PASJ:Publ.Astron.Soc.Japan,1–??, (cid:13)c 2008.AstronomicalSocietyofJapan. Observations of SiO Maser Sources within a Few Parsec from the Galactic Center Shuji Deguchi Nobeyama Radio Observatory, National Astronomical Observatory, Minamimaki, Minamisaku, Nagano 384-1305 [email protected] Takahiro Fujii Institute of Astronomy, The University of Tokyo, Mitaka, Tokyo 181-0015 [email protected] Makoto Miyoshi 2 VERA Office, National Astronomical Observatory, Mitaka, Tokyo 181-8588 0 [email protected] 0 2 and Jun-ichi Nakashima n Department of Astronomical Science, The Graduate University for Advanced Studies, a J Nobeyama Radio Observatory, Minamimaki, Minamisaku, Nagano 384-1305 7 [email protected] 1 (Received 2001 May 17; accepted 2001 October 24) v Abstract 7 7 Mapping and monitoring observations of SiO maser sources near the Galactic center were made with 0 the Nobeyama 45-m telescope at 43 GHz. Rectangular mapping an area of approximately 200′′×100′′ 1 in a 30′′ grid, and triangular mapping in a 20′′ grid toward the Galactic center, resulted in 15 detections 0 2 of SiO sources; the positions of the sources were obtained with errors of 5–10′′, except for a few weak 0 sources. Three-yearmonitoringobservationsfoundthatthe componentatVlsr=−27kms−1 ofIRS10EE / flaredtoabout1.5Jyduring2000March–May,whichwasafactorofmorethan5brighterthanitsnormal h intensity. UsingtheradialvelocitiesandpositionsoftheSiOsources,weidentified5whicharecounterparts p - of previously observed OH 1612 MHz sources. The other 10 SiO sources have no OH counterparts, but o two were previously detected with VLA, and four are located close to the positions of large-amplitude r t variables observed at near-infrared wavelengths. A least-squares fit to a plot of velocities versus Galactic s longitudes gives a rather high speed for the rotation of the star cluster around the Galactic center. The a : observedradial-velocitydispersionisroughlyconsistentwithavalueobtainedbefore. Itwasfoundthatall v of the SiO sources with OH 1612 MHz counterparts have periods of light variation longer than 450 days, i X while SiO sources without OH masers often have periods shorter than 450 days. This fact suggests that r lower-mass AGB stars are more often detected in SiO masers than in the OH 1612 MHz line. a Key words: Galaxy:center — Galaxy: nucleus — Masers — Stars: late-type 1. Introduction tion of Sgr A* to be pinpointed with an accuracy better than 0.′′1 on near-infrared images. Detections of accel- The Galactic-center star cluster consists of mixed stel- erating motions of stars near the Galactic center fixed lar populations (000 [cite]cite.kra95Krabbe et al. (1995); the mass of the cental object at about 3×106M (000 ⊙ 000 [cite]cite.mor96Morris, Serabyn (1996)). It in- [cite]cite.ghe00Ghez et al. (2000)). Because radio inter- volves a number of late-type stars which are poten- ferometers have a potential of measuring the proper mo- tial candidates for OH/SiO maser emitters. Deep tions of stars relative to Sgr A* [e.g., Sjouwerman et al. surveys in OH 1612 MHz, H O 22 GHz and SiO (1998a)]moreaccuratelythanthepresentoptical/infrared 2 43 GHz masers (000 [cite]cite.sjo98bSjouwerman et al. telescopes (000 [cite]cite.gen96Genzel et al. (1996); 000 (1998b)b; 000 [cite]cite.men97Menten et al. (1997); 000 [cite]cite.eck97Eckart,Genzel(1997)),to detectmoreSiO [cite]cite.izu98Izumiura et al. (1998)) have been made; maser sources near the Galactic center and to investigate a dozen sources have been detected within a few par- their properties will be quite important. sec of the Galactic center. The accurate H O/SiO Using the 45-m telescope at 2 maser positions of these sources observed by the Very Nobeyama, te]cite.izu98Izumiura et al. Large Array (VLA) provided a precise alignment be- (1998) ([cite]cite.izu98Izumiura et al. (1998)) de- tween the near-infraredand radio coordinate frames (000 tected 14 SiO sources around the Galactic center. These [cite]cite.men97Menten et al. (1997)), enabling the posi- observations(by a beam width of about 40′′) provedthat 2 S. Deguchi et al. [Vol. , the SiO maser source density is peaked at the Galactic center. Because this was a set of pointed observations . toward the Galactic center and two one-beam offset 1 positions, the SiO source positions were not determined with an accuracy better than the beam size. To remedy 0.5 the positionaluncertainties to some degree,wemade new mapping observations of SiO masers near the Galactic b(') 0 center using the 45-m telescope in the year 2000. The D present mapping observations on a 30′′ grid can be used -0.5 to derive the source positions with an uncertainty of about 5–10′′, depending on the signal-to-noise ratio. -1 2 1.5 1 0.5 0 -0.5 -1 -1.5 -2 Also, because the intensities of SiO masers are expected D l (') tovarystronglyonatimescaleofoneyear,wemonitored 1 the intensities of SiO masers toward the Galactic center 12 during 1999–2001. During the three-year monitoring 1 0.5 observations,we found an SiO maser flare in the −27 km 14 sp−re1secnotmtphoendenettaoilfsIoRfSth1es0eEoEbseinrva2t0i0o0nsMinartchhi–sJpuanpee.r.We b (') 0 11 5 4 3 D 15 -0.5 13 2. Observations 2 -1 Simultaneous observations of the SiO J = 1–0, v = 1 2 1.5 1 0.5 0 -0.5 -1 -1.5 -2 and 2 transitions at 42.122 and 42.821 GHz, respectively, D l (') weremadewiththe45-mradiotelescopeatNobeyamaon 400 1999June8–12,2000May20–29,and2001February9. A 300 cooled SIS receiver (S40) with a bandwidth of about 0.4 200 GHz was used, and the system temperature (including -1)s 100 atmospheric noise) was 200–250 K (SSB). The aperture m 0 k epffiowcieernbcyeaomf twhiedtthele(HscPopBeWw)awsa0s.6a0boautt4338G′′Hatz.43TGheHhz.alAf- V (lsr --210000 factorof2.9JyK−1 wasusedtoconverttheantennatem- -300 peratureto flux density. An acousto-opticalspectrometer -400 2 1.5 1 0.5 0 -0.5 -1 -1.5 -2 array of low resolution (AOS-W) was used. Each spec- D l (') trometerhad250MHzbandwidthand2048channels,giv- ing a velocity coverageof about 1700 km s−1 and a spec- tral resolution of 1.7 km s−1 (per two binned channels). Observations were made in a position-switching mode, Fig. 1. Mappedpositionsofthetelescope(top),positionsof and the off-position was chosen to be 10′ away from the the 15 detected sources with error bars (middle), and posi- tion–velocity diagramofthedetected sources(bottom). The Galactic center in azimuth. The telescope pointing was coordinates,∆land∆b,indicatepositiondifferencesfromSgr checked using a strong SiO maser source, OH 2.6−0.4. A*inGalacticcoordinatesinunitofarcminute. Theextreme Thecalibrationofthe telescopeantennatemperaturewas sources with |Vlsr|>300 km s−1 are shown as open circles made by observing the intensities of the 29SiO (thermal), in the middle and bottom panels. The numbers in the mid- dle panel correspond to the SiO source number in tables 1 H52α (recombination), U42.767, and SiO J =1–0, v=1 and 2. The solid and broken lines in the bottom panel are (maser) lines toward Sgr B2 MD5 [e.g., te]cite.shi97Shiki least-squares linearfits forallsources andforallbutthe ex- et al. (1997) ([cite]cite.shi97Shiki et al. (1997))]. The to- tremesources,respectively. tal on-source integrationtime was approximately 2 hours per day. this frequency range: U42.767, U43.018, U43.026, and In the 1999 June observations, only the direc- U43.178. These U-lines should appear at V =378 km lsr tion toward Sgr A* was observed; (R.A., Decl., s−1 for the J = 1–0 v = 2 transition, and V = 723, lsr epoch)=(17h42m29.314s, −28◦59′18.3′′, 1950) (000 668, and −390 km s−1 for the J = 1–0 v = 1 tran- [cite]cite.rog94Rogers et al. (1994)). At this time, we sition, if the circumnuclear molecular ring (cf., 000 spentthree dayslookingforextreme velocitycomponents [cite]cite.wri01Wright et al. (2001)) contains a sufficient (|Vlsr| > 350 km s−1); no extreme velocity component number of molecules responsible for these U-lines. The above 0.04 and 0.03 K for the J = 1–0, v = 1 and 2 spectra toward the Galactic center above |V |> 350 km lsr transitions, respectively, was detected in the frequency s−1 were carefully checked for contamination by these range between 42.73–43.35 GHz (approximate velocity lines, but we found no such feature at the corresponding span ranging from −2200 to 2700 km s−1 for the SiO velocities (except for weak features due to 29SiO, H52α, J = 1–0 v = 1 transition). In fact, several unidentified and U42.767 at several10′ offset positions from Sgr A*). lines were found in the spectra toward Sgr B2 MD5 in No. ] SiO Maser Sources near the Galactic Center 3 In the 2000 May observations, we made mapping ob- strong feature. The resulting three-year spectra toward servations in an area covering approximately 200′′×100′′ Sgr A* [(∆l,∆b)=(0,0)] are shown in figure 2. centeredtowardSgrA*. Twodifferentmodesofmapping Detections were judged by criteria similar wereused: 3-point(triangular)mappingwith20′′ separa- to those given in te]cite.izu98Izumiura et al. tion toward Sgr A*, and 9-point (3×3 square) mappings (1998) ([cite]cite.izu98Izumiura et al. (1998)). All toward (∆l,∆b)=(0 or ±60′′, 0′′) on a 30′′ grid. The 9- of the peaks for a single channel with S/N > 3 and point mapping mode was utilized to obtain a reasonably features for several channels with (S/N) > 5 were broad high level of signals in one run of the typical mapping treated as detection candidates. Then, all of the features time of about two hours (approximately 10-minutes inte- were checked as to whether or not they were detected in gration per point). The mapped points are shown in the the spectra of both the v=1 and 2 transitions of SiO, at top panel of figure 1. The large circle in figure 1 indi- nearby positions, or at different epochs of observations. cates the effective beam of the telescope (HPBW ∼40′′). Dubious features were discarded. After these careful Because the telescope pointing has been known to be in- checks, we selected 15 SiO spectral components as fluenced strongly by wind, we made 3-point mapping on confident detections, which are listed in table 1; the those days with wind speed less than 5 m s−1, when the number, telescope positions (relative to Sgr A*), V , lsr telescope pointing was excellent (error < 5′′ accuracy), peak antenna temperatures, integrated intensities, r.m.s. ∼ and 9-point mapping on relatively windy days with wind noise values, and signal-to-noise ratios integrated over speeds of 5–10 m s−1 (approximately < 10′′ accuracy). the emission profile, are given. We show the spectra ∼ Because the grid separation is larger than HPBW in the of the detected sources in figures 3, 4, and 5. The 9-pointmapping, the relativeintensities ofthe SiO maser components at −117 and −337 km s−1 were quite weak components in the grid points were supposed to be kept and only slightly above the critical level of detections in constant,eventhoughthewindvelocitywasslightlyhigh. 2000 May. These components were detected at several Inthe 2001Februaryobservations,we couldmanageonly different positions and epochs. They were, in fact, more a few hours of observation time, and we made observa- clearly detected previously: the −117 km s−1 component tions in the 3-point mapping mode toward Sgr A*; the in 1999, and the −338 km s−1 component in 1997 (000 signal-to-noise ratio of the spectra obtained was not very [cite]cite.izu98Izumiura et al. (1998)). For the −117 km high. Therefore, we averaged the spectra of three posi- s−1 component (No. 9) in figure 3, the spectra taken in tions (a half-beam width each away). They are shown at 1999 are shown. the bottom in figure 2. A shallow survey of SiO sources in the Galactic center 3. Discussion areawasalsomadewiththe45-mtelescopeusingamulti- 3.1. Source Positions and Identifications beam receiver (S40M) prior to the present observations (2000 March–April). This shallow observation detected Becausemostofthecomponentsweredetectedatmore 9 sources in the 7′×13′ area toward the Galactic center than one telescope position, we computed the most likely (000 [cite]cite.miy01Miyazaki et al. (2001)). The present positions of the SiO features from the relative intensities observations concentrated on a smaller and much closer at the several observed positions. We assumed that an- areato the Galactic center with a longerintegrationtime tenna temperature ofa SiO maser componentvaries with perpointandusingamoresensitivesingle-beamreceiver, the angular separationfrom the telescope pointing center S40. accordingto the Gaussianbeamshape with HPBW=40′′ As noted in te]cite.izu98Izumiura et al. (which is taken to be slightly larger than the nominal (1998) ([cite]cite.izu98Izumiura et al. (1998)), the HPBWbecauseofpointingfluctuationduetowind). The AOS-W spectra toward the strong continuum source, most likely position of a component was calculated by Sgr A* (∼11 Jy at 43 GHz; 000 [cite]cite.sof86Sofue et minimizing the sum, Σ[(F −F )/T ]2, where the obs exp rms i al. (1986) ; 000 [cite]cite.bec96Beckert et al. (1996)), sumwasmadeoverthedetectedandundetected(F =0) obs exhibited a baseline distortion of about 0.3 K at the positions, i; F , F , and T are the observed in- obs exp rms maximum, and ripples due to a standing wave in the tegrated intensity of the component, the expected inte- telescope system. The ripples in the velocity range of gratedintensityofthecomponentattheobservedposition ± 350 km s−1 were relatively weak. In order to remove assuming the Gaussian beam shape, and the rms noise these complex ripple features from the spectra, we took temperature, respectively. The errors in ∆l and ∆b are running means of the spectra (average of about 100 also calculated from the distance which gives twice the channels, or about 80 km s−1 width), and the averaged minimum value of the above sum. The resulting most- spectra were subtracted from the originals. With this likely positions obtained are given in table 2; the compo- procedure, the baselines of the resulting spectra became nentnumber,Galacticlongitudeandlatitude,therelative quite flat. Because the SiO maser lines are quite narrow positions from Sgr A*, errors of positions, SiO radial ve- (width less than 10 km s−1) and weak (T < 0.2 K), locity (averaged in J =1–0, v=1 and 2 transitions), the a this method seems to work well, except for a strong line referenceandnameofthepreviouslydetectedsources,the of T ∼ 0.5 K at −27 km s−1 (No. 6 in table 1). An radialvelocity in the literature, and the separationof the a additional baseline adjustment was made by taking a positions from the identified source are given. The ob- parabolicfitto thetroughbetween±40kms−1 nearthis tained positions are plotted in the middle panel in figure 4 S. Deguchi et al. [Vol. , Table 1. ObservedLineParameters. J =1–0, v=1 J =1–0, v=2 No (∆l,∆b)1 V T F T S/N V T F T S/N Date obs lsr a rms lsr a rms (′′, ′′) (kms−1) (K) (Kkms−1) (K) (kms−1) (K) (Kkms−1) (K) (yymmdd) 1 (−90,30) −26.0 0.096 0.368 0.019 7.4 −27.4 0.084 0.496 0.021 7.2 000530 2 (−90,−30) −308.0 0.668 2.764 0.018 55.1 −308.3 0.403 2.173 0.021 33.3 000530 3 (−90,0) −110.2 0.172 0.700 0.019 13.6 −110.1 0.144 0.775 0.023 10.7 000530 4 (−60,0) 6.3 0.109 0.436 0.018 8.8 7.4 0.087 0.296 0.014 8.5 000530 5 (0,0) −335.8 0.031 0.123 0.009 5.2 −335.7 0.044 0.314 0.014 6.1 000525 6 (0,−12) −27.8 0.349 1.228 0.008 59.8 −27.4 0.583 1.732 0.010 72.3 000526 7 (0,0) −60.8 0.102 0.206 0.012 9.0 −62.0 0.109 0.243 0.014 8.6 000525 8 (0,-12) 85.2 0.073 0.265 0.008 12.4 84.8 0.046 0.103 0.010 5.0 000526 9 (10,6) ··· ··· ··· 0.009 ··· −117.3 0.036 0.160 0.010 5.5 000526 10 (−10,6) −12.5 0.051 0.157 0.010 6.9 −13.7 0.044 0.288 0.007 11.8 000526 11 (30,30) 71.7 0.302 1.027 0.022 18.8 71.3 0.249 1.151 0.028 14.4 000530 12 (30,30) 37.7 0.092 0.161 0.014 6.4 37.7 0.078 0.260 0.014 7.4 000530 13 (30,−30) 51.1 0.202 0.951 0.022 15.1 54.4 0.160 0.852 0.025 11.0 000531 14 (90,0) 23.3 0.311 1.037 0.021 20.5 23.2 0.416 1.690 0.026 24.3 000531 15 (90,0) 86.9 0.161 0.686 0.021 12.1 86.5 0.152 0.915 0.026 10.8 000531 1 Observed telescope position. Table 2. Obtainedpositionsandidentifications tothepreviouslydetected objects. No. l b ∆l ∆b (∆l) (∆b) VSiO Ref∗ Identification† VRef ∆r er er lsr lsr (◦) (◦) (′′) (′′) (′′) (′′) (kms−1) (kms−1) (kms−1) (′′) 1 359.915 −0.041 −105 18 7 5 −26.7 3 3–57 6 2 359.919 −0.055 −92 −32 5 5 −308.1 1,3 359.918−0.055,3–2855 −307.9 5 3 359.921 −0.047 −84 −3 5 5 −110.1 2,3 M4 −105.1 4 359.930 −0.045 −47 0 5 7 6.9 2,3 M3,3–88 17.5 6 5 359.944 −0.045 0 4 10 9 −335.8 2 C7 −341.9 6 359.945 −0.047 2 -2 11 11 −27.6 1,2,4 359.946−0.048,C4 −26.4 6 ,IRS10EE 7 359.945 −0.045 2 3 14 12 −61.4 2 C5 −69.6 8 359.946 −0.046 5 −1 18 13 85.0 9 359.946 −0.047 5 −4 12 8 −117.3 2,4 C6,IRS7 −121 7 10 359.946 −0.045 6 3 13 15 −13.1 2,4 C2,IRS15NE −14.6 5 11 359.952 −0.040 28 22 8 5 71.5 1,2 359.954−0.041,P1 70.6 10 12 359.953 −0.032 30 52 6 30 37.7 3 3–885 8 13 359.957 −0.051 46 −19 5 5 52.8 1,2,3,5 359.956−0.050,P2,3–5 48.5 8 14 359.970 −0.043 93 10 8 5 23.3 3 3–6 3 15 359.973 −0.048 105 −8 12 5 86.7 1,3 359.970−0.049,3–3 88.8 8 ∗ References: 1—Sjouwerman et al. (1998b), 2—te]cite.izu98Izumiura et al. (1998) ([cite]cite.izu98Izumiura et al. (1998)), 3—te]cite.gla01Glass et al. (2001) ([cite]cite.gla01Glass et al. (2001)), 4—te]cite.men97Menten et al. (1997) ([cite]cite.men97Menten et al. (1997)), 5—te]cite.lev95Levine et al. (1995) ([cite]cite.lev95Levine et al. (1995)). † The names, 3–57, etc. (indicating the survey field and the star number) refer to the large-amplitude variables in te]cite.gla01Glass et al. (2001)([cite]cite.gla01Glass et al. (2001)). The names of the sources in te]cite.izu98Izumiura et al. (1998) ([cite]cite.izu98Izumiura et al. (1998)) are designated as Mn, Cn, and Pn, where M, C, and P stand the telescope positions at (−40′′,0), (0′′, 0′′), and (+40′′,0), respectively, relative to Sgr A*, and n the number given to the detected SiO source in te]cite.izu98Izumiura et al. (1998) ([cite]cite.izu98Izumiura et al. (1998)). No. ] SiO Maser Sources near the Galactic Center 5 0.4 . . 0.3 S00iO0.5n2o9.1.9 (-90,30) v=2 1.5 S00iO0.5n2o9.2.9 (-90,-30) 0.4 Ta (K)00..12 v=1 Ta (K)0.15 v=2 0.3 G99C0 6 (008,0.9) v=2 0 0 v=1 -0.1 K) 0.2 -100 -50 0 50 -400 -350 -300 -250 -200 Ta ( 0.1 no.9 0.5 0.4 v=1 0.4 S00iO0.5n2o9.3.9 (-90,0) 0.3 S00iO0.5n2o9.4.9 (-60,0) v=2 -0.01 Ta (K)000...123 vv==21 Ta (K)00..12 v=1 -400 -300 -200 -100 0 100 200 300 400 0 0 Vlsr (km s-1) -0.1 -0.1 0.4 -200 -150 -100 -50 0 -100 -50 0 50 100 GC (0,0) 0.3 000525.9 0.5 0.5 Ta (K) 00..12 no. 5 vv==21 Ta (K)0000....1234 S99iO0.6n0o8.9.n0 o (.09, 0) vv==21 no.7 Ta (K)0000....1234 S00iO0.5n2o8.n1.9o1. 1&2 1 2 ( 3 0 , n3o0.)11vv==21 0 0 0 -0.1 -0.1 -200 -150 -100 -50 -50 0 50 100 150 -0.1 -400 -300 -200 -100 0 100 200 300 400 00..34 G01C0 2a0v9e.r2age Vlsr (km s-1) v=2 Ta (K)00000.....12345 S00iO0.5n2o8.1.93 (30,-30) vv==21 Ta (K)00011.....146824 S00iO0n.5no3o10.41.9 4 & 1 5 ( 9n0o,.01)5vv==21 0 0.2 a (K) 0.2 -0.1-50 0 50 100 150 -0.02-50 0 50 100 150 T 0.1 v=1 V (km s-1) V (km s-1) lsr lsr 0 -0.1 -400 -300 -200 -100 0 100 200 300 400 V (km s-1) lsr Fig. 3. SiO maser spectra of the detected sources. The source number shown upper left of each panel corresponds Fig. 2. TimevariationoftheSiOmaser spectratowardSgr to the number intables 1 and 2. The observed positions (in A*. Thedataweretakenon1999June8(top),2000May25 unitofarcsecfromSgrA*)areshowninparentheses. (middle),and2001February9(bottom). Infact,thebottom spectra are averages over the three 12′′-offset spectra taken (1997)) with the VLA; the VLA position agrees with the inthe3-pointmappingmodetoimproveS/N. position obtained in the present paper within an error of 5′′. The positions of the other 4 weaker sources (No. 2, 1 with uncertainty bars. 11, 13, and 15) also coincide well with the positions of The identifications were made by using both the po- theidentifiedOHsourceswithinaccuraciesof3–10′′,thus sitions and radial velocities. A list of the previously verifying this method for weaker sources. detected OH 1612 MHz sources near the Galactic cen- The radial velocity of SiO source No. 4, V = ter (000 [cite]cite.sjo98bSjouwerman et al. (1998b)b) lsr 7 km s−1, is close to the center velocity of the OH was used for the identification. The positions for these 1612 MHz double peaks of OH 359.931−0.050, (000 OH sources are known to an accuracy of better than [cite]cite.sjo98bSjouwermanetal. (1998b)b)atV =17.5 ∼ 1′′. Previous identifications of SiO sources with OH lsr km s−1 (the expansion velocity is about 14 km s−1). sources were also made by te]cite.izu98Izumiura et al. Because the SiO position obtained is separated from the (1998) ([cite]cite.izu98Izumiura et al. (1998)). No. 6 in OHpositionby11′′, itisprobablyadifferentsource. The table 2 is the OH/IR source, 359.946−0.048(IRS 10 EE; positionofSiO sourceNo. 8 (85kms−1)agreeswellwith 000[cite]cite.men97Mentenetal. (1997)),whichhasara- thatofOH359.947−0.046,whichisknowntoanaccuracy dial velocity of −27 km s−1. The location of this source of4′′ (butnotethatthereisalargeuncertaintyintheSiO is well known from near-infrared,OH, and SiO maser ob- position). However,the radialvelocity ofthe SiO masers, servations (000 [cite]cite.men97Menten et al. (1997); 000 85 km s−1, is slightly outside of the OH double-peak ve- [cite]cite.sjo98bSjouwerman et al. (1998b)b). The posi- locities (89.4 and 105.3 km s−1). Therefore, we left these tion obtained in the present paper agrees well with the two SiO sources unassigned. OH position within an accuracy of about 6′′. The high ItisnotsurprisingthatmorethanhalfoftheSiOmaser signal-to-noise ratio of this component confirms that this sources have no OH 1612 MHz counterpart. Previous method for determing the positions of SiO masers works studies of SiO masers near the Galactic center (000 nicely. Source No. 10 with V =−13 km s−1 has been lsr [cite]cite.lin91Lindqvist et al. (1991)) and in the inner identifiedasIRS15NE(000[cite]cite.men97Mentenetal. bulge (e.g., Deguchi et al. 2000a,b) revealed that ap- 6 S. Deguchi et al. [Vol. , proximately2/3ofthe SiOsourceshadnotbeendetected 0.8 previously in spite of a very sensitive OH 1612 MHz sur- vey (000 [cite]cite.lin92bLindqvist et al. (1992b)b; 000 Galactic Center no.6 SiO J=1-0 v=1 2000.05.30 [cite]cite.sev97Sevenster et al. (1997)) with the VLA and 0.6 the ATCA. Large-amplitude variables (including long-period and semiregular variables, as well as supergiants) are poten- (0,-12) tial candidates for SiO maser emitters. We compared 0.4 the positions of SiO sources without OH identification with the knownpositions ofthe large-amplitudevariables Ta (K) no. 5 no.7 no.10 no.8 (-10,6) within12′ oftheGalacticcenter(000[cite]cite.gla01Glass 0.2 et al. (2001)). The positions of these variables were measured with the K-band array camera, and are of no.11 (10,6) a few arcsec accuracy. We found that 4 SiO sources (No. 1, 4, 12, and 14) are located near to these long- 0 period variables within the estimated positional uncer- no.13 tainty. These identifications are given in columns 9 and 10 in table 2. Because the radial velocities of these -0.2 large-amplitude variables are not known, the identifica- -400 -300 -200 -100 0 100 200 300 400 tions of these sources may be slightly less certain than V (km s-1) the OH identifications. So far, SiO sources No. 3, 5, 7, lsr and 8 have no OH 1612 MHz counterpart and no cor- Fig. 4. SiO spectra in the J =1–0 v=1 transition at the responding large-amplitude variable; nevertheless, these three12′′-offsetpositionsaroundSgrA*. Thepositionoffsets were detected before in SiO by te]cite.izu98Izumiura et fromSgrA*inunitsofarcsecareshownontherightside. The al. (1998) ([cite]cite.izu98Izumiura et al. (1998)). component number inthis figurecorresponds to the number We also checked the corresponding near-infrared ob- givenintables 1and2. jects in the 2MASS image server for unidentified sources. However, because the star density within 30′′ of the Galactic center is too high [e.g., te]cite.blu96Blum et al. (1996)([cite]cite.blu96Blumetal. (1996))],itis quitedif- ficult to identify the SiO sources in this region. On the 2MASSimages,wecouldonlyfindoneredcandidate(∼10 magintheK-band)forsourceNo. 3withina5′′ errorcir- cle. This redstar,however,accompaniesa faintextended (∼5′′) feature which is considerably elongatedin galactic 1 longitude, probably because of the coalescence of several no.6 stars due to the low spatial resolution of the 2MASS im- 0.8 Galactic Center SiO J=1-0 v=2 ages. 2000.05.30 3.2. Velocity Distribution of the SiO Masers Sources 0.6 Thebottompaneloffigure1showsalongitude-velocity (0,-12) diagram for the 15 detected sources. The positions given no.8 No.9 ifint ttoabtlhee2SiwOerreaduisaeldvefloorcimtiaeskiwnagstmhiasdde,iaagnrdamth.eAresluinlteaisr Ta (K) 0.4 no. 5 no.10 (-10,6) given in the first row of table 3. There are two extreme no.7 sourceswithverylargenegativevelocities,−336and−308 0.2 km s−1. It is possible that the sources with very large no.11 radial velocities are bulge stars with a very small angu- (10,6) lar momentum, rather than stars in the Galactic-center 0 stellar cluster (000 [cite]cite.van92van Langevelde et al. no.13 (1992); 000 [cite]cite.izu95Izumiura et al. (1995)). They may be located far from the Galactic center in distance, -0.2 -400 -300 -200 -100 0 100 200 300 400 buthappentobeseenneartoitinprojection,thoughthis V (km s-1) possibility is small. We also made a linear fit excluding lsr thesetwosources. Theresultsarealsogiveninthesecond Fig. 5. Sameasfigure4,butfortheJ=1–0,v=2transition. row of table 3. Theslopeofabout1.1(±0.5)kms−1perarcsec(or3890 kms−1 deg−1, ora rotationperiod of2.3×105 yr around the Galactic center) is a factor of a few larger than the No. ] SiO Maser Sources near the Galactic Center 7 previouslyobtainedvaluesfromSiOandOHmaserobser- thecharacteristicbindingenergiesarepositive. Therefore, vations on much larger scales [e.g., te]cite.miy01Miyazaki these sources are not dynamically bound to the central et al. (2001) ([cite]cite.miy01Miyazaki et al. (2001))]. compact object. Even for the sample excluding the 2 extreme-velocity Because the SiO intensity of the extreme object No. 5 sources, the slope is ∼2290 km s−1 deg−1. The was quite weak in the year 2000,the position uncertainty present value of the slope is not compatible with the given in the present paper is quite large. The charac- previously obtained low value by te]cite.izu98Izumiura teristic binding energy of object No. 5 could be posi- et al. (1998) ([cite]cite.izu98Izumiura et al. (1998)) tive if the true position is a few arcsec further away from (∼0.09 km s−1 per arcsec), probably because the po- the Galactic center than that observed. The SiO inten- sitions of the sample in te]cite.izu98Izumiura et al. sity of this component was much stronger in 1997 (000 (1998) ([cite]cite.izu98Izumiura et al. (1998)) were not [cite]cite.izu98Izumiura et al. (1998)). Therefore, a more knowntoanaccuracybetterthanthetelescopebeamsize accurate position should be obtained at a future date. of about 40′′; the positional accuracy was considerably 3.3. Time Variation better in our present observation. The obtained high ro- tational speed of the SiO maser cluster is, however, sim- The intensities of SiO maser lines change significantly ply the slope of the least-squares fit, and must be inter- on a time scale of one year. Though the number of SiO preted very carefully. Non-circular motions of stars and sources detected in this study was similar to the num- the gravitational potential made from the central com- ber given previously (000 [cite]cite.izu98Izumiura et al. pact mass and the nuclear star cluster are responsible (1998)), the two lists are not completely the same; 9 ob- for the observed velocity structure. The high rotational jects in the SiO spectra in 2000 are inferred to be the speedofthese starsis comparableto the rotationalveloc- same as those found in 1997 because of velocity coinci- ityofthe circumnucleargasringwhichhasbeenobserved dences. The identifications are givenin columns 9 and 10 in various molecular lines [e.g., te]cite.wri01Wright et al. of table 2. If we compare figure 1 in the present paper (2001) ([cite]cite.wri01Wright et al. (2001))]. It is pos- with the middle panel of figure 1 of te]cite.izu98Izumiura sible that these rapidly rotating AGB stars were born as et al. (1998) ([cite]cite.izu98Izumiura et al. (1998)), we a result of star formation in the circumnuclear ring (000 can recognize significant variations in SiO maser intensi- [cite]cite.lev95Levine et al. (1995)). ties during the last 5 years. Thevelocitydispersionfromtheaveragelinearfitisap- Wenoticedin2000MarchthattheSiOmasersofsource proximately 108 km s−1 (or 54 km s−1 excluding the ex- No. 6 (the −27km s−1 component, knownasIRS 10 EE; treme sources). This value gives the mass of the Galactic 000[cite]cite.men97Mentenetal. (1997))hadflaredupto center (integrated to 100′′; ∼ 4.1 pc) as ∼ 1×107M , morethanT ≃0.5K(∼1.5Jy). TheSiOmaserintensities ⊙ a when we use the virial theorem, and is consistent with of source No. 6 are plotted against time in figure 6. We previous estimates of the mass of the Galactic center took the 1996 and 1997 data from te]cite.men97Menten region (000 [cite]cite.kra95Krabbe et al. (1995); 000 et al. (1997) ([cite]cite.men97Menten et [cite]cite.mor96Morris, Serabyn (1996)). The radial ve- al. (1997)) and te]cite.izu98Izumiura et locity dispersion obtained for SiO maser sources, 110 km al. (1998) ([cite]cite.izu98Izumiura et al. s−1, is slightly smaller than the value of 154 km s−1 for (1998)), respectively, and the 2000 March– the late-type stars within 12′′ from the Galactic center April data from te]cite.miy01Miyazaki et al. (000 [cite]cite.kra95Krabbe et al. (1995)). In fact, if we (2001) ([cite]cite.miy01Miyazaki et al. (2001)). take the SiO sources only within a 20′′ radius from the The peak and integrated intensities of the −27 Galactic center in our sample (No. 5–10; including the km s−1 component of te]cite.miy01Miyazaki et al. extremesourceNo. 5),wegetavelocitydispersionof133 (2001) ([cite]cite.miy01Miyazaki et al. (2001)) agree km s−1, giving a reasonable agreement with the result of quite well with the 2000 May results in this work. te]cite.kra95Krabbe et al. (1995) ([cite]cite.kra95Krabbe Therefore, the flare lasted for more than two months, et al. (1995)). from the end of 2000 March to the end of 2000 May. Asimplejudgementonwhetherornotaparticularstar Because we did not observe the Galactic center in 1998, isdynamicallyboundtotheGalactic-centermassivecom- it is not known if there was any brightening in 1998. pact object can be obtained from the characteristicbind- Near-infrared monitoring observations of this source ingenergyperunitmass,E =(1/2)V2 −GM/r . Here, over an interval of 4 years (000 [cite]cite.woo98Wood c l.o.s. p Vl.o.s., rp, G, and M are the observed radial velocity, the et al. (1998); assigned as LWHM65) gave a period projectedradiusfromthegalacticcenter,thegravitational of 736 days for the light variation. Extrapolating the constant,and the massof the centralcompactobject (for fit of the light curve given in te]cite.woo98Wood et which we adopt 2.8×106 M ; 000 [cite]cite.ghe00Ghez al. (1998) ([cite]cite.woo98Wood et al. (1998)), we ⊙ et al. (2000); 000 [cite]cite.gen00Genzel et al. (2000)), estimate that the infrared maxima of this source came respectively. Considering a perpendicular velocity com- around mid November of 1998 and late November of ponent and some depth along the line of sight, the char- 2000. However, because the observed light curve (000 acteristic binding energy, E , gives a lower limit to the [cite]cite.woo98Woodetal. (1998))hasalargeambiguity, c real binding energy when the central mass dominates the the estimated time of the light maximum is quite uncer- gravitationalfield. FortheSiOsources(No. 2,3,and15), tain. It may have occured in advance by several months 8 S. Deguchi et al. [Vol. , Table 3. Fittingresults. Sample Number Best fit r.m.s. of sources (km s−1) (km s−1) All sources 15 −42.0(±30.0)+1.07(±0.51)[∆l(′′)] 108.4 (including extreme sources) Low-velocity sources only 13 −3.7(±16.3)+0.67(±0.28)[∆l(′′)] 53.9 (excluding extreme sources) 2 periods is shown in figure 7. Because the sample in- volves only 8 stars, we divided it into two bins, i.e., Time variation of v=2 -1m s) 1.5 the S i O (sIRouCr1c0eE NEo). 6 aabllosvoeuarcnedsbweiltohwO4H50cdoauynst.erIptaristswhoarvthe mpeerniotidosnionfgmthoaret k x (K 1 than 450 days (see figure 7). Note that the average Flu v=1 period of the sample with 412 large-amplitude variables ne 0.5 near the Galactic center given by te]cite.gla01Glass et al. Li (2001) ([cite]cite.gla01Glass et al. (2001)) is 427 days. 0 Therefore,the Galactic-center SiO source sample is prob- 1996 1997 1998 1999 2000 2001 2002 ably weighted more toward those variables with longer Date (yr) periods than the average. SiO masers are, however,occa- sionally detected in stars with shorter periods than those Fig. 6. Time variation of the SiO lines from No. 6 (IRS in which OH masers were found. Considering the mass– 10 EE). The data of 1996 and 1997 were taken from te]cite.men97Menten et al. (1997) ([cite]cite.men97Menten luminosity–period relation [e.g., te]cite.vas93Vassiliadis, et al. (1997)) and te]cite.izu98Izumiura et al. Wood (1993) ([cite]cite.vas93Vassiliadis, Wood (1993))], (1998) ([cite]cite.izu98Izumiura et al. (1998)). The we conclude that lower-mass AGB stars are more often data of 2000 March were taken from te]cite.miy01Miyazaki detected in SiO maser observations than in OH 1612 et al. (2001) ([cite]cite.miy01Miyazaki et al. (2001)). The filledandunfilledcirclesindicatetheSiOJ=1–0,v=1and MHz observations. Because the size of the sample is 2 lines, respectively. The line fluxes in 1997 and 1999 were slightly small, the conclusion may not be free from sta- scaled bya factor of 1.4 because IRS10 EE islocated at an tistical fluctuations. However, the present conclusion is offsetofabout10′′ fromSgrA*. quite consistent with the finding by te]cite.gla01Glass et al. (2001) ([cite]cite.gla01Glass et al. (2001)) that before 2000 November (in fact, 2000 March–June), if we the OH 1612 MHz sources have longer periods than the assume that the light maximum occurs at the same time average period in their sample. According to an SiO as the SiO maser intensity maximum. maser study of the Galactic disk IRAS sources (000 The intensities of the SiO lines from IRS 10 EE, Ta≃ [cite]cite.nak01Nakashima, Deguchi (2001)), stars with 0.5 K (about 1.5 Jy), in 2000 May were comparable bluercolorsintermsoftheIRAS25/12µmintensityratio to the intensity of the SiO J = 1–0 v = 1 maser from (relatively thin dust envelope)tend to be detected in SiO Sgr B2 MD5 (000 [cite]cite.shi97Shiki et al. (1997); 000 masers more often than in OH 1612 MHz masers. This [cite]cite.mor92Moritaetal. (1992)). Iftheyarescaledto fact also seems to agree qualitatively with the above re- adistanceof500pc(∼400Jy),theyarecomparablewith sult for the Galactic-center AGB stars, though the IRAS theintensitiesoftheOrionSiOmasers. Therefore,theSiO 25/12µmcolorsdonotnecessarilycorrelatewellwiththe maser flare of IRS 10 EE may indicate that the mass-loss periodsofAGBstars(000[cite]cite.whi91Whitelocketal. rateofthisstaristemporarilyenhancedtoratescompara- (1991); 000 [cite]cite.nak00Nakashima et al. (2000)). bletothosefromOrionIRc2andSgrB2MD5. Iftheflare The total number of SiO sources detected in the of IRS 10 EE is repeatedperiodically, it is probably asso- present paper is 15. If we include all of the SiO ciated with a pulsation activity of the central star. If it sources which were detected by te]cite.izu98Izumiura et is irregular,it is possible to consider another mechanism, al. (1998) ([cite]cite.izu98Izumiura et al. (1998)) and in for example, a wind–wind collision between nearby mas- the present work, the total number of SiO maser stars in sivestars(000[cite]cite.yus92Yusef-Zadeh,Melia (1992)). the region of 200′′×100′′ from the Galactic center is 20. Inaddition,atidaleffectduetoacloseencounterbetween FurthermonitoringobservationsoftheSiOmaserintensi- nearby stars might not be negligible, because the density ties ofthe these Galactic centerSiO sourcesaredefinitely of stars in the central star cluster is quite high. required. For the 8 SiO sources (No. 1, 2, 4, 6, 12, 13, 14, and 15), the periods of light varia- tion are known (000 [cite]cite.blo98Blommaert et al. (1998); 000 [cite]cite.woo98Wood et al. (1998); 000 [cite]cite.gla01Glass et al. (2001)). A histogram of the No. ] SiO Maser Sources near the Galactic Center 9 [Blum et al. (1996)] Blum, R.D.,Sellgren, K.,& DePoy,D.L. 1996, ApJ,470, 864 [Deguchiet al. (2000a)] Deguchi, S., Fujii, T., Izumiura, H., 7 Kameya, O., Nakada, Y., Nakashima, J., Ootsubo, T., & Ukita, N., 2000a, ApJS,128, 571 6 [Deguchiet al. (2000b)] Deguchi, S., Fujii, T., Izumiura, H., Kameya, O., Nakada, Y., & Nakashima, J. 2000b, ApJS, 5 130, 351 [Eckart, Genzel (1997)] Eckart, A., & Genzel, R. 1997, er 4 With OH MNRAS,284, 576 b [Genzel et al. (2000)] Genzel, R., Pichon, C., Eckart, A., m Nu 3 [GenzGeelrehtaardl.,(O19.9E6.),]&GOentzt,elT,.R20.,00T, MhaNttRe,ASN,.3,17K,r3a4b8be, A., Kroker, H., & Tacconi-Garman, L. E. 1996, ApJ, 472, 153 2 [Ghez et al. (2000)] Ghez, A. M., Morris, M., Becklin, E. E., Tanner, A., & Kremenek, T. 2000, Nature,407, 349 1 [Glass et al. (2001)] Glass, I.S.,Matsumoto,S.,Carter,B. S., & Sekiguchi,K. 2001, MN, 321, 77 0 [Izumiuraet al. (1998)] Izumiura, H., Deguchi, S., & Fujii, T. 0 — 450 > 450 1998, ApJ,494, L89 [Izumiuraet al. (1995)] Izumiura, H., Deguchi, S., Hashimoto, Period (day) O., Nakada, Y., Onaka, T., Ono, T., Ukita, N., & Yamamura, I. 1995, ApJ, 453, 837 [Krabbeet al. (1995)] Krabbe, A,, Genzel, R., Eckart, A., Fig. 7. Histogram of the periods for the SiO detected Najarro, F., Lutz, D., Cameron, M., Kroker, H., Tacconi- sources. TheblankandshadowpartsindicatetheSiOsources Garman, L. E., et al. 1995, ApJ, 447, L95 with and without OH counterparts. All of the SiO sources [Levineet al. (1995)] Levine, D. A., Figer, D. F., Morris, M., withOHcounterparts arestarswithperiodslongerthan450 & McLean, I. S., 1995, ApJ, 447, L101 days. [Lindqvist et al. (1991)] Lindqvist, M., Winnberg, A., Johansson, L. E. B., & Ukita, N. 1991, A&A, 250, 4. Conclusion 431 [Lindqvist et al. (1992a)] Lindqvist, M., Habing, H. J., & We have made mapping and monitoring observations Winnberg, A.1992a, A&A,259, 118 of SiO maser sources near the Galactic center and have [Lindqvist et al. (1992b)] Lindqvist, M., Winnberg, A., detected 15 SiO sources. Approximate positions were ob- Habing, H. J., & Matthews, H.E. 1992b, A&AS,92, 43 tained with accuracies of about 5–10′′; five sources were [Menten et al. (1997)] Menten, K. M., Reid, M. J., Eckart, A., identified with the previously observed OH 1612 MHz & Genzel, R. 1997, ApJ, 475, L111 sources. Among the sources without OH counterparts, [Miyazaki et al. (2001)] Miyazaki, A., Deguchi,S., Tsuboi, M., Kasuga, T., & Takano, S., 2001, PASJ,53, 501 fourarecloseto the positionsofthe large-amplitudevari- [Morris, Serabyn (1996)] Morris, M., & Serabyn, E. 1996, able stars observed at near-infrared wavelengths and two ARA&A,34, 645 to previously detected SiO sources with accurate posi- [Morita et al. (1992)] Morita, K-I., Hasegawa, T., Ukita, N., tions from the VLA. Three-year monitoring observations Okumura,S. K., & Ishiguro, M. 1992, PASJ44, 373 of these objects found that SiO masers from IRS 10 EE [Nakashima et al. (2000)] Nakashima, J., Jiang, B. W., flared up by a factor of more than 5 during March–May Deguchi, S., Sadakane, K., & Nakada, Y., 2000, PASJ 52, 2000. A least-squares linear fit of the velocities to the 275 Galacticlongitudeinthelongitude–velocitydiagramgives [Nakashima, Deguchi (2001)] Nakashima, J., & Deguchi, S. a high rotational speed for the star cluster around the 2001, submitted to PASJ Galactic center. [Rogers et al. (1994)] Rogers, A. E. E., Doeleman, S., Wright, The authors thank I. Glass for stimulating discussions M. C. H.,Bower, G.C., Backer, D.,C., Padin, S.,Philips, J. A.,Emerson, D. T., et al. 1994, ApJ, 434, L59 and reading the manuscript. They also thank M. Morris, [Sevensteret al. (1997)] Sevenster, M. N., Chapman, J. M., H. Izumiura, H. Imai, and A. Miyazaki for comments. Habing, H. J., Killeen, N. E. B., & Lindqvist, M. 1997, ThisresearchwaspartlysupportedbyScientificResearch A&AS,122, 79 Grant (C2) 12640243 of Japan Society for Promotion of [Shikiet al. (1997)] Shiki, S., Ohishi, M., & Deguchi, S. 1997, Sciences. ApJ, 478, 206 [Sjouwerman et al. (1998a)] Sjouwerman, L. O., van References Langevelde, H. J., & Diamond, P. J. 1998a, A&A, 339, 897 [Beckert et al. (1996)] Beckert, T., Duschl, W. J., Mezger, P. [Sjouwerman et al. (1998b)] Sjouwerman, L. O., van G., & Zylka,R. 1996, A&A,307, 450 Langevelde, H. J., Winnberg, A., & Habing, H. J. [Blommaert et al. (1998)] Blommaert, J. A. D. L., van der 1998b, A&AS,128, 35 Veen,W.E.C.J.,vanLangevelde,H.J., Habing,H.J.,& [Sofue et al. (1986)] Sofue, Y., Inoue, M., Handa, T., Tsuboi, