 3
O. S. Salafia,
4, 5, 6
Z. Paragi, J. W. T. Hessels,
2, 3
E. Petroff, andR. Karuppusamy Joint Institute for VLBI ERIC, Oude Hoogeveensedijk 4, 7991 PD Dwingeloo, The Netherlands ASTRON, Netherlands Institute for Radio Astronomy, Oude Hoogeveensedijk 4, 7991 PD Dwingeloo, The Netherlands Anton Pannekoek Institute for Astronomy, University of Amsterdam, Science Park 904, 1098 XH Amsterdam, The Netherlands Universit`a degli Studi di Milano-Bicocca, Dip. di Fisica “G. Occhialini”, Piazza della Scienza 3, 20126 Milano, Italy INAF - Osservatorio Astronomico di Brera, via E. Bianchi 46, 23807 Merate, Italy INFN - Sezione di Milano-Bicocca, Piazza della Scienza 3, 20126 Milano, Italy Max-Planck-Institut f¨ur Radioastronomie, Auf dem H¨ugel 69, D-53121 Bonn, Germany (Received XXX; Revised XXX; Accepted XXX)
Submitted to ApJLABSTRACTOfek (2017) identified FIRST J141918.9+394036 (hereafter FIRST J1419+3940) as a radio source sharingsimilar properties and host galaxy type to the compact, persistent radio source associated with the first knownrepeating fast radio burst, FRB 121102. Law et al. (2018) showed that FIRST J1419+3940 is a transient sourcedecaying in brightness over the last few decades. One possible interpretation is that FIRST J1419+3940 is anearby analogue to FRB 121102 and that the radio emission represents a young magnetar nebula (as severalscenarios assume for FRB 121102). Another interpretation is that FIRST J1419+3940 is the afterglow of an‘orphan’ long gamma-ray burst (GRB). The environment is similar to where most such events are produced.To distinguish between these hypotheses, we conducted radio observations using the European VLBI Networkat 1.6 GHz to spatially resolve the emission and to search for millisecond-duration radio bursts. We detectFIRST J1419+3940 as a compact radio source with a flux density of ± µ Jy (on 2018 September 18)and a source size of . ± . (i.e. . ± . given the angular diameter distance of
83 Mpc ). Theseresults confirm that the radio emission is non-thermal and imply an average expansion velocity of ( . ± . ) c . Contemporaneous high-time-resolution observations using the 100-m Effelsberg telescope detected nomillisecond-duration bursts of astrophysical origin. The source properties and lack of short-duration burstsare consistent with a GRB jet expansion, whereas they disfavor a magnetar birth nebula. Keywords: radio continuum: transients – gamma-ray burst: individual: FIRST J1419+3940– fast radiobursts – galaxies: dwarf – radiation mechanisms: non-thermal – techniques: high angularresolution INTRODUCTIONVery-long-baseline radio interferometric (VLBI) ob-servations are a powerful way to study astrophysicaltransients because they provide milliarcsecond angu-lar resolution imaging and astrometry. Such transientevents are produced by blast waves and slowly-evolvingsynchrotron afterglows, whose temporal evolution and
Corresponding author: B. [email protected] interaction with the surrounding medium are well char-acterized by VLBI observations that can measure theprojected size and proper motion of such emission.For example, this technique was successfully used tospatially resolve the emission and measure the expansionspeed of the afterglow associated with the long gamma-ray burst (GRB) 030329 (Pihlstr¨om et al. 2007). VLBIobservations have also been used to study the first de-tected binary neutron star merger, GRB 170817A (Ab-bott et al. 2017). The obtained measurement of theproper motion and physical size constrained the nature a r X i v : . [ a s t r o - ph . H E ] A p r Marcote et al. of the source to be a relativistic jet (Mooley et al. 2018;Ghirlanda et al. 2019). Furthermore, VLBI observa-tions contributed to the first precise localization of a fastradio burst (FRB), the repeating source FRB 121102(Spitler et al. 2014, 2016; Scholz et al. 2016). The burstsource was associated with a compact ( < . ; Mar-cote et al. 2017), persistent radio source with a lumi-nosity of ν L ν ≈ × erg s − at 1.7 GHz (Chatterjeeet al. 2017), located inside a low-metallicity star-formingregion in a dwarf galaxy at a redshift of . ( ) (Ten-dulkar et al. 2017; Bassa et al. 2017). The environmentof FRB 121102 is remarkably similar to the ones wherelong GRBs (as well as superluminous supernovae) typ-ically occur (Modjaz et al. 2008; Metzger et al. 2017),favoring several scenarios that consider repeating FRBsto be produced by newly-born magnetars created insuch events (see e.g. Margalit & Metzger 2018; Piro &Gaensler 2018). FRBs could thus be detectable at thesites of long GRBs, and the persistent source associ-ated with FRB 121102 could be the longer-lived nebulafollowing the afterglow of one of these events. In anycase, FRBs are expected to be produced in relativelyyoung objects ( ∼ –
