Astronomy Letters, 2016, Vol. 42, No. 4, p. 240–250. Translated from Pis’ma v Astronomicheskij Zhurnal. Sample of Cataclysmic Variables from 400d X-ray Survey R. A. Burenin*1, M. G. Revnivtsev1, A. Yu. Tkachenko1, V. S. Vorobyev1, A. N. Semena1, A. V. Meshcheryakov1, S. N. Dodonov2, M. V. Eselevitch3, M. N. Pavlinsky1 1Space Research Institute RAS (IKI), Moscow, Russia 2Special Astrophysical Observatory RAS, Nizhnij Arkhyz, Russia 3Institute of Solar-Terrestrial Physics SB RAS, Irkutsk, Russia ReceivedNovember24,2015 7 Abstract—Wepresentasampleofcataclysmicvariables(CVs)identifiedamongtheX-raysourcesfromthe 1 400squaredegreeX-raysurveybasedonROSATpointingdata(400d). TheprocedureoftheCVselection 0 among the X-ray sources using additional optical and infrared data from Sloan Digital Sky Survey and 2 WISEsurveyisdescribed. Theresultsoftheopticalobservationsoftheselectedobjectscarriedoutmainly with the Russian–Turkish 1.5-m telescope (RTT-150) and the 6-m telescope of the Special Astrophysical n a Observatory of the Russian Academy of Sciences (BTA) are presented. Some observations have also been J performed with the Sayan Observatory 1.6-m AZT-33IK telescope. Currently we selected eight CVs, four 7 of which were found for the first time in our work. Based on this sample, we have obtained preliminary 1 constraints on the CV X-ray luminosity function in the solar neighborhood in the low luminosity range, L ∼ 1029–1030 erg s−1 (0.5–2 keV). We show that the logarithmic slope of the CV X-ray luminosity X ] function in this luminosity range is less steep than at L >1031 erg s−1. From our CV X-ray luminosity E X function estimates it follows that few thousand CVs will be detected in the Spectrum-R¨ontgen-Gamma H (SRG) observatory all-sky X-ray survey at high Galactic latitudes, which will allow to obtain much more . accurate measurements of CV X-ray luminosity function in the luminosity range L <1030–1031 ergs−1. h X p Key words: cataclysmic variables, X-ray surveys, luminosity function - o r t s INTRODUCTION erate objects (the so-called “post-bounce” CVs, e.g., a [ Kolb, 1993; Howell et al., 2001; Knigge et al., 2011). Studies of samples of binary systems with accret- However,onlyasmallnumberofpossiblepost-bounce 1 ing white dwarfs (cataclysmic variables, CVs) provide CV candidates are known to date (see, e.g., Littlefair v important information on physical processes in these 0 systems that are difficult or even impossible to in- et al., 2008; Aviles et al., 2010). Among the prob- 4 lems that can be solved using studies of the statisti- vestigate by other means. For example, it turns out 2 cally complete samples of CVs, there is a question of thatthemainmechanismofangularmomentumlosses 5 how and at what rate the masses of the white dwarfs 0 for CVs with periods shorter than 2–3 h is the emis- (WDs) in these systems grow (e.g., Zorotovic et al., . sion of gravitational waves (Faulkner, 1971; Paczyn- 1 2011). Since accreting white dwarfs are apparently 0 ski, 1981). This prediction is confirmed by charac- the progenitors of type Ia supernovae, all such ques- 7 teristic features in the properties of the CV popula- tions appear to be closely related to the studies of the 1 tion, such as the minimum in the distribution of or- : properties of type Ia supernovae, which are used as v bital periods, Pmin ≈ 80 min (G¨ansicke et al., 2009), “standard candles” for cosmological measurements. i which agree well with the improved theoretical esti- X mates of this quantity for the case of angular momen- ItisalsoturnsoutthatCVscontributesignificantly r a tumlossestroughgravitationalradiationinbinarysys- totheGalacticridgeX-rayemission(Revnivtsevetal., tem (Knigge et al., 2011). 2009) and to the total X-ray luminosity of the Galaxy Nevertheless, there are also some discrepancies be- and other galaxies after the subtraction of the con- tween observations and theoretical views. For exam- tribution from low-mass X-ray binaries (Sazonov et ple, theoretical modeling of the CV population shows al., 2006; Revnivtsev et al., 2008a). To obtain more thatmostofthemmustbeinastatewithalowaccre- accurate measurements of the contribution from the tion luminosity, and their companions must be degen- CV population to the total X-ray luminosity of galax- ies,thestudiesofthestatisticallycompletesamplesof *e-mail: [email protected] these objects and the data that allow to measure the 240 CV sample from 400d X-ray survey 