The multi-frequency parsec-scale structure of PKS 2254-367 (IC 1459): a luminosity-dependent break in morphology for the precursors of radio galaxies?

0.24 MB·English

by  S. J. Tingay

Mon.Not.R.Astron.Soc.000,1–7(2014) Printed20January2015 (MNLATEXstylefilev2.2)
The multi-frequency parsec-scale structure of PKS
2254−367 (IC 1459): a luminosity-dependent break in
morphology for the precursors of radio galaxies?
5
1
0 S.J. Tingay1⋆ and P.G. Edwards2†
2 1International Centre for Radio Astronomy Research, Curtin University,Bentley, WA 6102, Australia
n 2Australia Telescope National Facility, CSIRO Astronomy and Space Science, Epping, NSW 1710, Australia
a
J

9 Submitted: 2014
1

] ABSTRACT
A
We present the first multi-frequency VLBI images of PKS 2254−367, a Giga-
G hertz-Peaked Spectrum (GPS) radio source hosted by the nearby galaxy IC1459
. (D=20.5Mpc). PKS 2254−367 and the radio source in NGC1052 (PKS 0238−084;
h D=17.2Mpc) are the two closest GPS radio sources to us, far closer than the next
p closestexample, PKS1718−649(D =59Mpc). As such, IC 1459 andNGC1052offer
-
o opportunities to study the details of the pc-scale radio sources as well as the environ-
r mentsthattheradiosourcesinhabit,acrosstheelectromagneticspectrum.Giventhat
t
s somemodelsfortheoriginandevolutionofGPSradiosourcesrequireastrongconnec-
a tion betweenthe radio sourcemorphologyandthe gaseousnuclear environment,such
[
opportunities for detailed study are important. Our VLBI images of PKS 2254−367

1 show that the previously identified similarities between IC 1459 and NGC1052 con-
v tinue onto the pc-scale. Both compact radio sourcesappear to have symmetric jets of
3 approximatelythe same luminosity,much lowerthan typically notedin compactdou-
9 ble GPS sources. Similarities between PKS 2254−367and NGC1052,and differences
3 withrespecttootherGPSgalaxies,leadustospeculatethatasub-classofGPSradio
4
sources, with low luminosity and with jet-dominated morphologies, exists and would
0
belargelyabsentfromradiosourcesurveyswith∼1Jyfluxdensitycutoffs.Wesuggest
.
1 that this possible low-luminosity, jet-dominated population of GPS sources could be
0 ananalogof the FR-I radio galaxies,with the higher luminosity lobe-dominatedGPS
5 sources being the analog of the FR-II radio galaxies.
1
: Key words: galaxies: active — galaxies: individual: PKS 2254−367 (IC 1459) —
v
techniques: high angular resolution
i
X
r
a
1 INTRODUCTION populationofrelativisticelectrons(althoughtheseinterpre-
tations can be complicated in high density regions by the
Giga-hertz–Peaked Spectrum (GPS) sources represent at
effects of absorption).
least10%centofthebrightradiosourcepopulation(O’Dea
1998) and are thought to represent the early evolutionary Multi-frequency, multi-epoch observations with the
stages of the FR-II, and perhaps FR-I, radio galaxies (e.g., Australia Telescope Compact Array (ATCA)(Tingay et al.
Snellen et al. 2003). High angular resolution studies are 2003a)establishedPKSB2254−367asaGPSsource.Inad-
therefore of great importance for understanding the early dition to the peaked radio spectrum, PKS B2254−367 dis-
stages of their evolution, particularly for low-redshift GPS plays all of the typical characteristics of GPS sources, with
galaxies where high angular resolution corresponds to high compact structureseen on themaximum ATCA 6kmbase-
linear resolution. Multi-frequency observations additionally lines, low fractional polarisation, and low variability over
allow the spectral index distribution to be studied: flat or the 3.5 year period of the ATCA monitoring (Tingay, Ed-
inverted spectra identify regions of active particle accelera- wards&Tzioumis 2003b;Edwards&Tingay2004).Ingen-
tion, whereas steeper spectra are consistent with an aging eral,GPSobjectshavemarginalvariability,representingthe
variations due to the evolution of a jet originating from a
black hole and accretion disk system. In GPS objects, the
⋆ E-mail:[email protected] contribution to the total flux density of the jet is small in
† E-mail:[email protected] fractional terms, leading to low levels of variability. Long
(cid:13)c 2014RAS
2 S.J. Tingay & P.G. Edwards
term gradual flux density variability has been reported for the spectral properties of the source structure, are difficult
OQ208(Wuetal.2013).PKS2254−367hasamoreirregular to achieve given the poor (u,v) coverage of the VLBA at
variability, similar to high frequency peaking GPS objects southern declinations. Theabsence of Brewster, North Lib-
(e.g.Orienti, Dallacasa & Stanghellini(2007)andreferences erty,andHancockfromthearrayresultsinalargeholeina
therein). Observations at 20GHz (Murphy et al. 2010) and largely east-west (u,v) coverage. On the long baseline side
95GHz(Sadleretal.2008)confirmthePKS2254−367spec- oftheholearebaselinestotheMaunaKeaandStCroixan-
tral index steepens beyond 8.6GHz (the highest frequency tennas. On the short baseline side of the hole are baselines
of the original ATCA observations). between the five antennas concentrated in the south west
Assuming a Hubble constant of 75 kms−1Mpc−1 and US.
correcting for the motion of our Galaxy in the direction A consequence of the (u,v) coverage is that approxi-
of PKS 2254−367 relative to the Cosmic Microwave Back- mately matched resolution images can be made at our four
ground,thePKS2254−367redshift(0.006011,Zwaan et al. frequenciesusing acombination of restrictions on the(u,v)
(2004)) indicates a distance of 20.5Mpc. PKS 2254−367 data and weighting of the visibilities (described in a sub-
is therefore one of the closest known GPS radio sources. section below). This approach reveals useful qualitative in-
NGC 1052 (z=0.005037) at 17.2Mpc and PKS 1718−649 formation on the structure of the compact radio source.
(NGC 6328: 0.014428) at 59Mpc, in the same frame of ref- However,theuncertainties in deconvolution inducedby the
erence, are the only two other GPS radio sources within poor (u,v) coverage do not allow quantitative information
100Mpc,forwhichthedetailedkinematicsofthehostgalax- on the spectral characteristics of the structure.
ies can be studied (see, e.g., Labiano et al. 2007). All three In order to best determine spectral information, the
galaxies present strong evidence for merger activity, an ac- long baselines need to be discarded and restrictions on the
tivelyfuelledblackhole,andhighdensityenvironmentswith shortbaseline(u,v)coverageareappliedasafunctionoffre-
which theradio sources interact (Tingay et al. 2003b). quency to provide visibilities matched in spatial frequency
The host galaxy of PKS B2254−367, IC 1459, is a gi- (without the need for additional weighting) for component
ant elliptical galaxy in a loose group of spirals. IC 1459 is fittinginthe(u,v)plane(describedinasub-sectionbelow).
a LINER (Verdoes Kleijn et al. 2000), and has one of the Usingthesetwoapproaches,wedeterminelimitedinfor-
strongestcounter-rotatingcorecomponentsofanyobserved mation on the structure and spectra of the compact radio
elliptical, suggestive of merger activity (Forbes, Franx, and source.
Illingworth1994).Thedustdistributionisirregularnearthe
nucleus, indicating that infalling material may currently be
fuelling the active nucleus (Forbes et al. 1994). The cen-
tral black hole mass in IC 1459 is between 4×108M⊙ and 2.1 Approximately matched resolution imaging
3×109M⊙ (Cappellari et al. 2002). In order to produce the highest angular resolution approxi-
matelymatchedimagesofthecompactstructure,westarted
with the full (u,v) coverage at 1.7GHz. At 2.3GHz, the
same (u,v) coverage was used but a Gaussian taper giving
2 OBSERVATIONS AND RESULTS
a weight of 0.5 at a (u,v) distance of 20 Mλ (denoted 0.5,
Very Long Baseline Interferometry (VLBI) observations of 20) was applied to produce an image which approximately
PKS 2254−367 were made using the VLBA (excluding the matchedtheresolutionobtainedat1.7GHz.At5.0GHz,no
northernmoststationsatBrewster,NorthLibertyandHan- Gaussian taper was used, but the long baselines to Mauna
cock)atfrequenciesof1.655and4.975GHzon2003Novem- Kea to Saint Croix were removed, to approximately match
ber 29 and at frequencies of 2.266 and 8.416 GHz on 2003 theresolution at thetwolower frequencies. At8.4GHz,the
December11.Standardobservingsetups,withtwo-bitquan- dataonallbaselinestoMKandSCwerealsoremovedfrom
tisation, were used. the dataset and a Gaussian taper of 0.5, 30 was applied to
The 1.7 and 5.0GHz observations switched between thedata,toapproximatelymatchtheresolutionattheother
these two bands on a 15 minute timescale over the 6 hour frequencies.Allimageswereproducedusinguniformweight-
durationoftheobservation.A32MHzaggregatebandwidth ing of the data, image sizes of 256 × 256 pixels, and pixel
wasusedatLCP.The2.3and8.4GHzobservationsusedthe sizesof1.0mas.Standarddeconvolutionandself-calibration
dual“S/X”bandreceivertorecordRCP.Switchingbetween routineswereusedtoformtheimages.Asphasereferencing
these frequencies was therefore not required, as both are wasnotusedfortheseobservations,theself-calibrationstep
available simultaneously. In this case, the aggregate band- in imaging (and indeed via fringe-fitting in AIPS), forces
width per band is half of that which is available for the1.7 the brightest component in the image at the centre of the
and 5.0GHz observations, but the integration time is ap- image. As seen below, the brightest component is the same
proximately doubled. component at all frequencies, thusthealignment of theim-
The data were correlated on the VLBA processor and agesatthedifferentfrequenciesissecuretofirstorder.Small
initial standard data reduction was performed in AIPS,ac- frequency-dependent shifts in the centroid of the brightest
cordingtostandardroutines(amplitudecalibration, instru- componentarelost duetoself-calibration. However,acom-
mental phase and delay calibration, fringe-fitting etc). The parisonoftheoverallstructureofthesourceallowsasecure
data were then exported to DIFMAP (Shepherd 1997) for identificationofthedifferentcomponentsofemission across
editing, imaging, and modelfitting of the(u,v) data. the different frequencies. These identifications are used to
Theaimsofimagingthedataatthemaximumpossible guide the model-fitting process described in the next sec-
angular resolution, to investigate the structure of the com- tion.
pactradiosource,andtoobtainquantitativeinformationon Table 1 lists the tapers used at each frequency, the di-
(cid:13)c 2014RAS,MNRAS000,1–7
PKS 2254−367, a low redshift GPS radio source 3
Table 1.Parameters of theimages. Thebeam parameters aresemi minoraxis FWHM× semimajor axis
FWHM @ beam position angle in degrees. (u,v) coverage refers to the full VLBA (Full) or without the
baselines to Mauna Kea (−MK) or Saint Croix (−SC). Taper is described in the text. RMS is the Root
MeanSquareintheresidualimagepixels.
Parameter 1.7GHz 2.3GHz 5.0GHz 8.4GHz
Beam(mas×mas@PA◦) 3.3×[email protected] 3.4×[email protected] 3.5×[email protected] 3.2×[email protected]
(u,v)coverage Full Full −MK,−SC −MK,−SC
Taper(weight, uvdistance) NA 0.5,20 NA 0.5,30
RMS(mJy/beam) 1.0 1.1 0.9 0.8
mensionsoftheresultantsynthesisedbeams,andtheimage objects (CSOs). CSOs are compact (<1 kpc in projected
RMS values achieved. extent)andhavehighlysymmetricstructures(Wilkinsonet
Finally, allimages wererestored with acommon Gaus- al.1994);theyarethoughttobetheprogenitorsofFanaroff-
sian restoring beam of 3.5 mas x 11 mas at a PA of 6.0 Riley typeII radio galaxies (Readhead et al. 1996). Snellen
degrees and are shown in Figure 1. This procedure follows etal.(2004)showthataveryhighpercentageoflowredshift
theimaging techniquesused byTingay & Murphy(2001). (0.005<z<0.16), compact (<2′′),andrelativelyweak (>
100 mJy at 1.4GHz) radio sources associated with galaxies
brightatopticalwavelengthscanbeclassifiedasGPSorCSS
2.2 Matched resolution model-fitting
(compact steepspectrum) radiosources, andare, therefore,
As mentioned above, we seek to obtain limited spectral in- perhapsyoungradio galaxies.
formation on the compact radio source by discarding the WethereforeexaminePKS2254−367todeterminehow
longest baselines to Mauna Kea and St Croix and restrict- itrelatestotheclassesofobjects(GPS,CSS,CSO)thought
ing the (u,v) coverages at all four frequencies to a range of to be the precursors of powerful radio galaxies, and in par-
0–9 Mλ. This allows the full short baseline (u,v) coverage ticular other nearby radio sources in these categories.
at 1.7GHz, and progressively more restricted coverages at Tingay et al. (2003b) identified PKS 2254−367
thehigher frequencies. (D=20.5Mpc)asoneofthreelowredshiftGPSradiogalax-
The 1.7GHzdataset was fitwith a model consisting of ies, along with PKS 1718−649 (D=59Mpc) and NGC1052
five circular Gaussian components, using the task modelfit (D=17.2Mpc).PKS1718−649hasaso-calledclassicaldou-
in DIFMAP. We allowed the flux densities, positions, and ble morphology. Tingay et al. (2003b) listed a number of
sizes ofthefivecomponentstovaryuntilconvergence.This similarities between NGC1052 and the host galaxy of PKS
model was then fit to the data at the three higher frequen- 2254−267 (IC 1459). Both galxies have counter-rotating
ciesbykeepingthepositionsandsizesofthefivecomponents cores (Forbes et al. 1994; Bettoni et al. 2001) and LINER
constant and allowing the flux densities to vary. Good fits spectra(VerdoesKleijnetal.2000;Gabeletal.2000).Both
werefoundatallfourfrequenciesusingthismethod,provid- radio sources are predominantly compact and have peaked
ingintegrated fluxdensities for each component in orderto spectra (de Vries, Barthel, & O’Dea 1997; Tingay et al.
estimate the spectrum for each component. Model compo- 2003b). All three low redshift GPS sources support mod-
nent types other than circular Gaussians were tested using elsin which a densenuclearenvironmentand mergeractiv-
this model-fitting approach and were not found to provide ity are the cause for the mostly compact radio structure.
significantly better fits to the data. We justify the use of PKS 2254−367 and NGC1052 are hosted in cluster domi-
circular Gaussians as the simplest set of components that nant (CD) elliptical galaxies.
provideagoodfittothedatawiththeminimumnumberof
Of the three bright components that dominate the pc-
free parameters. scalestructureofPKS2254−367,thoselabelledA,B,andC
TheerrorsontheVLBAfluxdensityscaleareapproxi-
inTable1,Ahasamarginally peakedspectrum,sitsat the
mately5%.Themodelfitresultsforallfourfrequenciesare
centre of the structure, is the brightest component, and is
given in Table 2 and thespectra are shown in Figure 2.
themost compact of thecomponents.Wethereforeidentify
The results of the model-fitting are broadly consistent
component A as likely to correspond to the nucleus of the
withtheresultsoftheimagingshowninFigure1,discussed
radio source, presumably coincident with the supermassive
below.
black hole known toreside in the galaxy.
