New active galactic nuclei detected in ROSAT All Sky Survey galaxies; Part I: Verification of selection strategy

18 Pages·0.92 MB·English

by  W. Pietsch

A&A manuscript no. ASTRONOMY
(will be inserted by hand later) AND
Your thesaurus codes are: ASTROPHYSICS
11; (13.25.2; 11.01.2; 11.02.1; 4.19.1;
New active galactic nuclei detected in ROSAT All Sky
Survey galaxies
⋆
Part I: Verification of selection strategy
W. Pietsch1, K. Bischoff2, Th. Boller1, S. D¨obereiner1, W. Kollatschny2, and H.-U. Zimmermann1
8
9 1 Max-Planck-Institutfu¨r extraterrestrische Physik,Giessenbachstraße, D-85740 Garching, Germany
2
9 Universit¨ats-Sternwarte, Geismarlandstr. 11, D-37083 G¨ottingen, Germany
1
Received 24 November1997 /accepted 20 January 1998
n
a
J

2 Abstract. We present the first results of a program to 1. Introduction
2 identify so far unknown active nuclei (AGN) in galax-
ies. Candidate galactic nuclei have been selected for op- Active galactic nuclei are dominated by huge amounts of
1 energyreleasefromtheirnuclearregions.Thiscentralcon-
tical spectroscopy from a cross-correlationof the ROSAT
v
tinuum source is responsible for the photoionization of a
AllSky Survey(RASS) brightsourcecatalogwithoptical
0
central emission-line region. From the optical spectra one
1 galaxycatalogs.AhighX-rayfluxhasbeenusedaspointer
2 to galaxies with a high probability to contain active nu- candistinguishseveraltypesofAGN,e.g.Seyfertgalaxies,
1 clei. Only galaxies have been accepted for the program LINERs, and BL Lac objects. While Seyfert galaxies and
0 LINERs can be classified according to their optical emis-
for which no activity was noted in NED. For many of the
8 sion lines (see Sect. 4 for the classification scheme used),
galaxies no radial velocity was reported before. The opti-
9
the main characteristic of BL Lac objects is a featureless
/ cal spectra demonstrate that the galaxies covera redshift
h optical spectrum; for some BL Lac’s absorption or emis-
range of 0.014 to 0.13 and that most of them host active
p
sionlines become visible during lowstages ofcontiunuum
nuclei.For75%ofthe33candidatesthe X-rayemissionis
-
o caused by the AGN. In addition several of the remaining emission in optical spectra with very high S/N ratio.
tr candidates host Seyfert 2/LINER nuclei that, however, AGN are well known as strong X-ray emitters often
s most certainly cannot explain the X-ray emission alone. dominating the X-ray emission of the galaxy they are
a
: Three BL Lac objects have been detected serendipi- hostedin.ThishasnicelybeendemonstratedbyFabbiano
v
et al. (1992) in their analysis of the galaxy content of the
touslyingalaxyfieldsthathavebeenfollowedupbyshort
i
X Einstein X-ray observatory archive. They systematically
ROSAT HRI observations to confirm the X-ray galaxy
r identification with improved position accuracy and point searched for the galaxy content of the pointed observa-
a tions ofthe Einsteinsatellite resulting in’The X-rayCat-
response function. The sources show X-ray to radio flux
alog and Atlas of Galaxies’. The catalog comprises 493
ratios typical for X-ray selected BL Lac objects.
galaxies, 450 of which were imaged well, resulting in 238
The results presented in the paper prove the selection
detectionsand2123σ upperlimits.WhiletheX-raysofE
strategy as very successful to detect previously unknown
andS0galaxies(luminositiesupto1043 ergs−1)aredom-
AGN of all Seyfert 1 types in nearby galaxies encourag-
inated by the emission of a hot interstellar medium, the
ing the extension of this program. The detection of new
X-ray emission of spiralgalaxies (<1042 erg s−1) is dom-
nearby AGN will be used to initiate a detailed investiga-
inatedby the integratedoutputofevolvedstellarsources,
tion of their multi-wavelength properties and a compari-
such as supernova remnants and close accreting binaries
son with the more distant AGN population.
with a compact stellar remnant. If an active nucleus is
present in E and S0 galaxiestheir X-ray luminosity is be-
Keywords:X-rays:galaxies–Galaxies:active–BLLac-
tween 1041 erg s−1 and 1043 erg s−1 at the high end of
ertae objects: general – Surveys
the non-active galaxies of this type. For spiral galaxies
hosting an active nucleus the luminosity range extents to
above 1044 erg s−1, out-shining non-active spirals by up
to more than two orders of magnitude.