241 X-ray luminosity function are also required. The selection of the statistically complete CV sam- plesinX-raysisoneofthemostnaturalapproachesto obtain unbiased CV samples, because the X-ray emis- sion is generated in the accretion flow near the WD surface and, therefore, is a property of all types of CVs (for a more detailed discussion, see below). Such samples can be obtained in various X-ray sky surveys. TheX-rayluminosityfunctionsofCVsweremeasured earlier using the data from the RXTE all-sky survey (Sazonov et al., 2006) and the ROSAT North Eclip- tic Pole survey (Pretorius et al., 2007b; Pretorius and Knigge, 2012) as well as in harder X-ray band, us- ing INTEGRAL (Revnivtsev et al., 2008b) and Swift (Pretorius and Mukai, 2014) data. Almostalltheobjectsdetectedinthesesurveyshave X-ray luminosities1 L >1030 erg s−1. Since the CV X X-rayluminosityfunctionincreasestowardlowerlumi- nosities, the measurements of the CV number density Fig.1. ExamplesoftheCVX-rayspectra. Thesolidlineindi- at luminosities LX < 1030 erg s−1 should allow to re- catesapower-lawspectrumdN/dE∝E−1.6. fine the contribution from the CV population to the total X-ray luminosity of the Galaxy. In addition, at those low luminosities, one could expect to detect the CV SELECTION IN X-RAYS AND OPTICAL large number of post-bounce CVs. X-raysurveysprovideoneofthemostefficientmeth- Thus, the CV X-ray luminosity function measure- ods to select CVs. Indeed, all types of CVs appear to ments at low X-ray luminosities L < 1030 erg s−1 be the sources of X-ray emission associated with the X are of great interest. To obtain these measurements optically thin plasma emission of the accretion flow one should use deeper X-ray surveys, with flux limits heatedneartheWDsurface(see,e.g.,Aizu,1973). De- ∼ 10−15–10−14 ergs−1cm−2. In this work for these pending on the specific physical picture of accretion, purposes we used the 400 square degree (400d) X-ray the fraction of the X-ray emission at energies above survey, based on ROSAT pointing data, which were 0.5keVinthebolometricluminosityofCVscandiffer used earlier to detect galaxy clusters (Burenin et al., significantly,butitalwaysremainsfairlylarge,atleast 2007). More than 37000 point X-ray sources with ∼1% (see, e.g., Beuermann and Thomas, 1993). fluxes above 10−14 ergs−1cm−2 were detected in this The spectrum of CVs at the energies above 0.5 keV survey; about 22000 of these sources were detected in isdominatedmainlybytheopticallythinplasmaemis- the ROSAT fields overlapped with Sloan Digital Sky sionthatoriginateseitherinanaccretioncolumnorin Survey (Burenin et al., 2016b). The optical observa- anopticallythinboundarylayerneartheWDsurface. tionswiththeaimtosearchtheCVsamongtheX-ray The hot region near the WD surface is significantly sourcesfromthissurveywerestartedbyourgroupear- inhomogeneousintemperature,butthebulkoftheX- lier (Tkachenko et al., 2015); in this paper, we discuss rayemissionisgeneratedattemperaturesof1–20keV, the sample of CVs obtained to date. producing a spectrum with a characteristic power-law slopedN/dE ∝E−1.6. ItisshowninFig.1,wherethe Below,wedescribetheprocedureoftheCVselection examples of CV X-ray spectra from ROSAT, ASCA, amongtheROSATX-raysourcesusingadditionalop- and RXTE data are presented. tical and infrared data from SDSS and WISE surveys. Fig. 1 shows the spectra of the magnetized WDs WeprovidealistofCVsselectedinthe400dsurveyto EFEri and AMHer (the latter in two states: in the date and discuss the properties of these objects. We high state, in which a powerful soft component with also discuss the preliminary constraints on the X-ray a temperature of about 60 eV is seen, and in the low luminosity function of CVs in the solar neighborhood state, when there is no such component in the spec- obtained using this sample. trum)aswellasthespectrumofWZSge,anaccreting WD without a strong magnetic field. A more detailed discussionoftheX-rayspectraforaccretingWDswith 1Here and below all X-ray fluxes and luminosities are given a weak magnetic field can be found, for example, in in0.5–2keVenergyband. Byckling et al. (2010). It is clearly seen that the CV ASTRONOMY LETTERS Vol. 42 No. 4 2016 242 R. A. Burenin et al. spectra in the 0.5–10 keV energy band look approxi- from the spectroscopic sample of quasars (Richards et mately similar in shape, irrespective of the CV type al. 2002). and state, and that the general form of these spectra To eliminate a