Incoherent transient radio emission from stellar-mass compact objects in the SKA era 5 1 0 2 S. Corbel∗1†, J.C.A.Miller-Jones2‡, R.P.Fender3,E. Gallo4, T.J.Maccarone5, T.J. n O’Brien6, Z.Paragi7, M.P. Rupen8,A.P. Rushton3,S.Sabatini9, G.R.Sivakoff10, J. a Strader11 andP.A.Woudt12 J 1UniversityParisDiderot&CEASaclay,2ICRAR,CurtinUniversity,3OxfordUniversity, 1 2 4UniversityofMichigan,5TexasTechUniversity,6UniversityofManchester,7JIVE,8NRAO, 9INAF,10UniversityofAlberta,11MichiganStateUniversity,12UniversityofCapeTown ] E E-mail: [email protected],[email protected] H Theuniversallinkbetweentheprocessesofaccretionandejectionleadstotheformationofjets . h andoutflowsaroundaccretingcompactobjects. Incoherentsynchrotronemissionfromtheseout- p flows can be observed from a wide range of accreting binaries, including black holes, neutron - o stars, and white dwarfs. Monitoring the evolution of the radio emission during their sporadic r outburstsprovidesimportantinsightsintothelaunchingofjets, and,whencoupledwiththebe- t s haviourofthesourceatshorterwavelengths,probestheunderlyingconnectionwiththeaccretion a [ process.Radioobservationscanalsoprobetheimpactofjets/outflows(includingotherexplosive eventssuchasmagnetargiantflares)ontheambientmedium,quantifyingtheirkineticfeedback. 2 v The high sensitivity of the SKA will open up new parameter space, enabling the monitoring 6 of accreting stellar-mass compact objects from their bright, Eddington-limited outburst states 1 down to the lowest-luminosity quiescentlevels, whose intrinsic faintness has to date precluded 7 4 detailedstudies.Acensusofquiescentlyaccretingblackholeswillalsoconstrainbinaryevolution 0 processes.Byenablingustoextendourexistinginvestigationsofblackholejetstothefainterjets . 1 fromneutronstarandwhitedwarfsystems,theSKAwillpermitcomparativestudiestodetermine 0 theroleofthecompactobjectinjetformation.Thehighsensitivity,widefieldofviewandmulti- 5 beaming capability of the SKA will enable the detection and monitoring of all bright flaring 1 : transientsin theobservablelocalUniverse(within∼ 15Mpc), includingthe radiocounterparts v i ofultraluminousX-raysources,improvingourunderstandingofaccretionandjetejectionatthe X highestrates,withimportantimplicationsforthegrowthofthefirstquasars. r a As synchrotron events peak earlier at higher frequencies, and with higher flux densities, such studies will be best enabled by SKA1-MID, in the higher-frequencybands 4 and 5. With the highsensitivityavailablefromSKA1-MID,wewillalsobeabletoprobeisolatedquiescentblack holes undergoingBondi-Hoyle accretion from the nearby environment, both stellar-mass black holesinthefieldandtheputativepopulationofintermediateblackholesinglobularclusters.This chapterreviewsthesciencegoalsoutlinedabove,demonstratingtheprogressthatwillbemadeby theSKA instudyingincoherentsynchrotronemissionfromaccretingcompactobjects. We also discussthepotentialoftheastrometricandimagingobservationsthatwouldbepossibleshoulda significantVLBIcomponentbeincludedinphase1(andeventuallyphase2)oftheSKA. AdvancingAstrophysicswiththeSquareKilometreArray June8-13,2014 GiardiniNaxos,Italy (cid:13)c Copyrightownedbytheauthor(s)underthetermsoftheCreativeCommonsAttribution-NonCommercial-ShareAlikeLicence. http://pos.sissa.it/ Incoherenttransients S.Corbel 1. Introduction andcontext Stellar-masscompactobjectsprovideimportantlaboratoriesforstudyingthefundamentalcou- pling between the accretion process and the launching of energetic outflows, which often take the form of highly-collimated jets (Hughes 1991). A theoretical picture has been developed where the jets are composed of an electron/positron or electron/proton plasma (e.g. Bonometto & Rees 1971), whichismagnetically collimated asitflowsawayfromthecompactobject. Thesejetsmay bepoweredeitherbytappingtheenergyofarotatingblackhole(Blandford&Znajek1977),orby extractingenergyfromtheaccretionflow(Blandford&Payne1982). However,manybasicaspects of jet physics are uncertain, including their composition and structure, as well as the mechanisms that power and collimate them. Relativistic jets are relevant in almost all fields of astrophysics, and in some cases may be the dominant output channel for the accretion power from black holes (Fender et al. 2003). They provide an important source of feedback to the surrounding environ- ment, being able to either trigger or suppress star formation, accelerate cosmic rays, and seed the surrounding medium with magnetic fields. They have even been suggested to play a role in the reionisation oftheUniverse(e.g.Mirabeletal.2011). Despite their relative proximity, the lower masses of stellar-mass compact objects imply that theyareobservedatlowerangularresolution(intermsofgravitationalradii)thannearbyAGN,yet they evolve through their duty cycles on human timescales, typically undergoing entire outbursts over periods of days to months. Thus, they provide unique insights into the coupling between accretion andoutflow. Furthermore, comparative studies ofthedifferentclasses ofcompactobject can provide important insights into the necessary and sufficient ingredients for the jet launching processes. The non-thermal radio emission from stellar-mass compact objects typically arises via syn- chrotron emission from relativistic particles spiraling around the magnetic field lines of the jets. Whilesignificantprogresshasbeenmadeoverthepastfewdecadesinunderstanding thenatureof thejetsandtheircouplingtotheaccretionflow(seeFender2006,forareview),investigations have beenhamperedbythelimitedsensitivityofpastandcurrentfacilities (seealsoFenderetal.2014). Wedetailinthischapter howthesensitivity improvementbrought aboutbytheSKAwillallowus to study black hole outbursts throughout the Galaxy and out into the Local Group, determine the role of jetsin thelow-luminosity quiescent state, and extend ourexisting studies of black holes to theanalogous yetfainter neutronstarandwhitedwarfsystems. 2. Galacticblack holes Fromanobservationalstandpoint,blackholetransientsspendmostoftheirtimeinaquiescent state, at very low mass accretion rates. They occasionally undergo outbursts that last from a few months to ∼ a year, during which the flux rises by several orders of magnitude across the whole electromagnetic spectrum (McClintock & Remillard 2006). These outbursts are associated with globalchangesinthreemaincomponents: thejets,theaccretiondiskandthecorona. Theluminous ∗Speaker. †Co-chairofthechapter. ‡Co-chairofthechapter. 2 Incoherenttransients S.Corbel Figure1: RadiolightcurveofCI Cam(a B[e]X-raytransient, alsocalledXTEJ0421+560)followingits discovery. Asforallincoherentsynchrotrontransients,thelower-frequencyradioemissionpeakslaterand with lower amplitude (Figure from Fender et al. 2008). Depending on the rate at which a source evolves throughitsX-rayspectralstates,suchradioflarestypicallyoccurafewdaystomonthsfollowingthedetec- tionoftheX-rayoutburst. outburst phase, with a luminosity > 10–30% of the Eddington luminosity (the “soft state”), is dominated by thermal emission from the accretion disk. During the rise and decay phases of the outburst (the “hard” state), the bolometric luminosity of the source is dominated by non-thermal emission(synchrotronorinverseComptonemissionfromeitherthejetsorthecorona)extendingup tothehardX-rayband. Followingrecentresultswithhigh-resolutionX-rayspectroscopy,accretion disk winds are now recognized as ubiquitous in black hole X-ray binaries, and may carry away a significant fraction of the inflowing energy (Miller et al. 2006; Neilsen & Lee 2009; Ponti et al. 2012;DíazTrigoetal.2013). Observations of typical Galactic black holes have indicated twoforms of jets associated with theseprimaryaccretion states;theslowly-varying, partially self-absorbed compactjets(withradio emission usually ≤ 30-50 mJy for distances of a few kpc, and a flat or slightly inverted radio spectrum) observed in the hard state (Corbel et al. 2000; Fender et al. 2000; Dhawan et al. 2000; Stirling et al. 2001), and the bright (0.1–10 Jy, with an optically thin radio spectrum), strongly variable transient jets (occasionally showing apparent superluminal motion) detected during the transition from the hard to the soft state (Corbel et al. 2004; Fender et al. 2004b). The core radio emission is then strongly quenched during the soft state (Fender et al. 1999; Coriat et al. 2011). Relic radio emission can also be detected in some cases when the jets interact with the ambient medium,eitheraslarge-scale lobes(Galloetal.2005),orasfaint,transienthotspots(Corbeletal. 