Send offprint requests to: W. Pietsch TheX-raysatelliteROSAT(Tru¨mper1983)performed
⋆ based partially on observations collected at the European the first all sky survey in the soft X-ray band (0.1–2.4
Southern Observatory,La Silla, Chile keV) using an imaging telescope. Details of the ROSAT
Correspondence to: [email protected] All Sky Survey (RASS) are described by Voges (1993). A
2 W. Pietsch et al.: NewAGN detected in RASSgalaxies
RASS bright source catalog has been produced (Voges et Table 1. Log of ROSAT HRI follow up observations for
al. 1996) by visual screening of the sources found by the RASS galaxies included into our optical program
detectalgorithmsintheautomaticprocessingoftheRASS
data using the standard analysis software system (SASS, Object Date Active
Voges et al. 1992). This catalog contains 18811 sources time
with a count rate above 0.05 cts s−1. The sources have a
ESO 113- G010 28.12.1995 2.4 ks
detectionlikelihoodof≥15andcontainatleast15source
10.06.1996 2.7 ks
photons. The typical positional accuracy is 30′′. We have
NGC 427 17.06.1996–18.06.1996 1.9 ks
analyzed the catalog for its galaxy content by correlation
10.01.1996 3.9 ks
with optical galaxy catalogs (Zimmermann et al. 1998).
ESO 080- G005 22.04.1996 4.3 ks
The results reportedaboveon galaxiesin the Einstein
ESO 416- G002 28.01.1995 2.3 ks
energyband(0.2–3.5keV)alsoholdforthesofterROSAT
band (0.1–2.4 keV) with some restrictions. If the AGN is ESO 15- IG 011 30.07.1995–17.09.1995 6.5 ks
hiddenbehindcircumnuclearmaterial(forinstancedueto ESO 552- G039 21.03.1996 1.9 ks
a circumnuclear disk as proposed in the unification mod- ESO 254- G017 29.10.1996–30.10.1996 5.5 ks
els) they will be heavily absorbed (e.g. AGN in Seyfert
PMN J0623-6436 04.07.1995–16.07.1995 4.4 ks
(Sy) 2 galaxies and - if present - in LINERs). There-
PMN J0630-2406 11.04.1996–12.04.1996 4.7 ks
fore, their observed luminosity in the ROSAT band will
be highly reduced often preventing the detection of the ESO 490- IG026 24.10.1996–25.10.1996 3.2 ks
nucleus. In such a case only secondary – less luminous ESO 209- G012 21.12.1995 1.4 ks
– effects of present or past activity like anomalous arms, ESO 602- G031 25.05.1995–26.05.1995 4.5 ks
jets, or extended emission fromthe halo can be measured
by ROSAT (see e.g. the detailed ROSAT studies of the
nearbySy2galaxyNGC4258ortheLINERandstarburst
galaxy NGC 3079 (Pietsch et al. 1994, 1998)). wellknownfortheirX-rayemissionfrominvestigationsof
FortypicalAGNspectra(e.g.powerlawspectrumwith theEinsteinObservatoryobservations(seeFabbianoetal.
photonindexof2.3)andmoderategalacticforegroundab- 1992),butinadditionwithgalaxieslocatedinclustersand
sorptionwithN ofafew1020cm−2thecountratelimitof groups (in which the X-ray emission may originate from
H
the RASS bright source catalog corresponds to an unab- hot gas in the cluster or group) and with galaxies known
sorbedfluxof10−12erg cm−2 s−1 .Thisfluxlimitdoesnot to host an active nucleus. Besides these classes showing
depend strongly on the shape of the spectrum assumed. exceptionally high X-ray intensities there remained some
Thereforeatdistancesof100Mpconlygalaxieswithlumi- galaxies where no reasons for the high X-ray emission is
nositiesabove1042erg s−1 canbedetected.Extrapolating known. Several of these galaxies were proposed for short
theEinsteinresultsfortheRASSbrightgalaxylistweex- X-ray follow up observations using the ROSAT HRI to
pect to see some E and S0 galaxies to distances of 300 verify the origin of the emission from the galaxy. While
Mpc whilenon-activespiralsshouldonlybe visibleto100 in some cases the HRI observations clearly demonstrated
Mpc.Galaxygroupsandunrecognizedclustersofgalaxies that the RASS X-rays were extended emission from an
may also be found in this X-ray luminosity range. early type galaxy or even from a cluster of galaxies, in
many cases they showed unresolved emission from the
HavingtheseconsiderationsinmindweusedtheRASS
galactic nucleus as expected from an AGN. We selected
brightgalaxycatalogtoidentifynewAGNingalaxies.Our
only these galaxic nuclei for further spectroscopic inves-
firstspectroscopicfollowupobservationsthatarereported
tigations. Table 1 summarizes the HRI observation dates
here demonstrate the success of this procedure. Further
and exposure times for the galaxies that were included
optical observations have been granted to this program
into the optical observing run we report in this paper.
and will lead to an extended list of new X-ray selected
In a follow up of this workwe cross-correlatedthe im-
AGNinnearbygalaxies.Resultswillbepublishedinforth-
proved X-ray source catalog produced from re-processing
coming papers of this series,that also will discuss sample
of the RASS list (RASS bright source catalog, Voges et
properties.
al. 1996) with the Catalogue of Principal Galaxies PGC
(Paturel et al. 1989). The candidate galaxies have been
2. Sample selection categorized according to their X-ray emission, separat-
ing cluster candidates and active galaxies (Zimmermann
Cross-correlationsof sourcesdetected in the first process-
et al. 1998). Correlations with the NASA Extragalactic
ing of the RASS with a large galaxy catalog merged from
Database (NED)1 have been used to accumulate addi-
the most important optical catalogs of galaxies (RC2, de
tional information on these galaxies. This procedure re-
Vaucouleurs et al. 1976; RSA, Sandage & Tamman 1981;
Tully 1988; UGC, Nilson 1973; ESO, Lauberts 1982) not 1 The NASA/IPAC Extragalactic Database is operated by
only showed identifications with nearby galaxies that are theJetPropulsionLaboratory,CaliforniaInstituteofTechnol-
W. Pietsch et al.: NewAGN detected in RASSgalaxies 3
sulted in a list of galaxies as candidates for the X-ray X-ray emission from this source. However, the automatic
sources that show unusually high X-ray fluxes and were source detection was confused by X-ray emission from a
not known to host an active nucleus from observations nearbybrightstarandtherefore,theX-raycountrateand
atother wavelengths.No redshifts hadbeen measuredfor position for 1RXS J055559.4-612438are not reliable. The
mostofthesegalaxies.Toensuretheidentificationwepro- RASS bright source catalog also contains information on
ducedandinvestigatedoverlaysoftheX-raycontoursonto the extent ofthe X-raysources.If one uses a conservative
opticalimagesextractedfromthedigitizedskysurvey(see extentcriterium(extentlikelihood>10andextent>30”)
acknowledgments).To identify galaxiesonthe opticalim- threeofthesourcesintable2areX-rayextended(gal.no.
ages with the catalogedones we over-plottedgalaxyposi- 195, 238, and 276) indicating that the X-ray emission in
tions from the PGC catalog and NED extracts. these galaxies does not, or not exclusively originate from
To further characterize these X-ray bright galaxies we a nuclear source.
obtained optical observations for a subsample of these
galaxies in an observing run in November 1996 at the
2.2mESO/MPGtelescopeatLaSilla.Inthisinvestigation 3. Optical observations and data reduction
we included from our HRI follow up studies of the RASS
The optical observations were performed with the 2.2m
galaxiesthreebrightnearbyfieldobjects(onesourcenear
ESO/MPGtelescopeatLaSillaobservatoryfromNovem-
NGC 427anda sourcepair centeredonESO416–G002)
ber 2 to 7, 1996. We used EFOSC2 spectrograph with
andonecandidateopticalsourcewheretheX-rayemission
grism #6 and a 1′.′5-wide long slit and the 2048×2048
did not originate from the galaxy (ESO 490– G 008) but
15µm LORAL CCD, which gave a dispersion of 2 ˚A per
from a nearby radio source (PMN J0630-2406). Figure 1
pixel, a spectral coverage of 3800–8000 ˚A, and a spectral
shows X-ray contours for the HRI fields overlaid on digi-
resolution of 10 ˚A. The seeing was typically between 0.8”
tized optical images. In Fig. 2 the ESO 416– G 002 field
and 1.5”.
is extended to include the source pair mentioned above.
For acquisition we made images with Johnson R filter
All X-ray sources are unresolved and are supposed to be
and an integration time of typically 1 minute. These im-
nuclear point sources.
ages were also used to determine the morphological type
Table 2 summarizes the X-ray identification informa-
of the galaxies. For the spectra we chose a position angle
tion for the observed objects. Objects are identified by
◦
(PA) ofthe slitof90 (E-W).The only exceptionwasthe
ROSAT name (col. 1), prefix 1RXS J stands for a RASS
double nucleus galaxy ESO 120- G 016; in this case the
determinedposition(fieldnumberofgalaxycross-correla-
PAofthe slitwas60◦ inorderto takespectraofbothnu-
tionfromZimmermannetal.(1998)givenincol.2),prefix
clei simultaneously. The integration times of the spectra
RX J for a HRI detected source (HRI in col. 2). ROSAT
are given in Table 3, col. 2.
PSPC count rates for the RASS sources or HRI count
The data were reduced with standard MIDAS pro-
rates (0.1–2.4 keV) are given in col. 3, object name in
cedures. Wavelength calibration was done using He-Ar
col. 4. The optical positions (epoch 2000.0, col. 5 and 6)
comparisonlamps.The flux calibrationwasperformedby
were derived from the digitized sky survey (Irwin et al.
1994) and should be accurate to better than 2′′. Only meansofthefluxstandardstarsHD49798andNGC7293
(Turnshek et al. 1990). To extract the nuclear spectra we
the position of NGC 1217 was taken from NED as the
used an aperture of 3′′. The spectra are shown in Fig. 3.
galaxy information from the digitized sky list was con-
fused by a nearby source. Column 7 gives the separation Redshifts and line parameters were determined by fit-
of the optical position from the X-ray source location. ting line complexes ([Nii], Hα narrow and broad; [Oiii],
In all cases HRI positions are closer to the optical po- Hβ narrow and broad, [Sii]) and single lines. Only the
sitions of the galaxies than the RASS positions and the narrow Balmer line results have been used for the di-
nuclei of the galaxies are within the typical X-ray error agnostic line ratios. For galaxies with strong underlying
circle confirming the identification. For 1RXS J063059.4– stellar continua we determined redshifts from absorption
240636 the cross-correlation with the PGC catalog sug- lines. With the best accuracy (±80 km s−1) of all spec-
gested ESO 490– G 008 as identification. The position tra the redshift of the SW nucleus of ESO 122- IG 016
howeverwasoffby83′′.TheHRIfollowupobservationre- was measured via Mgi λ5175 and Nad λ5890, 5896,
jectedthegalaxyidentificationwithoutquestionandiden- and G band λ4304 absorption lines. In a second step all
tified a star-like nearby object coinciding with the PSPC spectra were normalized and rebinned to a logarithmic
and HRI position, which is also identified with the radio wavelength scale. Then the redshift was determined via
sourcePMNJ0630–2406(seeFig.1).ESO120–G023and cross-correlation with ESO 122- IG 016 SW. The radial
theRASSpositionhaveabigoffset.Fromtheopticalover- velocities were convertedfor the Earths motioninto a he-
lay in Zimmermann et al. (1998) it is clear that there is liocentricsystemusingaMIDASroutinewithacodegiven
by Stumpff (1980).
ogy, under contract with the National Aeronautics and Space
Administration.
4 W. Pietsch et al.: NewAGN detected in RASSgalaxies
Fig.1.ROSATHRIcontoursoverlaidontoopticaldigitizedskysurveyimages.TheX-rayimageshavebeensmoothed
with a Gaussian filter of 5′′ FWHM. Contours correspond to (0.1, 0.2, 0.5, 1.5, 4.5) cts arcsec−2
W. Pietsch et al.: NewAGN detected in RASSgalaxies 5
Fig. 1. continued
6 W. Pietsch et al.: NewAGN detected in RASSgalaxies
Fig.2. ROSAT HRI contour over-
layofthe extendedESO 416-G002
field onto an optical digitized sky
survey image. The X-ray image has
been smoothed with a Gaussian fil-
ter of 10′′ FWHM. Contours corre-
spondto(0.03,0.1,0.3)ctsarcsec−2
4. Results FromthespectralfitsweclassifiedtheobjectsasAGN
usingclassificationdiagrams(Fig.4)followingBaldwinet
al. (1981). We also determined the AGN type (Table 2,
The general information for our identification objects in
col. 10). Following the Osterbrock (1989) definition we
Table 2 is completed by redshift information(col.8,sum-
classified the AGN in Seyfert types according to the rel-
maryfromredshiftsfromTable3asdescribedbelow)mor-
ative strength of narrow to broad components of the H
phological and AGN type (col. 9 and 10) and comments α
andH emissionlines;Seyfert 1 galaxiesshowonly broad
concerning the identification (col. 11). We determined a β
components; Seyfert 1.2 to 1.8 galaxies show decreasing
morphologicaltype fromour acquisitionimagesandcom-
but visible broadcomponentsandnarrowcomponents in-
pared them with the morphological type given in NED
creasing in relative strength; Seyfert 1.9 galaxies show
(availableformostofthebrightergalaxiesofoursample).
only a weak broad component in H but no broad H
IfourtypediffersfromtheonegiveninNEDweshowour α β
component; Seyfert 2 galaxies show only narrow Balmer
type marked by *. NED types are marked by a preceding
3. AGN type (col. 10) was determined from the optical lines.NarrowlineSeyfert1galaxies(NLS1)areapeculiar
group of Seyferts where (1) the Balmer lines of hydrogen
spectra (see below). Objects for which some AGN infor-
are only slightly broader than the forbidden lines such as
mation was already available in NED are marked by *.
[Oiii],[Nii]and[Sii];(2)they oftenshow the presenceof
In col.11 we indicate X-rayvariability (X-var,see discus-
emission lines from Feii (e.g. the optical multiplets cen-
sionbelow)andcommentongrouporclusterenvironment,
tered at 4570˚A, 5190˚A and 5300 ˚A) or higher ionization
nearby stars and mark by ”ID?”, when the identification
iron lines such as [Fevii] λ6087 and [Fex] λ6375 (these
of the X-ray source with the galaxy is questionable (see
lines areoften seeninSeyfert1 galaxiesbutgenerallynot
Section5).NEDdetails andreferencesonsourcesmarked
in Seyfert 2 galaxies); and (3) the ratio of [Oiii] λ5007
by * in col. 9 to 11 are given in the notes at the bottom
to H is < 3, a level which Shuder & Osterbrock (1981)
of Table 2. β
foundto discriminate wellSeyfert1sfromSeyfert2s.The
In Table 3 we give results from the analysis of the op- full-width at half-maximum (FWHM) of NLS1 hydrogen
tical spectra. If there are redshifts available in NED (col. Balmerlinesisusuallyintherange≈500–1500kms−1(cf.
5, references col. 6) they are within the errors compara- Goodrich 1989 and Osterbrock & Pogge 1985). LINERs
ble to ourredshifts determined fromemission(col.3)and (LowIonizationNuclearEmission-lineRegion)wedefine-
absorption lines (col. 4). If possible, diagnostic line ratios followingHo (1997)-by[Oi]/Hα>0.17,[Nii]/Hα>0.6,
have been determined from fits to the line complexes and [Sii]/Hα>0.4 and[Oiii]/Hβ >3.Forsome ofour galax-
are given in columns 7 to 10. If a broad H component ies a separation between a Seyfert 2 or LINER nucleus is
α
was present we determined the FWHM from a multicom- not possible because [Oiii] and Hβ are not detected. The
ponent fit (col. 11). For very asymmetric broad line com- galaxy nuclei however clearly have to be classified as ac-
plexes the FWHM was estimated. In col. 12 we comment tive as their [Sii]/Hα and [Nii]/Hα ratios show. For RX
on special features of the spectra. J023454.8-293425theHα lineisshiftedoutofourspectral
W. Pietsch et al.: NewAGN detected in RASSgalaxies 7
range. The redshift has been derived using Mgii λ2798, The HRI rate could be lower, because in the RASS anal-
[Nev] λ3426, [Oii] λ3727, [Neiii] λ3869, 3968 and Hγ ysis additional unresolved emission is attributed to the
lines. From the absolute optical B magnitude it clearly source that is resolved by the HRI. Such additional emis-
has to be classified as a Quasar. BL Lac candidates are sion could result from an extended galaxy halo or from
identified from their featureless blue spectra.High f /f , othernearbyX-raysourcesorfromdiffuse X-rayemission
X B
detection in the radio, and a starlike appearance in the due to galaxycluster membershipof ourtarget.A careful
optical strengthen the BL Lac nature. check of the PSPC and HRI images most certainly rules
We investigated the galaxies from 32 RASS correla- out this possibility for all 5 sources. On the other hand
tionsand4fieldsourcesdetectedintheHRIfollowupob- HRI and PSPC fluxes could differ because we did not as-
servations. Two galaxies and one component of a galaxy sume the appropriatespectralmodelforthe flux determi-
with a double nucleus show no emission lines that would nation.WhilethiseffectcouldexplainreducedHRIfluxes
indicate activity. We therefore classify them as ”non ac- byuptoafactoroftwoitcannotexplaintheextremevari-
tive”.TheothersourcesareclassifiedasSeyferts(Sy1(5), ationsandespeciallythestrongHRIfluxincreasedetected
Sy1.2 (7), Sy1.5 (3), Sy1.8 (5), Sy1.9 (1), Sy2 (2)), NLS1 intwosources.Themostprobableexplanationfortheflux
(1), LINER (3), Sy2/LINER (4), BL Lac (3), and QSO differences therefore is true source variability.This expla-
(1). nation is also consistent with the fact that we see more
The soft X-ray (0.1–2.4 keV), the far-infrared (40− often a flux decrease than an increase from RASS to HRI
120µm), and radio (1.4 GHz) fluxes and luminosities are observations because most of the sources lie close to our
listed in Table 4. A Hubble constant of H = 75 km s−1 detectionthresholdfortheRASS.X-rayfluxvariabilityon
0
Mpc−1 and cosmological deceleration parameter of q = these time scales is a common feature for active galactic
0
0.5 were used throughout. For the objects (col. 1) and nuclei (see e.g. Ulrich et al. 1997).
redshifts from Table 2 we calculatedthe distance (col. 3). For two fields the HRI observations have been split in
To compute the soft X-ray (0.1–2.4 keV) energy flux twoobservationintervalsseparatedbyhalfayear(seeTa-
(col.6)fromthecountrateweassumedapower-lawspec- ble 1). This allowedto investigate variability on this time
trum scale using the same detector and therefore avoiding the
crosscalibrationproblemsmentionedabove.TheHRIflux
f dE ∝ E−Γ+1 dE , (1) ofESO113-G010decreasedby afactorof3 betweenthe
E
observations. While the average HRI flux was a factor of
where f dE is the galaxy’s energy flux between pho- 2.2abovethePSPC(seeabove)duringDecember1995the
E
ton energies E and E +dE. We assume a fixed photon HRI flux was higher by even a factor of 3.6 and dropped
spectral index Γ = 2.3, which is the typical value found till June 1996, when it was however still higher than the
for extragalactic objects with ROSAT (cf. Hasinger et al. PSPC flux by a factor of 1.3. The HRI flux of NGC 427
1991,Walter&Fink1993),andanabsorbingcolumnden- did not vary by more than 15% between the observations
sity fixed atthe individual Galactic hydrogenvalue N in January and June 1996 but stayed at the lower inten-
Hgal
along the line of sight (col. 2, Dickey & Lockman 1990). sity with respect to the PSPC observation. The BL Lac
With this procedure we do not correct for absorption of candidate in the same field (RX J011219.5-320140)how-
the X-rays within the galaxy. Our X-ray fluxes derived ever showed an increase in flux between the observations
withthismethodthereforehavetobeunderstoodaslower by a factor of 1.5.
limitsfortheintrinsicfluxesiftheX-raysareemittedfrom TheopticalBmagnitudeandthelogofthemonochro-
active nuclei within the galaxies. This absorption effect maticfluxattheopticalbandat4400˚A(usingtherelation
could be corrected for if one would be able to fit spectra givenby Allen (1976),p. 174)are listed in columns 4 and
to the individual X-ray sources.While such fits are possi- 5, respectively. The absolute B-magnitude is given in col-
ble for RASS sources with more than 500 counts the few umn9.Column7givestheintegrated40-120µmfluxand
photons (typically < 100 counts for our sample) do not column8 the monochromatic1.4 GHz radioflux fromthe
allow such a procedure. NVSSsurvey(Condonetal.1996)or(ifnotavailable)the
For the 12 sources with follow-up ROSAT HRI obser- 4.85 GHz radio flux from NED, if available.
vationsweusedtheHRIcountratetodeterminetheX-ray The total far-infrared (40-120µm) fluxes, f , were
FIR
flux. For 3 sources PSPC and HRI determined fluxes cor- computedwiththeformulaofHelouetal.(1985)fromthe
respond within 20% (ESO 15- IG 011, PMN J0630-2406, IRAS 60 µm and 100 µm band fluxes taken from NED:
ESO 209- G 012), for 7 sources the HRI flux is signifi-
cantlylower(NGC427,ESO080-G005,ESO416-G002, f =1.26×10−11(2.58f +f ) erg cm−2 s−1, (2)
FIR 60 100
ESO254-G017,PMNJ0623-6436,ESO490-IG026,and
ESO 602- G 031 by a factor of 0.6, 0.4, 0.4, 0.7, 0.7, 0.6, where f and f are given in Jansky. The soft X-ray,
60 100
and0.2,respectively),andfor2 sourceshigher(ESO113- optical, far-infrared and radio fluxes were converted to
G010andESO552-G039by2.2and2.0).Severalexpla- luminositiesusingequation(7)ofSchmidt&Green(1986)
nationscanbeputforwardforthederivedfluxdifferences. (col. 10 to 13). Column 14 to 17 give flux ratios between
8 W. Pietsch et al.: NewAGN detected in RASSgalaxies
differentbands.Fluxratioshavebeenplottedasfunctions IC1867,NGC1217,AM0426-625S,ESO120-G023,
of absolute B magnitude and far infrared flux in Fig. 5. andUGC12492.Thereare,however,galaxiesforwhich
these arguments to exclude a Sy2/LINER nucleus as
the origin of the X-ray emission do not hold. HRI ob-
5. Discussion
servations of ESO 254- G 017 show a X-ray variable
As shown in the last section the X-ray selected galaxies nuclearpoint sourceandstronglyarguefor annuclear
whichwerefollowedupopticallycoverredshiftsfrom0.014 origin of the X-ray emission (see 2 and 3 above). The
to0.13.Mostofthemturnouttoharvestanactivenucleus opticalspectrumofthenucleusisheavilydisturbedby
or even merging nuclei where at least one component is thehostgalaxyspectrum,butdoesnotshowanybroad
active. It is, however, not clear yet if these active nuclei, BalmercomponentthatwouldindicateaSy1type.For
that we have classified using optical, radio and infrared the Sy2 galaxy UGC 3134 neither the morphological
data, really are the origin of the X-rays. type (SBc) nor a known group or cluster surrounding
We have several ways to attack this problem and can can explain the X-ray luminosity of ∼ 1043 erg s−1.
propose solutions for individual galaxies: HRI observationscouldclarify the interpretationifwe
see emission from a nuclear source.
1. WecanusetheextentinformationfromtheRASScat- 5. Wehaveperformedalinefittinganalysisofthepromi-
alog. If a source is extended the major part of the X- nent optical emission lines. The location of our ob-
ray emission will not originate from the nucleus. This jects in the diagnostic diagrams of Osterbrock (1989)
is the case for AM 0426-625 S, ESO 120- G 017, and clearly demonstrates their AGN nature (cf. Fig. 4).
ESO 122- IG 016. From the optical spectra their nu- ThefluxratiosplottedinFig.5areanadditionalindi-
clei have been classified as Sy2/LINER, non active, cationfortheAGNcharacteroftheobjects.Usingthe
and LINER+non active, respectively. From this type nomographshowninFig.1ofMaccacaroetal.(1988),
of nucleus X-ray emission in the ROSAT band should one sees that the combined X-ray flux and optical B-
be highly absorbed. The extended X-rays may origi- magnitude agree well with those expected from AGN
nate from hot gas in the early type galaxies or in the (cf. the left panel of Fig. 5). The middle panel of Fig.
groups they are member of. 5 gives the ratio of the X-ray to far-infrared (40-120
2. An HRI detection ofa sourcewhichis centeredonthe µm) flux. All of the objects show flux ratios above a
nucleusandisunresolvedstronglyarguesforanuclear valueof0.01.AsshowninFig.5ofBolleretal.(1997)
originof the X-rays.This is the casefor 9 sources(see most AGN exhibit flux ratios above that value. The-
Table 2). In addition we searched the ROSAT public oretical models addressing the X-ray and far-infrared
archive for additional coverage of the galaxies in our emissionofgalaxiesindifferentstatesofnuclearactiv-
sample. The only galaxy covered is UGC 716. It was ity suggest that flux ratios above about 0.01 require
observed serendipitously at 10′ offset from the center AGN activity, whereas values below about 0.01 can
ofanHRI pointinginJanuary1996.The X-raysource be explained by starburst activity (Bertoldi & Boller
isclearlyextendedandmostlikelyrepresentsemission 1998).UsingtheformalismofKellermannetal.(1989,
from a galaxy cluster. see their Fig. 4) one can quantify the radio-quiet or
3. Time variability of the X-ray flux rules out emission radio-loud nature of our objects. The ratio of the ra-
from extended gas clouds and strongly argues for an dio to optical flux density for most objects is below
AGN. As described above the comparison of the HRI 10 (right panel of Fig. 5) and objects below such a
andRASSfluxesfor9sourcesindicatestimevariability value are considered as radio-quiet (of course, there is
supportingthe AGNoriginofthe X-rays.This further no strict dividing line between radio-quiet and radio-
pointsatanAGNidentificationinadditiontotheHRI loud objects and there is a continuous increase of the
point source detection argument in 2. radio activity with increasing radio to optical flux ra-
4. We can compare the measured X-ray luminosity with tio).Weconsiderthesixobjectswithfluxratiosabove
X-rayluminositiesexpectedfromtheAGN/morphol- 10 as radio-loud and the most intense radio emitters
ogy type of the galaxy or for the galaxy surround- in our sample (RX J011232.8-320140and NGC 1218)
ings (group or cluster environment). While Seyfert 2 show a radio to optical flux ratio of about 275.
and LINER nuclei should not have luminosities in the
ROSAT band above 1041 erg s−1 the hot gas in early The considerations above propose 25 AGN identifica-
type galaxies or in groups or cluster may be more lu- tions and 1 dubious case for the 33 X-ray sources. This
minousbyfactorsof100or1000.Thiswouldarguefor representsasuccessrateofabout75%fordetectingactive
an origin of the X-ray emission from the surrounding nucleiusingourselectioncriteria.Ifwedonotconsiderthe
grouporclusteraslistedinNEDforUGC716,MCG- HRI pre-selection we still have a success rate of ∼ 50%.
01-05-031,AM0426-625S,ESO120-G023,ESO254- Of specific interest are the 3 BL Lac objects identi-
G 017, and ESO 122- IG 016. Hot gas in the early fied (one as the centralsource of the field and two nearby
type host galaxy could explain the X-ray emission in sources). We tried in vain to identify possible optical ab-
W. Pietsch et al.: NewAGN detected in RASSgalaxies 9
sorption lines to get a handle on the distance of the ob- References
jects.Thereforenoluminositydeterminationwaspossible.
Allen, C.W., 1976, ’Astrophysical Quantities’, The Athlone
Theirf /f andf /f valuesclearlyputthemintheregime
x r o r Press, London,p. 198
ofX-rayselectedBLLacobjects(seeFig.11ofBrinkmann
Baldwin, J.A., Phillips, M.M, Terlevich, R., 1981, PASP93, 5
et al. (1995)). Bertoldi F., Boller Th., 1998, in preparation
Theresultsofthisfirstopticalfollowupobservingrun Boller Th., Bertoldi F., Dennefeld M., Voges W., 1997, MPE
have demonstrated that our selection strategy from the Preprint 423, A&AS,in press
RASSbrightsourcecatalog/galaxycorrelationsisrather Brinkmann W., Siebert J., Reich W., et al., 1995, A&AS109,
efficient in detecting new active galactic nuclei. The re- 147
sults encourage further optical observations that will be Brissenden R.J.V., Tuohy I.R., Remillard R.A., et al., 1987,
Proc.Astr.Soc.Austr. 7, 212
reported in future papers of this series.
Condon J.J., et al. 1996, in preparation
The detection of the more than 100 new members
Coziol R., DemersS., Pena M., et al., 1993, AJ 105, 35
of the nearby AGN population expected from our pro-
Coziol R., Demers S., Pena M., Barneoud R., 1994, AJ 108,
gramwillallowto attackseveralimportantquestionsand
405
will stimulate further investigations: (i) sub-arcsec near- DaCostaL.N.,PellegriniP.S.,DavisM.,etal.,1991,ApJS75,
infrared imaging and spectroscopy to study the physical 935
conditionsinthe broadandnarrowline regions;(ii)asal- de Vaucouleurs G., de Vaucouleurs A., Corwin H.G.Jr., 1976,
readyafewofourobjectsarestrongFeiiemitters(MCG- second reference catalogue of bright galaxies, Austin,Uni-
02-14-009,PMNJ0623-6436,MCG-02-59-006)thesample versity of Texas Press
offersnewpossibilitiestostudytheFeiiexcitationmecha- de VaucouleursG.H., et al., 1991, third reference catalogue of
bright galaxies, version 3.9
nisms,whichareexpectedtobeconnectedtothestrength
Dickey J.M., Lockman F.J., 1990, ARA&A28, 215
of the X-ray emission; (iii) high resolution optical spec-
Fabbiano G., Kim G.-W., Trinchieri G., 1992, ApJS 80, 531
troscopy of our objects with asymmetric broad Balmer
FisherK.B.,HuchraJ.P.,StraussM.A.,etal.,1995,ApJS100,
lines (VIII Zw 36, ESO 416- G 002, ESO 552- G 039,
69
PMN J0623-6436, ESO 419- IG 026, ESO 209- G 012)
Goodrich R.W., 1989, ApJ 342, 224
will allow to further constrain broad-line region models; Hasinger G., Tru¨mper J., Schmidt M., 1991, A&A 246, L2
(iiii) Future X-ray missions will allow sub-arcsec imaging HelouG.,SoiferB.T.,Rowan-RobinsonM.,1985,ApJ298,L7
ofthenuclearregions(AXAF)anddetailedspectralmod- Ho L., 1997, ASPConf. Ser.103, 105
eling (AXAF, XMM). Irwin M., Maddox S., McMahon R.,1994, Spectrum 2, 14
KeelW.C.,deGrijpM.H.K.,MileyG.K.,1988,A&A203,250