large number of quasars, which canbeapproximatelydescribedbyapowerlawwithin strongly contaminate the CV sample, we additionally a photon index of −1.6. discardedtheobjectswithcolorsw −w >0.6,where 1 2 w andw arethephotometricbandsoftheWISEin- 1 2 X-ray data frared all-sky survey with central wavelengths of 3.4 In our work we used the ROSAT pointed data in and 4.6 µm (Wright et al., 2010). For the vast major- the 0.5–2 keV energy band, at high galactic latitudes, ity of stars (except for the cool T dwarfs) and, con- which were used earlier to search for distant galaxy sequently, for the vast majority of CVs, this spectral clusters (400d survey, Burenin et al., 2007). In 400d range corresponds to the Rayleigh-Jeans part of the survey point X-ray sources were detected only in the spectrum;therefore,theCVcolormustbew1−w2 ≈0. central part of the ROSAT field of view at distances The criterion we used is fairly conservative; it should <18.5(cid:48) from its center, where ROSAT angular resolu- allow the possible contribution of the companion star tion is better than 70(cid:48)(cid:48) (PSF FWHM). The positional to be taken into account even if it is an L-type dwarf accuracy for an X-ray source is, on average, is better (see, e.g., Schmidt et al., 2015). than 10(cid:48)(cid:48) (at 95% confidence). We used 1605 ROSAT For the subsequent additional optical observations, pointings at high Galactic latitudes |b| > 25◦. The weselectedCVcandidateswithg(cid:48) magnitudes<20.0. CVs nearest to the Sun located within the Galactic Even when the contribution of the accretion disk is disk must be observed at such latitudes. smallandtheWDcontributiondominatesinthespec- The geometric area of the 400d survey for point tra of CVs, their absolute magnitudes are Mg(cid:48) ≈ 12 sources is 436.7 sq. deg. The geometric area of the (G¨ansickeetal.,2009;Revnivtsevetal.,2014). There- 400dsurveyoverlappingwithSDSSphotometricfields fore, the magnitude g(cid:48) ≈ 20.0 for such systems corre- is 262.3 sq. deg. Since the exposure distribution of sponds to a distance of ≈ 400 pc, which is approxi- ROSAT pointings is wide, the flux limit turns out to matelyequaltothethicknessoftheGalacticdisk. The be different in different fields of the survey. Half and minimumabsolutemagnitudeofCVsdependsstrongly 5% of the geometric area are gathered at X-ray flux on the WD mass and can reach Mg(cid:48) ≈ 14 for a WD of 2.5·10−14 and 8·10−15 ergs−1cm−2, respectively with a mass of ≈ 1.2M(cid:12) (Revnivtsev et al., 2014). (Burenin et al., 2016b). Even for such systems the magnitude limit g(cid:48) < 20.0 correspondsto a distanceof ∼160 pc, whichalso pro- Optical and Infrared data vides an appreciable search volume. Thereare53objectssatisfyingthecriteriadiscussed Asitwasdiscussedabove,intotalmorethan37000 aboveinthe400dsurvey. Someofthem(fourobjects) X-ray sourcesweredetected inthe400d survey, about turned out to be previously known CVs. To identify 22 000 of them — in the fields where there is an over- the nature of the remaining objects, we carried out lapwithSDSSphotometricfields. Obviously,thecom- additional optical observations. pleteopticalidentificationofallX-raysourcesfromthe 400dsurveyistoodifficult,andadditionaldataarere- quired to select CVs. For these purposes, we used the OPTICAL OBSERVATIONS OF CV CANDIDATES data from the 12-th release of the SDSS (Alam et al., Theadditionalopticalobservationswerecarriedout 2015) and the WISE infrared all-sky survey (Wright with the Russian–Turkish 1.5-m telescope (RTT-150) et al., 2010). using the medium- and low-resolution TFOSC spec- The stellar companions of the WDs in CVs trograph and the 6-m (BTA) telescope at the Special are mostly low-mass (< 1M ) main-sequence stars Astrophysical Observatory of the Russian Academy (cid:12) (Knigge, 2006). However, the WD in a CV and the of Sciences, where the SCORPIO-2 spectrograph accretion disk around it have a temperature no lower (Afanasyev, Moiseev,2005,2011)wasusedfortheob- than 104K (see, e.g. Townsley and G¨ansicke, 2009), servations. Recently, beginning in the fall of 2015, the and, hence, are blue in the optical band. In the blue observations have also been carried out with the 1.6- part of the optical spectrum and in the near ultravio- m AZT-33IK telescope at the Sayan Observatory of let,theCVspectrummustbealwaysdominatedbythe the Institute of Solar-Terrestrial Physics, the Siberian emission from the WD and the accretion disk; there- BranchoftheRussianAcademyofSciences, usingthe fore, this property can be used to select CVs in the newmedium-andlow-resolutionADAMspectrograph optical band (see, e.g., Green et al., 1982; Szkody et (Afanasyev et al., 2016; Burenin et al., 2016a). In al.,2002). FortheselectionofCVcandidates,weused all cases, we used grisms or volume phase holographic the criterion u(cid:48)−g(cid:48) <0.7, which was used previously gratingsoptimizedforarange4000–6000˚A,whichalso amongothercriteriaintheSDSStoeliminatetheWDs includes the blue part of the spectrum starting from ASTRONOMY LETTERS Vol. 42 No. 4 2016 CV sample from 400d X-ray survey 243 400d j001912.9+220736 400d j152212.8+080338 1 1 − − ˚A ˚A 2 2 − − m m c c 1 1 − − s s g g r r e e 7 7 1 1 − − 0 0 1 1 × × , , λ λ F F λ, ˚A λ, ˚A 400d j154730.1+071151 400d j160547.5+240524 1 1 − − ˚A ˚A 2 2 − − m m c c 1 1 − − s s g g r r e e 7 7 1 1 − − 0 0 1 1 × × , , λ λ F F λ, ˚A λ, ˚A Fig.2. —SpectraofCVsdetectedinthe400dsurvey. ASTRONOMY LETTERS Vol. 42 No. 4 2016 244 R. A. Burenin et al. B m r(cid:48) e, u(cid:48) g(cid:48) ud t ni g a m Time, hours Fig.3. BroadbandspectrumoftheCV400dj154730.1+071151. TheblackcurveshowstheRTT-150spectrum(seeFig.2). The Fig.4. B-bandlightcurveoftheCV400dj154730.1+071151ob- red curve shows the model spectrum of a binary system which tainedwiththeSayanobservatoryAZT-33IKtelescopeinMarch consist of an M2V red dwarf and a white dwarf with a tem- 2013. perature of ≈3·104 K (WD0320−539). The measurements in therange1500–2500˚Aaretakenfromtheobservationswiththe GALEXorbitalultraviolettelescope(Martinetal.,2005). The 000008canbefoundinBurwitzetal.(1999), Stokeet histogram also indicates the photometric SDSS measurements al. (1983), Liu et al. (1999) and Szkody et al. (2004), intheu(cid:48)g(cid:48)r(cid:48)i(cid:48) bands. respectively (for more details, see also below). 3500–3700˚A. The spectral resolution was from 7˚A for Notes on individual objects SCORPIO-2 at the BTA telescope to 12˚A for TFOSC at the RTT-150. 400d j050146.2−035914 (HY Eri) The spectroscopic observations of objects from this The CV HY Eri=RX J0501.7-0359 was discovered programmewerecarriedoutduringthelastfewyears. intheROSATall-skysurvey(Beuermannetal.,1999). We observed 38 objects, four of which turned out It was found to be an eclipsing binary system with a to be new, previously unknown CVs. The spec- magnetized WD, a polar (Burwitz et al., 1999). The tra for two of them (400d j001912.9+220736 and overall spectrum of this system and its measured or- 400dj152212.8+080338)werelateralsoobtainedinthe bital period, 2.855 h, are also given in Burwitz et al. SDSS.Theremainingobjectsturnedouttobequasars (1999). at various redshifts, typically, at z = 1–2. Examples Duetotheabsenceofanaccretiondiskinthebinary ofthespectraforsuchquasarsaregiveninTkachenko system, its optical brightness in the blue part of the et al. (2015). The spectra of CV found in our work spectrum is determined mainly by the emission from arepresentedinFig.2. Thespectraof15objectswith itshotWD.Byassumingcertaintemperature(104 K) magnitudes19.5<g(cid:48) <20werenotmeasuredtodate; and mass (0.6–0.8 M ) of the WD and, consequently, (cid:12) this work will be continued. its absolute magnitude M ≈11–12, we can estimate g(cid:48) the distance to the binary system, D ≈200–300 pc. THE CV SAMPLE The binary system was detected in the infrared in Basic information about the CVs in our sample is the 2MASS all-sky survey with J = 16.67 (Skrutskie presented in the Table 1. Four of them turn out to be et al., 2006). In this part of the spectrum, the com- previously known CVs, we detected four more objects panion star makes a major contribution. Using the using the optical observations described above, two of absolutemagnitudesofCVcompanionstarstabulated them were also independently detected in the SDSS in Knigge (2006), we can make another distance esti- duringthespectroscopicobservationsofquasarcandi- mate for the binary system: D ≈520 pc. In fact, the dates (Alam et al., 2015). The spectra of the CVs de- first distance estimate is only a lower limit, because tectedin400dsurveyarepresentedinFig.2. Theopti- thepossiblestrongheatingoftheWDsurfacethrough calspectraof400dj050146.2−035914,400dj124325.7+ accretionisnottakeninaccount;therefore,thesecond 025541, 400d j160002.4+331120 400d j204720.3+ distance estimate should be used. ASTRONOMY LETTERS Vol. 42 No. 4 2016 CV sample from 400d X-ray survey 245 Table1. Cataclysmicvariablesdetectedin400dsurvey Name α(J2000) δ (J2000) mg(cid:48) fX, ergs−1cm−2 LX, ergs−1 Othername 400dj001912.9+220736 001912.6 +220733 19.61 4.84·10−14 ∗6.4·1029 SDSSJ001912.58+220733.0 400dj050146.2−035914 050146.4 −035920 18.41 1.78·10−13 5.8·1030 HYEri 400dj124325.7+025541 124325.9 +025547 17.72 5.51·10−13 ∗1.3·1030 1E1240.8+0312 400dj152212.8+080338 152212.2 +080341 18.99 1.58·10−13 ∗1.2·1030 SDSSJ152212.20+080340.9 400dj154730.1+071151 154730.8 +071131 16.34 1.16·10−13 2.0·1029 400dj160002.4+331120 160003.7 +331114 19.89 8.87·10−14 ∗1.5·1030 VWCrB 400dj160547.5+240524 160548.0 +240531 19.78 5.47·10−14 ∗8.5·1029 400dj204720.3+000008 204720.8 +000008 19.36 4.19·10−14 ∗4.4·1029 SDSSJ204720.76+000007.7 ∗ —TheroughestimateoftheX-rayluminosityLX thatweuseforourpreliminaryestimatesoftheluminosityfunction(seetext formoredetails) 400d j124325.7+025541 (1E 1240.8+0312) distance to the source can be estimated to be about The source 1E 1240.8+0312 was discovered in the 120 pc. Einstein Observatory Medium Sensitivity Survey and Note,thattheX-rayluminosityofthebinarysystem identified with an optical object with bright emission basedonthisdistanceestimateisLX ≈2·1029ergs−1, lines (Stoke et al., 1983). In the same paper, it was which roughly corresponds to the accretion rate of suggested that the source is a dwarf nova, i.e., a CV 5 · 10−14 M(cid:12) for a WD of mass ≈ 0.5 M(cid:12) (if the with a WD without a strong magnetic field. X-ray luminosity accounts for a significant fraction of thebolometricluminosityoftheaccretionflow). Such an accretion rate is too low for a Roche-lobe-filling 400d j154730.1+071151 M2V companion star. Thus, this object may be clas- AbroadbandspectrumofanotherCVfromoursur- sified as the so-called “pre-cataclysmic variable”, i.e., vey,400dj154730.1+071151,ispresentedinFig.3. We abinarysystemwherethemasstransferfromthecom- approximatelydecomposedtheopticalRTT-150spec- panion star onto the compact star through the inner trum into two components corresponding to an M2V Lagrangian point has not yet completely started. red dwarf, whose spectrum was taken from the stel- lar spectral flux library of Pickles (1998), and a white dwarf(thespectrumWD0320-539fromtheHSTCAL- 400d j160002.4+331120 (VW CrB) SPEC library2). This model also describes well the The CV VWCrB=Var21CrB was discovered at ultraviolet measurements with the GALEX orbital ul- photographicplatesofthe40-cmSternbergAstronom- traviolet telescope (Martin et al., 2005). ical Institute astrograph at the Crimean Astrophysi- In Fig. 4 we show the B-band light curve of this cal Observatory (Antipin, 1996) as a dwarf nova that CV obtained with Sayan Observatory AZT-33IK tele- exhibits the outbursts with nearly 3 mag amplitude, scope in March 2013. The light curve of this object m = 14.5–17.5, 10–15 days in duration. Subse- phot exhibits variability at a level of about 30% of the flux. quently,theorbitalperiodofabinarysystemwasmea- Therefore, apart from the white and red dwarfs, an sured to be 107 min (Novak, 1997; Ritter and Kolb, appreciable contribution of the emission from the ac- 2003). The spectrum of this system, obtained during cretion flow around the WD must also be present in the outburst, at the object’s magnitude m = 15.8, B the emission from this object near the B photometric can be found in Liu et al. (1999). band. Unfortunately,inSDSSthisobjectturnsouttobein 400d j204720.3+000008 saturationinr(cid:48),i(cid:48)andz(cid:48)bands. However,inFig.3one canseethatther(cid:48)magnitudeis,nevertheless,approxi- This object was discovered in SDSS (Szkody et al., 2004). This CV was identified as a dwaft nova with matelycorrect,anditcanbeusedforroughestimates. 4.475 h orbital period (Ritter and Kolb, 2003). Since the contributions of the WD and the accretion disk to the total flux from the binary system are in- significant in the r(cid:48) band and assuming that the red CONSTRAINTS ON THE CV X-RAY dwarfinourbinarysystemisamain-sequencestar,we LUMINOSITY FUNCTION IN THE SOLAR can roughly estimate the distance. The absolute mag- NEIGHBORHOOD nitudeofM2Vreddwarfinr(cid:48) bandisM ≈10.1,the r(cid:48) As it was discussed above, we detected eight CVs, sourceobservedbrightnessisr(cid:48) =15.47,therefore,the with more or less reliable distance estimates being 2http://www.stsci.edu/hst/observatory/crds/calspec.html available only for two of them. To obtain rough, pre- ASTRONOMY LETTERS Vol. 42 No. 4 2016 246 R. A. Burenin et al. liminary estimates of the constraints on the luminos- ityfunction,letusassumethattheremainingsystems have a low X-ray luminosity. In this case, their ab- solute magnitudes must be near minimum value of M ≈12 which is defined by the presence of a WD in g(cid:48) thesystem(Revnivtsevetal.,2014). Inthesesystems, −3 c theaccretionrateontotheWDmustbelow,andtheir p , orbital periods must be not far from the period min- X L imum. Therefore, this value of Mg(cid:48) agrees well with lg the estimate of M ≈ 11.6 for systems with periods d g(cid:48) / N near the minimum (G¨ansicke et al., 2009). d Note, that signatures of WD emission are clearly seen in the spectra of two systems (400d j001912.9+ 220736 400d j160547.5+240524, see Fig. 2), which is expected for such systems (G¨ansicke et al., 2009). The system 400d j160002.4+331120 is known to have a short orbital period, Porb =107 min, which is quite L (0.5-2 keV), ergs−1 X close to the period minimum (86 min). Therefore, its absolute magnitude should also be close to the above Fig. 5. Constraints on the CV X-ray luminosity function. Our one. constraintsareshownwiththeredcrosses. Inaddition,theblue The X-ray luminosities that are derived for sys- circlesshowthemeasurementsoftheX-rayluminosityfunction temswithunknowndistancesundertheassumptionof basedonRXTEdata(Sazonovetal.,2006)recalculatedtothe 0.5–2 keV energy band under the assumption of a power-law M ≈12 are given in the table and marked by aster- g(cid:48) spectrum dN/dE ∝ E−1.6. The dotted blue crosses indicate isks. Thedistanceestimatesforthesesystemsturnout theconstraintsontheluminosityfunctionfromthetheROSAT to be within the range 200–300 pc. Some of the sys- NorthEclipticPoleSurvey(Pretoriusetal.,2007b). tems(400dj124325.7+025541,400dj152212.8+080338 and 400d j204720.3+000008 ) can actually have a higherluminosity. However,theluminosityofallthese theX-rayfluxbutalsoontheg(cid:48)magnitude. Therefore, systems, on average, cannot be much higher, because, the system 400d j050146.2−035914 (HY Eri), whose inthiscase,theobjectsshouldbelocatedatapprecia- distance can be estimated to be 520 pc, is not taken blylargerdistances,butathighGalacticlatitudesthe in account in the estimates below. space density of sources at distances larger than the The sample incompleteness can be taken into ac- thickness of the Galactic disk drops rapidly. count statistically, appropriately increasing the mea- Basedonthesedata,wecanobtainpreliminarycon- surementerrors. Aswaspointedoutabove,atpresent straints on the CV X-ray luminosity function. For we carried out optical observations for 21 out of the simplicity,wewillassumethatthefluxlimitforX-ray 36 selected objects with magnitudes 19.5 < g(cid:48) < 20, sources in the 400d survey is 2.5·10−14 ergs−1cm−2 with two previously unknown CVs having been de- irrespectiveoftheROSATpointingfield,andthatthe tected among them. Therefore, the number of previ- area overlapping with the SDSS is 262.3 sq. deg. The ously unknown CVs among the remaining 15 objects dependence of the density of sources in the Galactic can be estimated to be, on average, about 1.4. It can disk on the perpendicular distance from the Galac- be seen from the table that these CVs will most likely tic plane, ρ(z) = ρ e−|z|/h, can be taken into ac- have luminosities L ≈1030 ergs−1. 0 X count trough the calculation of generalized volume, Six systems from our sample are found in in the as it was done by Tinney et al. (1993) and Preto- range of X-ray luminosities 3 · 1029–3 · 1030 ergs−1. rius et al. (2007b). We will assume that the expo- Taking into account all mentioned above, we obtain nentialscaleheightoftheGalacticdiskforourCVsis thefollowingconstraintontheCVspacedensitybased 260 pc, which corresponds to the disk scale height for on this sample: ρ = 1.0 ± 0.4 · 10−5 pc−3. Only 0 old, short-period systems (Pretorius et al., 2007a). onesystem,400dj154730.1+071151,withaluminosity It should be also noted that only systems with close to the upper boundary of the X-ray luminosity m < 20 are included in our sample. This constraint range 3 · 1028–3 · 1029 ergs−1 is found in this lumi- g(cid:48) does not affect the selection of systems located at dis- nosity range. In this case, the upper limit on the CV tanceslessthanabout400pc,becausethereisamini- space density for this range is ρ < 4.1·10−5 pc−3 0 mumabsolutemagnitudeM ≈12forCVs(Revnivt- (at 95% confidence). However, this system is appar- g(cid:48) sev et al., 2014). For systems located at larger dis- ently a pre-cataclysmic variable. If such systems are tances, the selection probability depends not only on not considered as CVs, then no object is found in the ASTRONOMY LETTERS Vol. 42 No. 4 2016 CV sample from 400d X-ray survey 247 luminosity range 3·1028–3·1029 ergs−1 and the limit bedetectedinthefutureSpecrtum-R¨oentgen-Gamma on the CV space density is then ρ <2.7·10−5 pc−3. all-sky X-ray survey with eROSITA telescope at high 0 Our constraints on the X-ray luminosity function of Galactic latitudes, which will allow to significantly CVsareshownwithsolidredcrossesinFig.5;theblue improve the measurements of their X-ray luminosity circles indicate the measurements of the X-ray lumi- function at X-ray luminosities L < 1031 ergs−1. X nosity function based on RXTE data (Sazonov et al., Further studies of CV samples at lower luminosities, 2006) recalculated to the 0.5–2 keV energy band un- L < 1029 ergs−1, will possibly allow to detect large X dertheassumptionofapower-lawspectrumdN/dE ∝ number of systems at late post-bounce evolutionary E−1.6; the dotted crosses show the constraints on the stages and to measure their space density. luminosity function from the ROSAT North Ecliptic Pole survey Pretorius et al. (2007b). It can be seen This work was supported by the Russian Founda- from the figure that our constraint on the CV space tion for Basic Research (project no. 13-02-00741-a) density near L ≈ 1030 ergs−1 agrees well with that and grant NSh-6137.2014.2. The calculations of the X fromPretoriusetal.(2007b). Notethatweobtainsig- area of the 400d X-ray survey used here were sup- nificantlystrongerupperlimitontheCVspacedensity ported by RSF grant no. 14-22-00271. The observa- near an X-ray luminosity L ≈1029 ergs−1. tions at the 6-meter BTA telescope were carried out X with the financial support of the Ministry of Educa- tionandScienceoftheRussianFederation(agreement DISCUSSION no. 14.619.21.0004, project ID RFMEFI61914X0004). We present a sample of CVs detected among the X- raysourcesofthe400squaredegree(400d)survey. As REFERENCES itwasshownabove, thisX-raysurveyiswellsuitedto searchforlowX-rayluminosityCVs. Notethatamong 2. S. Alam, F. D. Albareti, C. Allende Prieto, thesystemsselectedin400dsurvey, thereisanappre- F. Anders, S. F. Anderson, T. Anderton, et al., Astrophys. J. Suppl. Ser. 219, 12 (2015). ciable fraction of systems whose spectra clearly show 3. S.V.Antipin,InformationBulletinonVariableStars signaturesoftheWDemission(seeFig.2),whichisex- 4343, 1 (1996). pectedforsystemswithalowaccretionrate(G¨ansicke 4. K. Aizu, Progress of Theoretical Physics 49, 1184 et al., 2009). (1973). Based on this X-ray sample of CVs, we obtained 5. A. Aviles , S. Zharikov, G. Tovmassian, R. Michel, preliminary constraints on the X-ray luminosity func- M.Tapia,M.Roth,V.Neustroev,et al.,Astrophys. tion of CVs in the range of low X-ray luminosi- J. 711, 389 (2010). ties, L ≈ 1029–1030 ergs−1. From the compari- X 6. V.L. Afanasyev, A.V. Moiseev, Pis’ma v Astron. son of our constraints with the luminosity function Zhurn.31,194(2005).;[AstronomyLetters,31,194] at L > 1031 ergs−1 measured previously (Sazonov X 7. V.L. Afanasyev, A.V. Moiseev, Balt. Astr. 20, 363 et al., 2006), we conclude that its logarithmic slope is (2011). less steep at low luminosities (see Fig. 5). 8. V.L. Afanasyev, S.N. Dodonov, V.R. Amirkhanyan, The presented results are preliminary ones. They A.V. Moiseev, Astrophysical Bulletin, in press can be significantly improved if more or less reliable (2016); arXiv:1611.07572. distance estimates will be obtained for a larger num- 9. K. Byckling, K. Mukai, J. R. Thorstensen, berofsystemsandifthestatisticallycompletesample J. P. Osborne, Mon. Not. R. Astron. Soc. 408, 2298 will be expanded to include system with a lower opti- (2010). cal brightness. We are going to continue this work in 10. K. Beuermann, H.-C. Thomas, Advances in Space future. Research 13, (12)115 (1993). Our constraints on the CV luminosity function can 11. K. Beuermann, H.-C. Thomas, K. Reinsch, A. D. Schwope, J. Tru¨mper, W. Voges, Astron. be compared with theoretical estimates of the space Astrophys. 347, 47 (1999). density of these objects. For example, de Kool (1992) 12. V. Burwitz, K. Reinsch, K. Beuermann, H.- estimated the space density of all CVs to be ≈ 0.5– C. Thomas, Annapolis Workshop on Magnetic 2 · 10−4 pc−3 (recalculated to the Galaxy age of 10 Cataclysmic Variables, ASP Conference Series, v. Gyr). As follows from Fig. 5, we can eliminate such a 157, ed. by C. Hellier and K. Mukai, 127 (1999). number of CVs in the investigated luminosity range. 13. R.A. Burenin, A. Vikhlinin, A. Hornstrup et al., This may imply, for example, that a large number of Astrophys. J. Suppl. Ser. 172, 561 (2007). CVs have an even lower X-ray luminosity. It should 14. R.A.Burenin,A.L.Amvrosov,M.V.Eselevich,et al., also be taken in account that the CV luminosity can Pis’ma v Astron. Zhurn. 42, 333 (2016) [Astronomy change in a wide range due to accretion instability. Letters 42, 295 (2016)]. From our estimates of the CV X-ray luminosity 15. Burenin et al., Pis’ma v Astron. Zhurn., in function, we conclude that several thousand CVs will preparation (2016b). ASTRONOMY LETTERS Vol. 42 No. 4 2016 248 R. A. Burenin et al. 16. M. de Kool, Astron. Astrophys. 261, 188 (1992). 41. S. Sazonov, M. Revnivtsev, M. Gilfanov, 17. J. Faulkner, Astrophys. J. 170, L99 (1971). E. Churazov, R. Sunyaev, Astron. Astrophys. 18. . (B. T. Ga¨nsicke, M. Dillon, J. Southworth, 450, 117 (2006). J. R. Thorstensen, P. Rodriguez-Gil, A. 42. S. J. Schmidt, S. L. Hawley, A. A. West, Aungwerojwit, et al.), Mon. Not. R. Astron. J. J. Bochanski, J. R. A. Davenport, J. Ge, and Soc. 397, 2170 (2009). D. P. Schneider, Astron. J. 149, 158 (2015). 19. R.F.Green,D.H.Ferguson,J.Liebert,M.Schmidt, 43. M. F. Skrutskie, R. M. Cutri, R. Stiening, M. Astron. Soc. of the Pacific 94, 560 (1982). D. Weinberg, S. Schneider, J. M. Carpenter, et al., 20. S. L. Hawley, K. R. Covey, G. R. Knapp, Astron. J. 131, 1163 (2006). D. A. Golimowski, X. Fan, S. F. Anderson, et al., 44. J.T.Stoke,J.Liebert,I.M.Gioia,et al.,Astrophys. Astron. J. 123, 3409 (2002). J. 273, 458 (1983). 21. S.B.Howell,L.A.Nelson,S.Rappaport,Astrophys. 45. P. Szkody, S. F. Anderson, M. Agu¨eros, J. 550, 897 (2001). R. Covarrubias, M. Bentz, S. Hawley, et al., 22. C. Knigge, Mon. Not. R. Astron. Soc. 373, 484 Astron. J. 123, 430 (2002). (2006). 46. P. Szkody, A. Henden, O. Fraser, N. Silvestri, 23. C. Knigge, I. Baraffe, and J. Patterson, Astrophys. J. Bochanski, M. A. Wolfe, et al., Astron. J. 128, J. Suppl. Ser. 194, 28 (2011). 1882 (2004). 24. U. Kolb, Astron. Astrophys. 271, 149 (1993). 47. A. Yu. Tkachenko, M.G. Revnivtsev, R.A. 25. S. P. Littlefair, V. S. Dhillon, T. R. Marsh, Burenin, A.V. Meshcheryakov, V.S. Vorobyev, B. T. Ga¨nsicke, J. Southworth, I. Baraffe, M.N. Pavlinsky, Pis’ma v Astron. Zhurn. 41, 196 C. A. Watson and C. Copperwheat, Mon. Not. (2015) [Astronomy Letters 41, 174 (2015)]. R. Astron. Soc. 388, 1592 (2008). 48. C. G. Tinney, I. N. Reid, J. R. Mould, Astrophys. 26. W. Liu, J. Y. Hu, X. H. Zhu, Z. Li, Astrophys. J. J. 414, 254 (1993). Suppl. Ser. 122, 243 (1999). 49. D. M. Townsley, B. T. Ga¨nsicke, Astrophys. J. 693, 27. D. C. Martin, J. Fanson, D. Schiminovich, 1007 (2009). P. Morrissey, P. G. Friedman, T. A. Barlow, et al., 50. E.L. Wright, P.R.M. Eisenhardt,, A.K. Mainzer, Astrophys. J. 619, L1 (2005). M.E. Ressler, R.M. Cutri, T. Jarrett, 28. R. Novak, Information Bulletin on Variable Stars J.D. Kirkpatrick, D. Padgett, et al., Astron. J. 4489, 1 (1997). 140, 1868 (2010). 29. B. Paczynski, Acta Astronomica 31, 1 (1981). 51. M. Zorotovic, M. R. Schreiber, and B. T. Ga¨nsicke, Astron. Astrophys. 536, A42 (2011). 30. A. J. Pickles, PASP 110, 863 (1998). 31. M. L. Pretorius, C. Knigge, U. Kolb, Mon. Not. R. Astron. Soc. 374, 1495 (2007a). 32. M. L. Pretorius, C. Knigge, D. O’Donoghue, J. P. Henry, I. M. Gioia, C. R. Mullis, Mon. Not. R. Astron. Soc. 382, 1279 (2007b). 33. M. L. Pretorius and C. Knigge, Astron. Astrophys. 419, 1442 (2012). 34. M.L.PretoriusandK.Mukai,Mon.Not.R.Astron. Soc. 442, 2580 (2014). 35. M. Revnivtsev, E. Churazov, S. Sazonov, W. Forman, C. Jones, Astron. Astrophys. 490, 37 (2008a). 36. M. Revnivtsev, S. Sazonov, R. Krivonos, H. Ritter, R. Sunyaev, Astron. Astrophys. 489, 1121 (2008b). 37. M. Revnivtsev, S. Sazonov, E. Churazov, W. Forman, A. Vikhlinin, R. Sunyaev, Nature 458, 1142 (2009). 38. M.G. Revnivtsev, E.V. Filippova, V.F. Suleimanov, Pis’mavAstron.Zhurn.40,196(2014).[Astronomy Letters, 40, 177] 39. H. Ritter, U. G. Kolb, Astron. Astrophys. 404, 301 (2003). 40. G. T. Richards, X. Fan, H. J. Newberg, M. A. Strauss, D. E. Vanden Berk, D. P. Schneider, et al., Astron. J. 123, 2945 (2002). ASTRONOMY LETTERS Vol. 42 No. 4 2016