2002), depending onthedutycycleofthecentralblackhole. Asischaracteristicofexpandingsynchrotronsources,thebrightradioflaresfromthetransient jetsdisplayacharacteristicfrequencydependenceofthetransition(seeFig.1)fromoptically-thick 3 Incoherenttransients S.Corbel to optically-thin emission, which is consistent with an expanding plasmon model such as the one described by van der Laan (1966). Thus, the low-frequency radio emission (<1GHz) will peak later (usually by a few days to weeks) and at lower fluxes (by at least an order of magnitude) compared to theemission athigher frequencies (>3GHz). Tostudy thetransient radio emission, it is therefore of the utmost importance for the initial deployment of SKA1-MID to include band 4 (2.80-5.18 GHz) or (preferably) band 5 (4.6-13.8 GHz). This is also important to constrain the energy budgetofthecompactjetsseeninthehardstate,whichcanbestrongly affectedbyopacity effects at lower frequencies. Bands 4 and 5 have higher intrinsic sensitivity than bands 1–3 (see Table 7ofthe SKA1System Baseline Design Document; band 1=0.35-1.05 GHz, band 2=0.95- 1.76 GHz, band 3 = 1.65-3.05 GHz), as well as lower confusion limits and higher intrinsic levels of radio emission at the peak of an outburst (see Fig. 1 or any textbook example of a synchrotron flare;e.g.vanderLaan1966). Thisallowsthemtoprobealargerobservingvolumethanthelower frequencybands,byfactorsofbetween6(comparingband4toband3)and200(comparingband4 toband1),thereby extending themaximumdistance wecanprobebyfactorsof2–6. Thebrighter peakemissioninband5wouldincrease thesenumbersstillfurther. There is astrong correlation between the radio and X-ray emission in the hard and quiescent states (Corbeletal. 2003, 2013b; Galloetal. 2003, 2012), indicating apossible coupling between thelaunchofthejetsandthedynamicsoftheflowclosetotheaccretingblackhole(Fig. 2). Itcan beusedtodeterminetheexpectedsourcebehaviour atthelowestradioluminosities, whichremain relatively poorly explored owing to the limited sensitivity of current instruments (Miller-Jones etal.2011). Itis,forexample,notclearwhetherquiescentblackholesdohostjetsasinthebrighter states. Currentobservationsseemtoindicateweakjetactivityeveninquiescence(Galloetal.2006, 2014), butitisnotknownwhatfraction oftheliberated accretion poweriscarriedawayinthejets rather than being advected across the black hole event horizon (Fender et al. 2003). Furthermore, understanding therecently-discovered dichotomyintheradio/X-raycorrelation(Coriatetal.2011; Galloetal.2012)couldprovideanewwaytoexplorethepossibilityofdifferentcouplingsbetween accretion and ejection in black hole transients. With a sensitivity as good as 30 m Jy s−1/2 in band 4/5 (Dewdney et al. 2013), SKA1-MID is the ideal instrument to monitor all detected black hole outbursts (up to a few tens per year) all the way down to quiescence. Furthermore, a long observation(10hrs)withSKA1-MIDcoulddetectallquiescentblackholebinariesuptoadistance of5kpc. Thesetransient blackholes could eitherbediscovered bytargetted monitoring orablind radiosurveywithSKA(Fenderetal.2014),orbyanyactivemulti-wavelength monitoringfacility. The brightest phases of black hole outbursts (the hard to soft state transitions) can easily be studied via snapshot observations, even for the most distant systems in our Galaxy, meaning that SKA(eveninphase1)willprobeasignificant fraction oftheoutburst activity foralmostallblack holes in our Galaxy, including the very faint, sub-Eddington transients, which have rarely been observed at radio frequencies. With a resolution on the order of a few tenths of an arcsecond – if SKA1-MID is equipped with band 4 or 5 receivers – the bright radio emission associated with the transient jets (Mirabel & Rodriguez 1994) could also be resolved from a week after the onset of the radio flare, as typical proper motions are on the order of 15–20 masday−1 (see Section 5 for the additional benefits of VLBI). Good polarization measurements (linear and circular) from SKA1-MID, associated with good signal purity, will provide key probes of the composition and geometry ofthejets,andthestructure oftheirmagneticfields. 4 Incoherenttransients S.Corbel Figure 2: Radio and X-ray (1-10 keV) luminosities for Galactic accreting binary black holes in the hard and quiescent states (see arrows at bottom for approximate luminosity levels). It illustrates the standard correlation (defined by sourcessuch as GX 339−4or V404 Cyg with index ∼ 0.6; green points) and the new correlation for the so-called “outliers” (defined by e.g. H 1743−322 or Swift J1753.5−0127 with index ∼ 1.4; a significant fractionof the red points). Black triangles representneutronstar binaries. The solid line illustrates the fit to the whole 1997–2012 sample of GX 339−4 (Corbel et al. 2013b) with an extrapolationtothequiescentstate. ThedottedlinecorrespondstothefittothedataforH1743−322,oneof therepresentativesfortheoutliers(Coriatetal.2011). Upperlimitsareplottedatthe3s confidencelevel. ThehorizontaldashedlinesrepresentsomesensitivitylevelsforSKA1-MIDasdiscussedinthetext. Thehighsensitivityandabilitytodumpthevisibilitiesonrapid(.1s)timescaleswouldallow us to probe any short-timescale variability of the jets, should it be present (shortest timescales of ∼ 10 minutes up to now, e.g. Corbel et al. (2000); Middleton et al. (2013)). Casella et al. (2010) detected rapid (sub-second) infrared variability from the jets of GX 339−4, which was correlated with the observed X-ray variability. Observing these variations at multiple frequencies as they propagate downstream in the jets would allow us to directly measure the speeds and sizes of the compactjets,providing thebestobservational constraints ontheirLorentzfactors. An X-ray binary by nature includes several emission components (star, compact object, ac- cretion disk, corona, jets, andnearbyenvironment), allemitting overawideenergy range(thejets themselves can be observed to emit up to the gamma-ray band), meaning that SKA observations would greatly benefit from coordinated and simultaneous observations with multi-wavelength fa- cilities (e.g. Rodriguez etal. 2008; Corbel etal. 2013a). Someobvious new facilities fortransient black holes in the next decade include (but are not limited to) X-ray observatories (e.g. SVOM, ASTROSAT),ground-basedall-skyopticaltelescopessuchasLSST,orasensitivenewhigh-energy gamma-ray observatory (CTA). Furthermore, all SKA1 components have the potential to act as radio “all-sky monitors” (if commensal transient search capability is implemented in the early 5 Incoherenttransients S.Corbel phases; see Fender et al. 2014). This would be of prime importance, allowing SKA to provide alerts to multi-wavelength facilities. Such a radio sky monitor could be crucial for detecting new outbursts shouldnoX-rayall-skymonitorbeavailable inthecomingdecades. Withareducedsensitivityof50%duringthedeploymentofSKA1-MID,asignificantfraction of the above scientific goals could still be achieved as long as band 4 or 5 is included in the early phase. If we consider the improved performance of SKA2, with broadband frequency access up to 24 GHz, then it is likely that almost all transient black holes located up to the distance of the GalacticCentrecouldbestudied ingreatdetail,asoutlinedabove. 3. Neutron starbinaries Comparisonsbetweenaccretingneutronstarsandblackholesareidealfordetermining which effects seen fromaccreting black holesarefundamentally related tothepresence ofaneventhori- zon, instead of being generic to accretion onto objects with deep gravitational potential wells. Largely speaking, the phenomenology of accretion onto neutron stars is similar to that onto black holes (e.g. Psaltis 2006). Many of the known differences can be explained in a straightforward mannerbythepresence ofasolidsurfaceforneutron starsandthelackofoneforblackholes. Studies of accreting neutron stars will be bolstered by the enhanced sensitivity of the SKA. Systematic studies of the radio luminosities of accreting neutron stars asa function oftheir X-ray luminosities are far sparser than those for black holes. This is in part because the characteristic timescales on which accretion disks change scales inversely with the accretor mass, so