100 yr ) with possibly associated ra-dio nebulae (Murase et al. 2016; Kashiyama & Murase2017; Omand et al. 2018).Based on the properties of FRB 121102’s persistent ra-dio source and host galaxy, Ofek (2017) identified a num-ber of similar sources in the Very Large Array (VLA)FIRST catalogue. Law et al. (2018) showed that oneof these sources, FIRST J141918.9+394036 (hereafterFIRST J1419+3940) is a slowly declining transient. Us-ing archival observations, they showed that the sourcedeclined from ∼
26 mJy (at 1.4 GHz) in 1993 to (cid:46) . (at 3 GHz) in 2017. Both the light-curve and the inferredluminosities of ν L ν (cid:38) × erg s − are consistent withthe afterglow of a long GRB, requiring a released kineticenergy of ∼ erg at the time of the explosion (esti-mated to be ∼ –
30 yr ago). No convincing associationwith a previously detected GRB could be made, however(Law et al. 2018).FIRST J1419+3940 is associated with a small star-forming galaxy at redshift of z ≈ . . Both sources,FIRST J1419+3940 and FRB 121102, show similar en-vironments: both show compact and persistent radioemission with luminosities of ∼ erg s − located in-side star-forming regions with equivalent star formationrates in similar sized dwarf galaxies. Their physical na-ture could thus also be similar, and FIRST J1419+3940might be associated with a source capable of producingFRBs.Here we present European VLBI Network (EVN) ra-dio observations of FIRST J1419+3940 that provide the first constraints on the source compactness, cou-pled with simultaneous searches for millisecond-durationbursts. We present the observations and data reductionin Section 2. We describe the results in Section 3, andtheir implications for the nature of FIRST J1419+3940in Section 4. Finally, we present our conclusions in Sec-tion 5. OBSERVATIONS AND DATA REDUCTIONWe observed FIRST J1419+3940 on 2018 September18 between 12:00 and 19:00 UTC at 18 cm (1.6 GHz)with the EVN, involving a total of 12 stations: JodrellBank Mark2, Westerbork single-dish, Effelsberg, Medic-ina, Onsala 25-m, Tianma, Toru´n, Hartebeesthoek, Sar-dinia, and three stations from e-MERLIN (Cambridge,Defford, and Knockin). The data were recorded with atotal bandwidth of 128 MHz, and correlated in real time(e-EVN operational mode) at JIVE (The Netherlands)using the SFXC software correlator (Keimpema et al.2015). The data were divided into eight subbands of 64channels each, with full circular polarization products,and 1-s time averaging. We also buffered the basebandEVN data in parallel so that high-time-resolution corre-lations could be produced afterwards, if a millisecond-duration radio burst was detected.Furthermore, we simultaneously observed FIRST J1419+3940in the frequency range – using the -mEffelsberg telescope and the PSRIX pulsar data recorder(Lazarus et al. 2016). We recorded with two summedlinear polarizations, achieving a gain of . K Jy − anda receiver temperature of K. The total bandwidthof
156 MHz was divided into subbands — each onefurther divided into channels and recorded with -bit time samples. The ultimate time and frequencyresolution of the data were . µ s and . ,respectively. Before processing, the subbands were com-bined into a single band and the data were converted to -bit samples to ensure compatibility with the PRESTO pulsar analysis software suite (Ransom 2001).2.1.
Interferometric data
We observed J1642+3948 as fringe finder and J1419+3821(located at only . ◦ from FIRST J1419+3940) asphase calibrator. We scheduled a phase-referencingcycle of 4.5 min on the target and 1.5 min on thephase calibrator, achieving a total time of ∼ . onFIRST J1419+3940. esolving a decades-long radio transient source Figure 1.
The obtained visibility data (amplitudes andphases) for the phase calibrator source, J1419+3821, afterthe original calibration (top) and in the alternate calibrationwhere the gain calibration of the Tianma station was cali-brated using a source model from only the stations with arobust amplitude calibration (bottom). See Sect. 2 for de-tails. Red dots represent data from the baselines includingthe Tianma station. Blue lines represent the source modelin each case.
The interferometric data were reduced using
AIPS (Greisen 2003) and Difmap (Shepherd et al. 1994) follow-ing standard procedures. A-priori amplitude calibrationwas performed using the known gain curves and sys-tem temperature measurements recorded individually oneach station during the observation. We used nominalsystem equivalent flux density (SEFD) values for thefollowing stations: Jodrell Bank Mark2, Tianma, andthe e-MERLIN stations. We manually flagged data af-fected by radio frequency interference (RFI) and then wefringe-fitted and bandpass-calibrated the data using thefringe finder and the phase calibrator. We imaged and The Astronomical Image Processing System (
AIPS ) is a soft-ware package produced and maintained by the National RadioAstronomy Observatory (NRAO). self-calibrated the phase calibrator in
Difmap to improvethe final calibration of the data. The obtained solutionswere then transferred to the target, which was subse-quently imaged.We note that Tianma did not produce reliable sys-tem temperature values during the experiment. Most ofthese measurements failed and the existing ones exhib-ited a much larger scatter than usual. Therefore we usedthe nominal SEFD for amplitude calibration. This typi-cally produces a satisfactory a-priori calibration that canbe further improved during imaging and self-calibration.Tianma, however, provides the longest East-West base-lines in our array with no equivalent baselines to com-pare with, and it also does not have short spacings toestablish a reliable station calibration. In this case imag-ing and parametrization of source properties by model-fitting is complicated due to the fact that some sourceparameters may correlate with the Tianma station gain(Natarajan et al. 2017).Figure 1 displays the visibility amplitudes and phasesas a function of projected baseline length in units ofobserving wavelength. The top panel shows the initialcalibration, with the Tianma data highlighted in red.The low amplitudes may be consistent with a sourcethat is very compact in general, but well resolved inthe East-West direction. The bottom panel shows thedata after we apply an amplitude correction factor forTianma, based on a source model obtained by only us-ing the stations with robust calibration. The requiredscaling factor was about three, implying that the stationcould have been much less sensitive than expected.Due to the uncertainty with the Tianma calibration,we decided to take the following procedure to ana-lyze the data. Instead of fitting an elliptical-Gaussianmodel brightness distribution to the u v -data in modelfitting, we assumed a circular-Gaussian brightness dis-tribution. This is expected to be less sensitive to un-certainties in station gain calibration (Natarajan et al.2017). In addition, we looked at the results derivedfrom the following cases: Tianma removed from thedata set, Tianma present with nominal gain calibration,and Tianma present but with its gain scaled to be inagreement with the most compact possible solution (asexplained above). As we will see in Sect. 3, the fit-ted source sizes differ somewhat, but in all cases theysupport the same main conclusion: that our target isresolved on milliarcsecond scales.2.2. High-time resolution data
The high-time-resolution Effelsberg data were ana-lyzed to search for individual millisecond bursts or aperiodic signal. First, using
PRESTO ’s rfifind , we Marcote et al. identified specific time samples and frequency chan-nels contaminated by RFI. The regions highlighted by rfifind and the frequency range – , asso-ciated with RFI from the Iridium satellites, were maskedprior to conducting the analysis. We then dedispersedthe -bit data using the PRESTO tool prepsubband for trial dispersion measures (DMs) in the range – . − . The resulting dedispersed time serieswere then searched for single pulses above a - σ thresh-old using PRESTO ’s single_pulse_search.py , whichapplies a matched-filter technique using boxcar func-tions of various widths, and in our search was sensitiveto burst durations in the range . µ s and .