Acknowledgements. This research has made use of the SIM- KellermannK.I.,SramekR.,SchmidtM.,ShafferD.B.,Green
BAD database operated at CDS, Strasbourg, France and of R.,1989, AJ 98, 1195
the NASA/IPAC Extragalactic Database (NED) which is op- LaubertsA.,1982,ESO/UppsalasurveyoftheESO(B)atlas,
erated by the Jet Propulsion Laboratory, California Institute Garching bei Mu¨nchen, FRG: European Southern Obser-
of Technology, under contract with the National Aeronautics vatory
and Space Administration. MaccacaroT.,GioiaI.M.,WolterA.,ZamoraniG.,StockeJ.T.,
To overlay the X-ray data we used programs kindly pro- 1988, ApJ 326, 680
vided by Andreas Vogler and images based on photographic Maia M.A.G., Da Costa L.N., Latham D.W., 1989, ApJS 69,
data of the National Geographic Society – Palomar Obser- 809
vatory Sky Survey (NGS-POSS) obtained using the Oschin Maza J., Ruiz M.T., Gonzalez L.E., Wischnjewsky M., 1992,
Telescope on Palomar Mountain and the UK Schmidt Tele- Rev.Mex.Astron.Astrofis. 24, 147
scope. The NGS-POSS was funded by a grant from the Na- Nilson P., 1973, Uppsala general catalog of galaxies,Uppsala,
tional Geographic Society to the California Institute of Tech- Sweden:Societatis scientiarum Upsaliensis
nology.TheUKSchmidtTelescopewasoperatedbytheRoyal Osterbrock, D.E., 1989, Astrophysicsof Gaseous Nebulae and
Observatory Edinburgh, with funding from the UK Science ActiveGalactic Nuclei (University ScienceBooks)
andEngineeringResearchCouncil,until1988June,andthere- Osterbrock D.E., Pogge R.W., 1985, ApJ297, 166
afterbytheAnglo-AustralianObservatory.Originalplatema- Paturel G., Fouque P., Bottinelli L., Gouguenheim L., 1989,
terial is copyright (c) the Royal Observatory Edinburgh and A&AS80, 299
the Anglo-Australian Observatory. The plates were processed Pietsch W., Vogler A., Kahabka P., Jain A., Klein U., 1994,
intothepresentcompresseddigitalformwiththeirpermission. A&A284, 386
TheDigitizedSkySurveywasproducedattheSpaceTelescope Pietsch W., Trinchieri G., Vogler A., 1998, A&A, in prepara-
Science InstituteunderUSGovernment grant NAGW-2166. tion
The ROSAT project is supported by the German Bun- Ramella M., Focardi P., Geller M.J., 1996, A&A312, 745
desministerium fu¨r Bildung, Wissenschaft, Forschung und Sandage A.R., Tamman G.A., 1981, Revised Shapley-Ames
Technologie (BMBF/DARA) and by the Max-Planck-Gesell- catalogue of galaxies, Washington: Carnegie Institution of
schaft (MPG). Washington
This work has been partially supported by Deut- Schmidt M., Green R.F., 1986, ApJ 305, 68
sche Agentur fu¨r Raumfahrtangelegenheiten (DARA) grant Sekiguchi K.,Wolstencroft R.D., 1993, MNRAS 263, 349
50OR94089. ShuderJ.M., Osterbrock D.E., 1981, ApJ, 250, 55
10 W. Pietsch et al.: NewAGN detected in RASSgalaxies
StockeJ.T.,MorrisS.L.,GioiaI.M.,etal.,1991,ApJS76,813
StumpffP., 1980, A&AS41, 1
TakataT.,YamadaT.,SaitoM.,ChamarauxP.,KazesI.,1994,
A&AS 104, 529
Tru¨mper J., 1983, Adv.Space Res. 2, 241
Tully R.B., 1988, Nearby Galaxies Catalog, Cambridge Uni-
versity Press, Cambridge
TurnshekD.A.,Bohlin R.L.,Williamson R.L.,etal.,1990, AJ
99, 1243
UlrichM.-H.,MaraschiL.,UrryC.M.,1997,ARAA35,inprint
Voges W., 1993, AdSpR13, 391
Voges W., Gruber R., Paul J., et al., 1992, in: Guyenne T.D.,
HuntJ.J.(eds.),Proc.EuropeanISYconference,ESAISY-
3, p.223
Voges W., Aschenbach B., Boller Th., et al., 1996, IAU Circ.
6420
Walter R., Fink H.H., 1993, A&A 274, 105
Yentis D.J., Cruddace R.G., Gursky H., et al., 1992, in: H.T.
MacGillivray,E.B.Thomson(eds.)Proc.on”DigitizedOp-
tical SkySurveys”, KluwerAcad. Publ., p.67
Zabludoff A.I., Geller M.J., Huchra J.P., Vogeley M.S., 1993,
AJ 106, 1273
Zimmermann H.-U.,et al., 1998, in preparation (GAL-BSC)