that “typi- cal”outbursts ofaccreting neutronstarsareshorterthanthoseforaccretingblackholes;andpartly because both the peak X-ray luminosities of neutron star transients and theratio ofradio toX-ray fluxforneutronstarsarebothlowerthanforblackholes(Fender&Kuulkers2001;Wuetal.2010; Migliari & Fender 2006). As a result of these factors, most soft X-ray transients with neutron star accretors peak at flux densities of about 1 mJy or less (for typical distances of several kpc). Until recently it was not feasible to probe flux densities significantly fainter than 100 m Jy, and even with the upgraded VLA, the sensitivity of short target-of-opportunity (ToO) observations is limited to a few m Jy. Thus, it has traditionally been difficult to span more than a factor of 10 in radio luminosity for neutron stars, and even now it is nearly impossible to span much more than a factor of 100. Furthermore, since the outbursts progress quickly, scheduling a large number of epochs places strong pressure on the ToO scheduling of existing arrays. Finally, since many of these sources are located in the Galactic Bulge, or elsewhere in the Southern Hemisphere, often onlyshortobservationsarepossible. Despitetheseproblems,someprogresshasstillbeenmadeon understanding theradioemissionofneutronstarX-raybinaries. Afewthingsseemclear: thehigh magnetic fieldX-raypulsars arenotstrong radioemitters (Migliari &Fender2006); theradio and X-ray fluxes are correlated (see Fig. 2)for neutron star X-raybinaries that emitat less than about 10% of the Eddington luminosity (Migliari et al. 2003; Tudose et al. 2009); and the reduction of theradio powerofneutron stars inaccretion states dominated bythermal emission isless extreme thanthesameturn-downforblackholes(Migliarietal.2004). Theenhanced sensitivity oftheSKAshould allow morequantitative statements onthese top- ics. Forexample, Migliari etal.(2003) findthatL (cid:181) L1.4 fortheneutron star4U1728-34 inhard R X 6 Incoherenttransients S.Corbel spectral states, while Tudose et al. (2009) looked at Aql X-1 and found L (cid:181) L0.4, albeit while R X including datafromarangeofX-rayspectralstates. ThefindingofMigliarietal.(2003)madefor nice agreement with a theoretical picture in which the bulk power put into the jet scales with the mass accretion rate, and the radio power of the jet scales with its bulk power to the 1.4, and also bolsters the suggestion that most hard states of black hole X-ray binaries are in radiatively ineffi- cientstateswhereL (cid:181) m˙2. Developingasampleofsourcesthathasbothmanyobjects,andwhich X spans a few orders of magnitude in radio luminosity would bring the data quality into line with that from the black hole transients at the present time (Fig. 2). It also opens the door to making robust determinations of whether the spin period of a neutron star affects the normalization of its radio/X-ray relation (Migliarietal.2011). Additional studies of the thermal X-ray states of neutron stars in the radio may also hold important clues to understanding how jets are launched. In these systems, by analogy with the blackholes,theaccretiondisksthemselvesareunlikelytosupplymuchpowertoanaccretionflow. Ontheotherhand,theboundarylayers,wheretheneutronstar’saccretiondiskdissipatesitsexcess rotational energy, should have the large scale height thought to be needed to power jets, and may interact with the magnetic field of the neutron star itself. Studies of a large sample of these soft states rather than merely the two that have been detected already (Migliari et al. 2004), may help ustounderstand jetproduction inthisenvironment. A final class of systems that can probe the relationship between the accretion flow and the launching of jets are the recently-discovered class oftransitional binary pulsars, which switch be- tween accretion-powered and rotation-powered states on timescales of just weeks (e.g. Archibald et al. 2009; Papitto et al. 2013; Bassa et al. 2014). These systems have been shown to emit flat or inverted-spectrum radio emission in their accretion-powered states, which is consistent with partially-self absorbed synchrotron jets (Papitto et al. 2013; Bassa et al. 2014). Intensive radio monitoring ofthetransitions between thesetwostates wouldprovide unique insights into howthe jetsareformedanddestroyed, and,together withsimultaneous X-raymonitoring, wouldhighlight theconnection betweenthejetsandtheevolving accretion flow. A second point, unrelated to the disk-jet connection, which can be addressed with neutron star observations is whether particles can be launched in a jet at speeds significantly greater than the escape speed of the accretor. To date, one example has been identified, Cir X-1 (Fender et al. 2004a). In this system an apparent speed of the jet of 15c has been inferred from the appearance ofradioemissionfarfromtheneutronstaritselfrelativelysoonafterX-rayflaringepisodes. Since the escape speed from a neutron star is typically ∼0.3c this would imply that the bulk jet speed mustbelargerthantheescapespeedforaneutronstarbyalargefactor;however,whathasnotyet been seen, due to lack of sensitivity and angular resolution with existing Southern radio facilities, is actual motion of the radio-emitting structures at an apparent speed larger than the speed of light. As distances and inclination angles for many neutron star X-ray binaries will be directly measurable with the SKA, the combination of two-sided proper motions and detections of the counterjets should allow for a clean test of whether the pattern speeds and the bulk motions of jets are equal. If e.g. the jets are dominated by Poynting flux close to the neutron star, then it would not be surprising for them to appear to move very close to the speed of light, even if the particles energized bythejetmoveatmuchslowerspeeds. Suchmeasurements represent aunique opportunity totestamajorhypothesisfromtheoreticalworkthatcannotbetestedwithblackholes, 7 Incoherenttransients S.Corbel wherethejetspeedsareexpectedtobeclosetothespeedoflight. 4. Cataclysmicvariables andrelated objects The occurrence of jets and fast collimated outflows is by no means restricted to accreting black holes orneutron stars. Through sensitive andtimely observations inthelast decade, numer- ousaccreting whitedwarfs(incataclysmic variables, symbiotic starsandsupersoft X-raysources) have shown strong evidence for jets and jet-like shocked, collimated outflows, observed at radio frequencies and interpreted as synchrotron emission. However, the brightness temperatures of the radio emission from cataclysmic variables are small enough that thermal and/or cyclotron emis- sion are still viable possibilities, and given the multiple possible mechanisms, there may be some heterogeneity in emission mechanisms. Better spectral measurements, and searches for circular polarization, whichshouldbepossible withSKA,canhelpresolvetheseissues. Perhaps the most striking (and encouraging) aspect is that these transient radio jets have all occurred in the prototypes of subclasses of accreting white dwarfs – the symbiotic star Z And (Brocksopp et al. 2004), the recurrent nova1 RS Oph (Rupen et al. 2008) and the dwarf nova SS Cyg (Körding et al. 2008), suggesting they are more common in white dwarf accretors than previously assumed. Persistent jets arealsothought tooccurinnova-like variables (Körding etal. 2011); these are systems in a regime of steady high mass transfer rate (∼10−9−10−8 M yr−1) ⊙ where accretion onto the white dwarf occurs via a standard thin a -disc (see e.g. Potter & Balbus 2014). In close analogy to transient jets in X-ray binaries, the non-magnetic dwarf nova SS Cyg repeatedly exhibitsradiooutburstsassociatedwithitsdiscinstabilityoutburstcycle(Kördingetal. 2008). Theoutbursts indwarfnovaearedescribedbyathermal-viscous instability intheaccretion disc, leading to a brief period (days to weeks) of enhanced mass transfer (∼10−8 M yr−1) onto ⊙ the white dwarf. Even though radio luminosities are low compared to X-ray binaries – peak flux densities of ∼ 1 mJy at 1–10 GHz for a system at 114 pc (SS Cyg; Miller-Jones et al. 2013) – these systems provide an important link in understanding how accretion is coupled to the outflow of matter across arange ofcompact accretors. Withthe expected sensitivities achieved by SKA1- MID we can extend the sample of dwarf novae observed at radio frequencies out to kiloparsec distances, where optical transient surveys such asCRTSand iPTF(and LSSTintheSKAera) are finding thousands of new dwarf novae; see Drake et al. (2014) for an overview of dwarf novae in CRTS. The sheer numbers of systems, the accessible time scales of the disc instability cycle in cataclysmic variables (weeks to months), and the reasonably well-understood accretion discs around white dwarfs provide an excellent laboratory for the SKA to study accretion physics with targeted (andtarget-of-opportunity) observations. Whilst thermal emission is the dominant component of radio emission in novae (Seaquist & Bode 2008; Chomiuk et al. 2014b), a significant number of novae exhibit non-thermal (syn- chrotron)emissionassociatedwithcollimatedbipolarandjet-likeoutflows(e.g.,RSOph: O’Brien et al. 2006; Rupen et al. 2008; V445 Pup: Woudt et al. 2009; V959 Mon: Chomiuk et al. 2014a). 1For more on classical and recurrent novae and the SKA, see also the chapter on thermal emission from novae (Chomiuketal.2014b). 8 Incoherenttransients S.Corbel As recurrent novae are prime candidates for the progenitors of type Ia supernovae, questions sur- rounding thenatureandenergetics oftheoutflowofmaterialduringanovaoutburst areatthecore of the debate as to whether a white dwarf grows in mass during successive nova cycles. Regular, multi-frequency monitoring of Galactic novae with the VLA (an SKA1-MID pathfinder in terms of sensitivity) and e-MERLIN(apathfinder interms ofangular resolution) isongoing and willbe complemented by deep observations with MeerKAT ahead ofSKA1-MID.Atthe moment weare likelyonlypickingupthetipoftheicebergintermsofsynchrotron-emitting novae,i.e.thosewith significant circumstellar medium (e.g. from red giant secondary winds) to provide strong shock interactions, or those sufficiently nearby. A representative census of Galactic novae observed at thesensitivity ofSKA1-MID(including band4or5)isrequired todetermine whatfraction ofno- vae show evidence for synchrotron emission and to fully understand the processes that lead tothe formationofcollimatedjetsfollowingathermonuclear runawayonthesurfaceofthewhitedwarf. 5. Imagingand astrometry withVLBI High-cadence, high-resolution imagingusingSKA-VLBI(Paragietal.2014)duringrareout- burst events can provide a wealth of information about the jets in stellar-mass compact objects, and their coupling to the accretion process. Daily imaging during and after state transitions can determine thepropermotions(andanyobserved deceleration) oftheapproaching andreceding jet ejecta, allowing ustotrackthem backtozero-separation, andhence determine theexacttimingof theejection event. Bycomparison withX-rayspectral andtimingsignatures, thisenables ustoes- tablish thecausal connection betweenchanges intheaccretion flowandtheejection ofrelativistic jets. Since structural changes in the accretion flow occur on timescales of hours to days, such in- formation cannot begleaned fromradiolightcurvesalone, asbothopacity effectsandtimedelays between ejection and shock formation cause a lag between an ejection event and the associated peakintheintegrated radiolightcurve. From such observations, we can determine the product of the jet speed and inclination an- gle, b cosq , which can then be decoupled if the distance can be well constrained. Should we detect deceleration, wecanconstrain boththepowerofthejetsandthedensity ofthesurrounding medium. Spatially resolving the polarisation of the jets can provide information on the magnetic fieldstructure, helpingtodistinguish shocksfromsteadyflow. Polarizationobservations ofstellar- masssynchrotronsourcesonmasscaleshavebeenveryrarebecausetheseobjectsareusuallyvery faint (∼mJyorbelow), and sensitive VLBInetworks havenotbeen easytoarrange atshort notice during outbursts (but see Tudose et al. 2007). A flexible SKA-VLBI array will therefore play an importantrolehere,evenduringtheinitialdeploymentphaseofSKA1-MID(cf.Paragietal.2014). Todate,intensiveVLBImonitoringhasonlybeencarriedoutforasmallnumberofoutbursts (e.g. Yang et al. 2011; Miller-Jones et al. 2012a; Paragi et al. 2013). However, the peculiarities of the individual sources, together with scheduling considerations and sensitivity limitations have so far precludedtheidentificationofwhichchangesintheaccretionflowgiverisetothemajorjetevents. Thegreatsensitivityandresolving powerofSKA-VLBIwillalsobeinvaluable forthepartic- ularlyinterestinglower-massneutronstarandwhitedwarfsystems,whichshowsimilarpatternsof behaviour (Maitra & Bailyn 2004; Körding et al. 2008) but are typically significantly fainter than blackholes. Thelowerradioluminositiesofthesetwoclassesofsystemhavemeantthattheyhave 9 Incoherenttransients S.Corbel not been as intensively studied to date. Comparative studies of accretion-ejection coupling across different classes of compact object can provide important insights into the process of jet forma- tion and collimation, determining how the jet properties depend on the depth of the gravitational potential well,andthepresence orabsence ofastellarsurface andmagneticfield. Thehigh-precision astrometryenabledbySKA-VLBIcanalsoplayanimportantroleincom- pactobjectstudies. Compact,partiallyself-absorbedsynchrotronjetsshowdistinctstructuralprop- erties on (sub-)mas scales (see Paragi et al. 2013, and references therein). While flat-spectrum compactjetsaretypicallyobserved inthehardstateofblackholebinaries (seeSection2),thefirst (and only well-established) observation of the expected frequency-dependent position shift in the jet peak-brightness distribution (the “core shift”) – due to the different optical depths at different frequencies–wasinSS433(Paragietal.1999). Sincethesynchrotronopticaldepthdependsonjet parameters (e.g. magnetic fieldstrength andrelativistic particle density), thejetpeakposition will alsochangeasthesourcegetsbrighterorfainter(seeRushtonetal.2012;Paragietal.2013). SKA- VLBI will be sufficiently sensitive to reliably measure core shifts for a large number of Galactic blackholebinaries,giventheastrometricaccuracyofafewm arcsecforsourcesasfaintas∼1mJy. This will provide information on the structural properties of the jets, but also has the potential to constrain theaccretion physics, ashasrecently beenshownforAGN(Zamaninasabetal.2014). SKA-VLBIastrometrywouldbeimportantformeasuringaccurateparallaxesofGalactictran- sientsources. DistancestoX-raybinariesaretypicallyuncertainbyafactorof∼2(Jonker&Nele- mans 2004), rendering relatively uncertain our measurements of luminosities, jet speeds and even black hole spin. As faint, persistent, yet unresolved radio sources, X-ray binaries detected in the hardorquiescent statesmakeidealastrometrictargets(e.g.Bradshawetal.1999;Reidetal.2011; Miller-Jones et al. 2009). The high astrometric accuracy feasible with VLBI currently enables the measurement of atrigonometric parallax distance out to several kpc. While Gaiawill provide accurate distances for relatively bright quiescent systems that are not highly absorbed, VLBI will remain the only feasible alternative for distance determinations for optically-faint systems in the Galactic Plane, andthe only meanstoaccurately measure distances for transients discovered after theendoftheGaiamissionlifetime. Even for systems that are either too distant or too faint for parallax measurements, proper motions canbemeasured overatimebaseline asshort asthe few-monthduration ofanindividual outburst. When combined with the line-of-sight radial velocity (from optical or infrared obser- vations) and an estimate of the source distance, the full three-dimensional space velocity of the system can be determined (e.g. Mirabel et al. 2001). This can shed light on whether it received a natalkickinasupernova,andhenceplaceobservationalconstraintsontheformationofblackholes andneutron stars. Subtracting themeasuredparallax andproper motionsignatures fromastromet- ric observations of a given system can provide additional constraints on the physical dimensions of the binary. With sufficient astrometric accuracy, we can measure the orbital signature of high- mass binary systems, and the size scale of the unresolved jets, from the scatter of the astrometric residuals perpendicular andparalleltothejetaxis(seeMiller-Jones 2014formoredetails). 6. Isolatedblackholes There are likely to be ∼108 stellar-mass black holes in our Galaxy (Samland 1998), which, 10