02 s . Dy-namic spectra of the identified single-pulse candidateswere generated and inspected by eye to distinguish be-tween astrophysical signals and RFI.In addition, a Fourier-domain search was per-formed on each individual dedispersed time series using
PRESTO ’s accelsearch , in order to search for periodicsignals. Potential periodic signals were sifted using ACCEL_sift.py , and the remaining candidates were in-spected by eye after folding using prepfold .The RFI mitigation process was unable to remove allinstances of RFI in the data. We calculate that of the ∼ . - h Effelsberg on-source time, approximately . was examined for bursts and periodic signals. Note thediscrepancy between the Effelsberg on-source time ( ∼ . ) and the EVN on-source time ( ∼ . ) which isdue to Effelsberg’s longer slew time compared with otherantennas.The aforementioned analysis strategy was verified us-ing similar data targeting pulsar PSR B2020+28. Weperformed a blind search and detected both individualpulses and the known periodicity of this pulsar. RESULTS AND DISCUSSION3.1.
On the persistent emission
FIRST J1419+3940 is detected on 2018 September 18as a radio source that is compact on milliarcsecond scales(see Fig. 2), with a flux density of ± µ Jy at aposition of: α ( J2000 ) = h m . s ± . mas δ ( J2000 ) = ◦ (cid:48) . (cid:48)(cid:48) ± . mas , where the quoted uncertainties represent the 1- σ confi-dence interval and take into account the statistical un-certainties in the image (0.2 mas in both α and δ ), theuncertainty in the phase calibrator position (0.1 mas;Beasley et al. 2002; Gordon et al. 2016), and the es-timated uncertainties associated with the phase refer-encing technique (0.06 and 0.04 mas for α and δ , re-spectively; Pradel et al. 2006). The obtained position is + ◦ . . . D ec li n a ti on ( J ) . s h m . s Right Ascension (J2000) + ◦ . . . D ec li n a ti on ( J ) Figure 2.
Images of FIRST J1419+3940 at 1.6 GHz withthe EVN on 2018 September 18 derived from the two gaincalibrations performed on Tianma data (original, top, andscaled, bottom). Contours start at a 3- σ rms noise level of and µ Jy beam − , respectively, and increase by factors of √ . The synthesized beams are represented by the dark grayellipses at the bottom left corner of each image. consistent with the one reported from the FIRST sur-vey (Law et al. 2018), as well as the preliminary re-sults published in Marcote et al. (2018). The measuredflux density on 2018 September 18 follows the declin-ing trend of the light-curve reported from observationswith the Karl G. Jansky Very Large Array, VLA (seeFig. 3). Given the luminosity distance of 87 Mpc, theobtained flux density corresponds to an isotropic lumi-nosity ν L ν = ( . ± . ) × erg s − . Together with thelast published VLA observation at 3.0 GHz, and con-sidering the same value for our epoch (i.e. no decliningtrend is assumed), we can place a conservative 3- σ up- esolving a decades-long radio transient source . S ν ( m J y ) Figure 3.
Light-curve of FIRST J1419+3940 during thelast 25 yr at 1.4–1.6 GHz (blue circles and open square) and3.0 GHz (orange circles). Errors bars represent 1- σ uncer-tainties (hidden by the size of the markers in most cases).Arrows represent 3- σ upper-limits. per limit on the spectral index between 1.6 and 3.0 GHzof α (cid:46) − . (where S ν ∝ ν α ).FIRST J1419+3940 is significantly resolved in the ob-tained images given the size of the synthesized beam( . × . ), as the measured size is larger than theminimum resolvable size by the array (see Mart´ı-Vidalet al. 2012; Natarajan et al. 2017, for a detailed expla-nation). By fitting a circular Gaussian to the u v data wemeasure a source size of . ± . , where the uncer-tainty has been estimated through a χ test. However,we note that the gain calibration of Tianma constitutesa potential source of systematic errors in the size mea-surement. To provide a more reliable measurement, weproduced images without this station. Despite havingpoorer resolution (synthesized beam of × . ), weobtained a size that is significant and consistent withinuncertainties with the value quoted above ( . ± . ).Finally, we imaged the source with the gain correctionapplied to Tianma (as mentioned in the previous sec-tion), which provides the most stringent lower limit onthe source size (i.e. assuming a point-like source duringcalibration). In this case we measured a source size of . ± . . The contribution of the longest baselinesis therefore not critical as we obtain consistent resultsfrom all cases. We summarize the results of these differ-ent analyses in Table 1.For comparison, the phase calibrator, J1419+3821, ex-hibits a main compact component with a measured sizeof . – . in all cases. The fact that the measuredsizes are siginificantly different – while they are seenalong almost the same Galactic line of sight – meansthat they are most likely intrinsic sizes, rather than dueto scatter broadening. At the high Galactic latitudes of ∼ ◦ scatter broadening at GHz frequencies is almostnegligible, and one would not expect strong variationsof scattering size on small angular scales either (see e.g.Pushkarev & Kovalev 2015).We thus conclude that FIRST J1419+3940 is signifi-cantly resolved, with an angular size of . + . − . + . − . mas ,where the first uncertainties take into account the dis-persion of the values from the different analyses, and thesecond ones consider the estimated statistical uncertain-ties on the value. Given that the angular diameter dis-tance to the source is 83 Mpc (Law et al. 2018), we derivea projected physical size of . ± . . This size also im-plies a brightness temperature of T b ∼ . × K , whichclearly points to a non-thermal origin for the emission.Law et al. (2018) estimated that the putative GRBproducing the observed afterglow likely took placearound ∼ –
30 yr ago. Considering an estimated cen-tral date for the explosion of ∼ and taking intoaccount the given uncertainties, the afterglow must havea mean expansion velocity of v = ( . ± . ) × km s − ,or ( . ± . ) c , consistent with a mildly relativisticexpansion. We note that the calculated expansion ve-locity is an average over the whole lifetime, during whicha significant deceleration has likely occurred.3.2. On the single burst searches
We detected no astrophysical single pulses or periodicsignals in the high-time-resolution Effelsberg data. Wecan estimate the expected dispersion measure (DM) to-wards FIRST J1419+3940 using Galactic electron den-sity models (NE2001; Cordes & Lazio 2002, YMW16;Yao et al. 2017). For an extragalactic source, the ob-served DM can be divided into four components alongthe line of sight: DM obs = DM MW + DM MW halo + DM IGM + DM host . (1)The Milky Way contribution to the DM along the line ofsight is divided into the disk and spiral arm component, DM MW , and the Galactic halo component, DM MW halo .The former, DM MW , is and
39 pc cm − calculated us-ing the NE2001 and YMW16 models, respectively. Theuncertainties in these contributions are not well quan-tified, but are likely on the order of . Using this,we can derive an approximate range of: (cid:46) DM MW (cid:46)
50 pc cm − . We apply a Galactic halo contribution of ∼ –
100 pc cm − to the DM (Prochaska & Zheng 2019).Given that the redshift of FIRST J1419+3940 is . (Law et al. 2018), the mean intergalactic medium (IGM)contribution to the DM is DM IGM (cid:39)
20 pc cm − (Ioka2003; Inoue 2004). We assume that the DM contributionof the host galaxy of FIRST J1419+3940, DM host , is com-parable to that of the host galaxy of FRB 121102: (cid:46) Marcote et al.
Table 1.
Properties of FIRST J1419+3940 measured following different imaging approaches.rms Peak brightness Flux Density size synthesized beam( µ Jy beam − ) ( µ Jy beam − ) ( µ Jy ) (mas) (mas × mas, ◦ )Default calibration
19 300 620 ±
20 4 . ± . . × . , − ◦ Without Tianma
30 510 630 ±
30 3 . ± . × . , ◦ Corrected Tianma
25 459 620 ±
30 3 . ± . . × . , ◦ DM host (cid:46)
225 pc cm − (Tendulkar et al. 2017). Combin-ing all individual components using equation (1) resultsin the approximate range (cid:46) DM obs (cid:46)
400 pc cm − .From the single pulse candidates reported using sin-gle_pulse_search.py , an astrophysical burst would beidentifiable provided the signal-to-noise ratio exceeds ∼ . We can estimate the fluence limit of our searchusing F = ( S / N ) min T sys G (cid:115) W b n pol ∆ ν (2)(following Cordes & McLaughlin 2003), where ( S / N ) min is our detection threshold of , T sys is the system tem-perature, G is the telescope gain, n pol is the number ofrecorded polarizations, ∆ ν is the total bandwidth, and W b is the observed width of the burst. The observedwidth, W b , accounts for broadening of the intrinsic widthdue to the finite time sampling of the data, intra-channelsmearing, smearing due to DM-trial spacing, and scatterbroadening. FRB 121102 has been shown to exhibit in-dividual bursts with widths (cid:46) µ s (Michilli et al. 2018)and there have been observations of FRBs with widthsas large as ∼
30 ms (Petroff et al. 2016) . Taking a DMof
300 pc cm − and intrinsic widths µ s –
30 ms , we findour fluence limit ranges from . to . INTERPRETATION4.1.
Measured source size
In order to compare our measurements with the sce-narios proposed by Law et al. (2018), we need to com-pute the expected apparent size of the source. Atthe time of our observation, t obs ∼
30 yr ∼ d af-ter the initial explosion, the external shock producedby the GRB jet upon deceleration into the interstel-lar medium (ISM) is expected to be non-relativistic,and to have become essentially spherical. Detailed,long term numerical relativistic hydrodynamics simu-lations (Zhang & MacFadyen 2009) indeed show thatthe blast wave is well-described by the spherical, non- All published FRBs and their properties can be found in theFRB Catalogue: .
10 20 30 40 50 t obs − t (yr)123456 A pp a r e n t s i ze ( m a s ) t = Figure 4.
Apparent source size evolution. The red lineshows the predicted apparent size evolution for a jet withparameters as those proposed by Law et al. (2018). Theblue line and the lighter blue band show our measured ap-parent size and its 1- σ uncertainty of θ s = . ± . . Thegrey vertical line marks the source age at the time of ourobservation, assuming that it originally exploded in 1993. relativistic Sedov-Von Neumann-Taylor solution after atime t NR ≈ × E / , n − / d , where t NR is the light-crossing time of the Sedov length associated to the jet.The initial relativistic expansion phase, however, canstill have effects on the relation between observed timeand projected size. For that reason, we compute thejet deceleration dynamics and spreading employing the“trumpet” model from Granot & Piran (2012), whichhas been shown to be in good quantitative agreementwith results from numerical relativistic hydrodynamicssimulations. The observed size is estimated as the max-imum projected size of the equal-arrival-time surface,relativistic beaming of radiation being negligible in ourlate-time observations. Using the same jet parameters asLaw et al. (2018), namely an isotropic equivalent energy E iso = × erg , an ISM number density n =
10 cm − and a viewing angle θ v = . , and further assuming ajet half-opening angle θ j = . (which implies a totaljet energy E jet ∼ erg ), we obtain the size evolutionshown by the red solid line in Fig. 4, which is fully com-patible with the measured one, assuming that the GRBtook place in ∼ . This disfavors the alternative sce- esolving a decades-long radio transient source (cid:46) . ), due to the muchlower expansion velocity (Murase et al. 2016).4.2. Flux density
While the measured size agrees well with the GRB sce-nario proposed by Law et al. (2018), our measured fluxdensity S . = ± µ Jy is low when comparedto the extrapolation of their model. More precisely,adopting the same assumptions as Law et al. (2018),namely quasi-isotropic, adiabatic expansion, Deep New-tonian regime and an electron power law index p = . ,the flux density should follow S ν ∝ ν − . t − . . Using thelatest VLA detection as reference, which yielded a fluxdensity of . ± . at .
52 GHz on 2015 May 11,and assuming the GRB to have happened in 1993, weshould have measured S . ≈ ± µ Jy at the timeof our observation, which is ∼ . σ (summing the un-certainties in quadrature) above our measured flux den-sity. As noted by Law et al. (2018), the latest VLASSnon-detection S ν < µ Jy at on 2017 October11 already pointed to a faster decline after 2015. Sev-eral physical processes could lead to a steepening in thedecay of the lightcurve, e.g.: • The conditions in the shocked fluid could be chang-ing as a consequence of the transition to the non-relativistic, Deep Newtonian phase, e.g. the frac-tion (cid:15) e of shock energy given to electrons coulddecrease, or the electron momentum distributionpower law index p could decrease from p = . to-wards ∼ (Sironi & Giannios 2013). Both theseeffects would result in a steepening of the flux de-cay; • Contrary to what is stated by Law et al. (2018),the steepening could also be due to the shock cross-ing a dip in the ISM density. According to theargument by Law et al. (2018), based on Nakar &Granot (2007) and Mimica & Giannios (2011), anISM density drop would result only in a smooth,slow change in the lightcurve. This is essentiallya consequence of the assumption that the shockis relativistic ( Γ (cid:29) ), in which case the angulartime scale R / Γ c would be of the same order asthe observer time t obs , and therefore any change inthe shock conditions would be smeared out overthat time scale. In our case, conversely, the shockexpansion speed is non-relativistic, and the angu-lar time scale is ∼ R / c (cid:28) t obs (using our size mea-surement, we have R / c ∼ . , which is signifi-cantly smaller than the explosion age t obs (cid:38)
25 yr ),so that a drop in the ISM density at a radius slightly smaller than the observed size ∼ . may justify the flux deficit. Such a drop couldmark the outer radius of the star-forming regionwhere the GRB exploded. Let us caution, though,that this requires some fine-tuning. In order forthe shock to entirely cross the outer edge of thestar-forming region in a short enough time, the lat-ter should be approximately spherical and nearlyconcentric to the shock. By using the same shockdynamics model as in the previous section, we es-timate the current shock expansion velocity to be v s ∼ . c ≈ − . The shock thus trav-elled a distance ∆ R ∼ .
03 pc between the latestVLA detection at . and our observation.The centers of the shock wave and the star-formingregion, assumed spherical, should therefore be lo-cated less than ∆ R away from each other. Sincethe flux deficit we observe amounts to a factor ∼ / reduction with respect to the expected value,we can partially relax these requirements, allow-ing the outer edge of the star-forming region tohave some structure – e.g. bumps, filaments, a non-spherical shape – as long as ∼ / (in terms of solidangle) of the shock wave still experiences a sharpdensity drop in the required time. This leads us toconclude that, while the arguments against an am-bient medium density drop proposed by Law et al.(2018) do not hold in this case, this kind of expla-nation for the flux variation, while not impossible,remains rather unlikely.Finally, we note that the flux deficit cannot be under-stood as due to scintillation-induced fluctuations, as theapparent size of the source is too large (Goodman 1997).4.3. Comparison with FRB 121102
The association of FRB 121102 with a persistent ra-dio counterpart (Chatterjee et al. 2017; Marcote et al.2017) led to the discovery of FIRST J1419+3940, thecharacteristics of which match those of the persistentsource coincident with FRB 121102 (Ofek 2017; Lawet al. 2018): i.e., a compact radio source with a sim-ilar luminosity and co-located with a star-forming re-gion of a dwarf galaxy. It is, therefore, natural to com-pare FIRST J1419+3940 with the radio counterpart ofFRB 121102.The declining light curve of FIRST J1419+3940 con-trasts with the persistent emission from FRB 121102(Chatterjee et al. 2017, Plavin et al. in prep). Anotherdiscrepancy that arises is that the obtained source sizeof FIRST J1419+3940 is significantly larger than theone associated with FRB 121102 ( < . ; Marcoteet al. 2017), implying a much higher expansion veloc- Marcote et al. ity. These differences can, naturally, be explained bya younger age of FIRST J1419+3940 ( ∼
30 yr ) whencompared with FRB 121102 ( ∼
100 yr ; Metzger et al.2017; Piro & Gaensler 2018). We note that although weconclude our observations are consistent with GRB jetexpansion, we do not rule out the presence of a nebuladriven by a highly magnetized neutron star contribut-ing to a fraction of the radio emission observed. Thepresence of such nebula could cause the light curve toplateau at late-times.It has been hypothesized that the birth of a mil-lisecond magnetar can connect FRB 121102 with longGRBs or superluminous supernovae (SLSNe; Metzgeret al. 2017). If we assume that millisecond magne-tars can produce FRBs similar to what is observed inFRB 121102, and that a magnetized neutron star re-sides within the radio source FIRST J1419+3940, wemight expect FRBs from this source. The comparableages of FIRST J1419+3940 and FRB 121102 leads toour assumption that the compact object residing withinFIRST J1419+3940 is emitting bursts with a compara-ble energy distribution and duty cycle to FRB 121102.Bursts from the repeating FRB 121102 have beenobserved with fluences of ∼ .
02 Jy ms (Gajjar et al.2018) to (cid:38) (Marcote et al. 2017) and widthsranging from (cid:46) µ s (Michilli et al. 2018) to ∼ . (Spitler et al. 2016). Taking this range of fluencevalues at the luminosity distance of FRB 121102(
972 Mpc ) and scaling to the luminosity distance ofFIRST J1419+3940 (
87 Mpc ), gives an estimated flu-ence range of . –
870 Jy ms . For bursts with widths ex-ceeding ∼ , the fluence limit of our search increasesbeyond . (see Section 3.2). Under the assump-tion that FIRST J1419+3940 is producing bursts withwidths comparable to that of FRB 121102 and with analignment (with respect to the observer) consistent withFRB 121102, single bursts from this source would beidentifiable in the data.The lack of short-duration bursts in our observationcould imply that the source is in a quiescent state, sim-ilar to the behaviour observed in FRB 121102 (Scholzet al. 2016; Gajjar et al. 2018). Alternatively, the hy-pothesized central compact object could be producingbursts that do not cross our line-of-sight. To ensure oursearch was not affected by self-absorption, future obser-vations of FIRST J1419+3940 at higher radio frequen-cies are required.The potential connection of FRB 121102 with longGRBs or SLSNe has sparked targeted searches formillisecond-duration bursts and compact persistent ra-dio sources at the positions of such events. In one sucha search, Eftekhari et al. (2019) discovered a persistent radio source coincident with the SLSN PTF10hgi. Anorphan GRB afterglow is explored as the potential ori-gin of the emission, but is considered unlikely due tothe high inferred isotropic jet energy, exceeding thatof most observed long GRBs. Since the discovery ofboth this radio source and FIRST J1419+3940 weremotivated by observations of FRB 121102 and its envi-ronment, we compare the inferred isotropic jet energyfor both sources. The inferred properties of the ra-dio source associated with PTF10hgi, assuming GRBjet expansion, is estimated as E iso ∼ ( – ) × erg , n = − – cm − (Eftekhari et al. 2019). Although thisenergy range is larger than the majority of observed longGRBs, it is comparable to that of FIRST J1419+3940( E iso = × erg , n =
10 cm − ; Law et al. 2018).The results shown in the work presented here supportthe scenario in which FIRST J1419+3940 is an orphanGRB afterglow. Whether this is the case for PTF10hgias well is not clear at this point, but we argue thatthe inferred high isotropic jet energy in itself does notexclude an off-axis jet origin. The ultimate probe ofthis scenario is very high angular resolution VLBI ob-servations. Accurately measuring the source size at theredshift of PTF10hgi (about five times more distantthan FIRST J1419+3940) – and especially consideringits low flux density of ∼ µ Jy – is very challenging,and may only be possible with a very sensitive futureSKA-VLBI array (Paragi et al. 2015) observing at highfrequencies ( (cid:38) ). CONCLUSIONSFIRST J1419+3940 was reported as a slowly fadingradio transient source. We provide the first constraintson the source size, using EVN data. These measure-ments confirm the non-thermal emission of the sourceand are consistent with jet expansion from a putative or-phan long GRB. The derived average expansion velocityis consistent with a mildly relativistic expansion, notingthat a significant deceleration has likely happened dur-ing these ∼
30 yr after the event. A flux density lowerthan expected is reported, suggesting a faster decline af-ter 2015. This decay could be explained by a change inthe post-shock microphysical parameters following thetransition to the non-relativistic phase, or by a dropin the ISM density (e.g. due to the shock reaching theouter edge of the star-forming region where the GRB ex-ploded). We exclude scintillation-induced fluctuationsas the origin of the reported variability.Finally, although FIRST J1419+3940 was discoveredin a search for persistent radio sources similar to thatassociated with FRB 121102, we note significant differ-ences between these sources (e.g. FIRST J1419+3940 esolving a decades-long radio transient source
Abbott, B. P., Abbott, R., Abbott, T. D., et al. 2017, ApJ,848, L12Astropy Collaboration 2013, A&A, 558, A33Bassa, C. G., Tendulkar, S. P., Adams, E. A. K., et al.2017, ApJL, 843, L8Beasley, A. J., Gordon, D., Peck, A. B., et al. 2002, ApJS,141, 13Chatterjee, S., Law, C. J., Wharton, R. S., et al. 2017,Nature, 541, 58Cordes, J. M. & Lazio, T. J. W. 2002, arXiv e-prints, arXiv:1901.04525
Cordes, J. M. & McLaughlin, M. A. 2003, ApJ, 596, 1142Eftekhari, T., Berger, E., Margalit, B., et al. 2019, arXive-prints, arXiv:1901.10479