The pc-scale radio source therefore appears to be
double-sided,withjet-likeextensionscomprisedofthebright
3 DISCUSSION
inner component B and the weaker outer component D to-
The analysis above provides information on the structure wards the north-west, and the bright inner component C
and spectral properties of PKS 2254−367. As one of only and the weaker outer component E towards the south-east.
a small number of GPS radio sources within 100Mpc, it is This structure is reflected in both the modelfitting results
important to determine the basic properties of the source andtheimagingresults.Thenorth-westcomponents(Band
and compare them to the other nearby, and more distant, D)havespectrathatsteepenabove2.3GHz.Thesouth-east
GPS sources. components(CandE)haveflatspectra.Thespectraforthe
ApointofinterestishowthenearbyGPSsources,such fivecomponentsare shown in Figure 2.
asPKS2254−367,relatetotheso-calledcompactsymmetric PKS2254−367 thereforeseemstobeacandidateCSO,
(cid:13)c 2014RAS,MNRAS000,1–7
4 S.J. Tingay & P.G. Edwards
Figure 1. VLBA images of PKS2254−367. All images have been restored witha beam of 11×3.5mas at a position angle of 6◦. The
imagepeaks are243, 262, 251and240 mJy/beam, at1.7,2.3,5.0and8.4GHzrespectively. Contours areplottedat −1(dashed), 1,2,
4,8,16,32,and64%oftheimagepeak.
Table2.Fluxdensitiesofcomponentsfittedto(u,v)dataintherange0–9Mλ.Sisthefluxdensity(inJy)
ofthe model component atthe frequency(inGHz) indicated bythe subscript.r isthe angular distance in
milli-arcsecondsofthecomponentfromthephasecentre.θisthepositionangleindegrees(eastofnorth)of
thecomponent.aistheFullWidthatHalfMaximumofacircularGaussiancomponent(inmilli-arcseconds)
Component S1.7 S2.3 S5.0 S8.6 r θ a
A 0.49±0.03 0.58±0.03 0.53±0.03 0.48±0.03 1.6 −174 0.0
B 0.33±0.02 0.30±0.02 0.16±0.01 0.10±0.01 15.2 −42 7.3
C 0.17±0.02 0.19±0.02 0.17±0.02 0.16±0.02 16.2 149 6.9
D 0.08±0.01 0.04±0.01 <0.03 <0.03 40.9 −29 10.6
E 0.04±0.01 0.05±0.01 0.04±0.01 0.05±0.01 42.0 153 6.8
(cid:13)c 2014RAS,MNRAS000,1–7
PKS 2254−367, a low redshift GPS radio source 5
Figure 2.Spectraforthemodelfitcomponents ofTable2.
with a symmetric structure centred on component A. In The integrated monochromatic luminosity at 5GHz of
comparison to the CSO candidates found in the COINS PKS 2254−367 is ∼ 7 × 1022WHz−1, and the compact
survey (Peck and Taylor 2000), PKS 2254−367 seems most radio source in NGC1052 has a very similar luminosity,
similar to the sources J0427+41 and J1546+00. Both these ∼ 5×1022WHz−1. PKS 1718−649 has a classical double
sourcesappeartohaveastrongcore(moredominantthanin morphology with an extent of 2 pc and a monochromatic
PKS2254−367,indicatingthatthesecoresmaybebeamed) luminosity at 5GHz of ∼ 2×1024WHz−1 (Tingay et al.
and double sided jets. However, such sources make up a 1997).UnlikeNGC1052,nomotioninthestructureofPKS
small minority of the candidates found by Peck and Taylor 1718−649 has been detected following multi-epoch VLBI
(2000). Orientiet al. (2004) report VLBA images of a sub- observations, with an estimated upper limit on the sepa-
sampleof CSSobjects and outof 18objects (from aparent ration speed of the double components of 0.08c (Tingay et
sampleof87objects),twohavemorphologiessimilartoPKS al. 2002).
2254−367, possibly without as strong a central core com-
InNGC1052, Vermeulen et al.(2003) foundstrong ev-
ponent (B3-VLA 1242+410 and B3-VLA 2358+406). Most
idence for free-free absorption within the central pc. We
candidate CSOs are dominated by strong edge-brightened cannot examine this possibility for PKS 2254−367, since
lobesandhaveweakorabsentjetsandweakorabsentcores. the limited frequency range, (u,v) coverage, and resolution
The more frequently observed morphology in the Peck and
of our data, compared to the NGC1052 data, make detec-
Taylor sample is more reminiscent of PKS 1718−649.
tionoffree-freeabsorptioneffectsdifficult.Ingeneral,distin-
Our new VLBA data for PKS 2254−367 shows that guishing between free-free absorption and synchrotron self-
similaritiesbetweenPKS2254−367andNGC1052continue absorption is difficult in GPS/CSO objects, as examined in
to beseen for thepc-scale radio source. detail by Tingay and de Kool (2003) for PKS 1718−649.
More recent work on the absorption mechanism in PKS
From comprehensive multi-epoch VLBI monitoring,
1718−649 by Tingay et al. (2015) shows that free-free ab-
NGC1052 has been revealed to possess bi-directional jets
with apparent speeds of up to ∼0.38c that appear to be sorption is likely to be the cause of its spectral peak at
∼3GHz.
orientated in the plane of the sky (Vermeulen et al. 2003;
Listeretal.2013).ThejetsinPKS2254−367alsoappearto ThesimilaritiesbetweenNGC1052andPKS2254−367
bebi-directional,with a highdegree ofsymmetry,although on the pc-scale are interesting, as are the differences be-
similarmulti-epochVLBIobservationswouldberequiredto tween these two sources and PKS 1718−649 and similar
unambiguously demonstrate this through the detection of lobe-dominated GPS sources.
motion in the jet components. With a spacing between ob- With the new images of PKS 2254−367, the two clos-
servingepochsofonly12days,obviouslynomotionhasbeen
est GPS sources are of low luminosity and have extended
detected in PKS 2254−367 but the opposing jets appear
low-powerjets(subluminalforNGC1052,unknownforPKS
highly symmetric with regard to their surface brightnesses, 2254−367). The nextclosest sources are substantially more
indicating that they are in the plane of the sky and/or not luminous – fifty times so in the case of PKS 1718−649 –
significantly relativistic.
andaredominatedbypowerfuldoublemorphologies(some-
The largest extent of the jets in NGC1052 is approxi- times with weak cores and symmetric jets). PKS 2254−367
mately3pc(Vermeulenet al.2003). InPKS2254−367, the and NGC1052 are below the lower edge of the distribution
jets havea projected extent of approximately 8pc. of luminosities of the sample of low redshift GPS and CSS
(cid:13)c 2014RAS,MNRAS000,1–7
6 S.J. Tingay & P.G. Edwards
sourcesofSnellenetal.(2004),asallsourcesinthatsurvey a sample of weak (10–100 mJy corresponding to luminosi-
are more luminous than ∼ 1023WHz−1 at 5GHz. The ob- tiesof1021−22WHz−1atdistancesof∼10–∼70Mpc)radio
jectswithsomewhatsimilarmorphologiestoPKS2254−367 sourcesinnearbyellipticalgalaxies.Suchradiosourceshave
from Orientiet al. (2004) have luminosities substantially been examined with rudimentary high resolution observa-
higher,oforder1027WHz−1h−1at400MHz(theseareCSS tions previously (e.g., Slee et al. 1994) but without exami-
objects, with spectra still rising at 400 MHz). nationofbroad-bandradiospectra(tofindpeakedspectra)
We speculate that objects such as NGC1052 and PKS or detailed imaging.
2254−367 represent a sub-class of GPS radio source which Some progress in this direction has been made by de
arecompact,oflowluminosity,andjetdominated.Theyare Vriesetal.(2009),whoreportVLBIobservationsofsources
differentiated from many other GPS sources (such as PKS in the CORALZ sample of Snellen et al. (2004). An inter-
1718−649) by thelack of powerful, compact lobes. estingexampleisJ105731+405646 witha1.4GHzfluxden-
Such a possibility recalls the Fanaroff-Riley sityof47mJyandaclearly peakedspectrumnear1.3GHz.
lumonosity−morphology break seen in large-scale ra- Snellen et al. give a redshift of 0.008, with a corresponding
dio galaxies (Fanaroff & Riley 1974). The low luminosity luminosity of L=1021.59WHz−1 at 5GHz, however a red-
radio galaxies (FR-I type) have prominent (two-sided shift of 0.025 is also given in the literature for this object
or sometimes one-sided) jets and centre-brightened or (Miller et al. 2002). In any case, de Vrieset al. find the ob-
indistinct radio lobes. The great majority of the higher ject is unresolved in their EVN + VLBA observation, with
luminosity radio galaxies (FR-II type) are dominated by an image peak of 14 mJy/beam at 5GHz.
edge-brightenedlobes.Theytypicallyhaveapparentlyweak
cores and weak jets. The apparent weakness of the jets is
ascribed to relativistic beaming along thejet direction. 4 CONCLUSIONS
NGC1052 and PKS 2254−367 may be the low-
The nuclear radio sources in IC 1459 and NGC1052 are
luminosity,jet-dominated analogs ofFR-Ityperadiogalax-
ies. Objects like PKS 1718−649 may be the higher lumi- remarkably similar, as are the properties of the host galax-
ies themselves. Both compact radio sources appear to have
nosity analog of FR-II type radio galaxies. A possible im-
plication of such a luminosity−morphology break could be symmetricjetsofapproximatelythesameluminosity,much
lower than typically noted in compact doubleGPS sources.
that the origin of the difference between FR-I and FR-II
SimilaritiesbetweenPKS2254−367 andNGC1052anddif-
type radio galaxies lies on the sub-parsec-scale, where the
ferenceswithrespecttootherGPSgalaxiesleadustospec-
jetisformedclosetotheblackholeandaccretiondisk.Such
ulate that a sub-class of GPS radio sources, with low lu-
schemes are favoured in the theoretical models of Meier et
minosity and with jet-dominated morphologies, exists and
al.(1997),whosuggestthattheFR-I/FR-IIbreakisdueto
wouldbelargely absentfrompreviousradiosourcesurveys.
themagnetic field strength near theblack hole.
Wesuggestthatalow-luminosity,jet-dominatedpopulation
However,otherexplanationsmayexistforthemorphol-
ogy of NGC1052 and PKS2254−367. Thenuclearenviron- of GPS sources may be the analog of FR-I radio galaxies,
withthehigherluminositylobe-dominatedGPSsourcesbe-
mentisthoughttoplayamajor rolein theappearanceand
ing theanalog of FR-IIradio galaxies.
evolution of GPSradio sources, and thisseemslikely in the
case of these two objects. In this case, the sub-pc environ-
mentoftheblackholeand accretion disksystem would not
beofprimaryrelevance.Supportforthisviewcanbefound ACKNOWLEDGMENTS
in the fact that the properties of the pc-scale jets in FR-I
We thank an anonymous referee for very useful comments
andFR-IIobjects,asseenwithVLBI,areindistinguishable
onthismanuscriptthatpromptedsignificantimprovements.
intermsofmorphology andapparentspeed,forexampleas
The National Radio Astronomy Observatory is a facility
reported byGiovannini et al. (2001).
of the National Science Foundation operated under coop-
VeryfewsourceslikePKS2254−367orNGC1052have
erative agreement by Associated Universities, Inc. SJT ac-
been studied in detail since even doubling their distance
knowledges the support of the Western Australian State
to a modest ∼ 40Mpc would mean that they lie a fac-
government, in the form of a Western Australia Premiers
tor of four below the 1Jy cutoff that is typical of previ-
Fellowship.
ous radio source surveys. Multi-frequency surveys of weak
radio sources would be required to find such objects. Be-
yond ∼50Mpc only the higher luminosity compact double
REFERENCES
and GPS quasars (relativistically beamed emission) could
expecttobedetected.Evensurveysthataimtoimproveour Bettoni D., Galletta G., Garc´ıa-Burillo S., Rodr´ıguez-
knowledgeofthelessluminousGPSsourcepopulation,such FrancoA., 2001, A&A,374, 421
asthosebySnellenet al. (2004) orPeckand Taylor(2000), CappellariM.,VerolmeE.K.,vanderMarelR.P.,Verdoes
would struggle to detect PKS 2254−367 or NGC1052 at KleijnG.A.,Illingworth G.D.,FranxM.,Carollo C.M.,
distances >50Mpc. deZeeuw P. T., 2002, ApJ,578, 787
With only PKS 2254−367 and NGC1052 to suggest de Vries W. H., Barthel P. D., O’Dea C. P., 1997, A&A,
the possibility of a low-luminosity, jet dominated popu- 321, 105
lation of GPS sources which is the analog of FR-I radio de Vries N., Snellen I. A. G., Schilizzi R. T., Mack K.-H.,
galaxies, it is important to collect further data for other Kaiser C. R., 2009, A&A,498, 641
nearby candidate objects. The information on NGC1052 Edwards P. G., Tingay S. J., 2004, A&A,424, 91
andPKS2254−367suggestawayforward.Weshouldtarget Fanaroff B. L., Riley J. M., 1974, MNRAS,167, 31P
(cid:13)c 2014RAS,MNRAS000,1–7
PKS 2254−367, a low redshift GPS radio source 7
ForbesD.A.,FranxM.,IllingworthG.D.,1994,ApJ,428,
L49
Gabel J. R., Bruhweiler F. C., Crenshaw D. M., Kraemer
S.B., Miskey C. L., 2000, ApJ,532, 883
Giovannini,G.,Cotton,W.D.,Feretti,L.,Lara,L.&Ven-
turi, T. 2001, ApJ, 552, 508
Labiano A., Barthel P. D., O’Dea C. P., de Vries W. H.,
P´erez I., Baum S. A., 2007, A&A,463, 97
Lister M. L., et al., 2013, AJ, 146, 120
Meier D. L., Edgington S., Godon P., Payne D. G., Lind
K.R., 1997, Natur,388, 350
MillerC.J.,KrughoffK.S.,BatuskiD.J.,HillJ.M.,2002,
AJ, 124, 1918
Murphy T., et al., 2010, MNRAS,402, 2403
O’Dea C. P., 1998, PASP,110, 493
Orienti, M., Dallacasa, D. & Stanghellini, C. 2007, A&A,
475, 813
Orienti, M. et al. 2004, A&A,426, 463
Peck A.B., Taylor G. B., 2000, ApJ, 534, 90
ReadheadA.C.S.,TaylorG.B.,PearsonT.J.,Wilkinson
P. N.,1996, ApJ, 460, 634
Sadler E. M., Ricci R., Ekers R. D., Sault R. J., Jackson
C. A., deZotti G., 2008, MNRAS,385, 1656
Shepherd M. C., 1997, ASPC, 125, 77
SleeO.B.,SadlerE.M.,ReynoldsJ.E.,EkersR.D.,1994,
MNRAS,269, 928
SnellenI.A.G.,MackK.-H.,SchilizziR.T.,TschagerW.,
2003, PASA,20, 38
SnellenI.A.G.,MackK.-H.,SchilizziR.T.,TschagerW.,
2004, MNRAS,348, 227
Tingay S.J., et al., 1997, AJ, 113, 2025
Tingay S.J., et al., 2002, ApJS,141, 311
Tingay, S.J. et al., 2003a, PASJ, 55, 351
Tingay, S. J., Edwards, P. G. & Tzioumis, A. K., 2003b,
MNRAS,346, 327
Tingay S.J., et al., 2015, AJ, in press, arXiv:1412.4216
Tingay S.J., deKool M., 2003, AJ, 126, 723
Tingay, S.J., Murphy,D. W., 2001, ApJ, 546, 210
Verdoes Kleijn G. A., van der Marel R. P., Carollo C. M.,
deZeeuw P. T., 2000, AJ, 120, 1221
Vermeulen R.C., Ros E., Kellermann K. I., Cohen M. H.,
ZensusJ. A., van LangeveldeH. J., 2003, A&A,401, 113
Wilkinson P. N., Polatidis A. G., Readhead A. C. S., Xu
W., Pearson T. J., 1994, ApJ, 432, L87
Wu et al. 2013, A&A550, 113
Zwaan, M.A. et al. 2004, MNRAS,350, 1210
(cid:13)c 2014RAS,MNRAS000,1–7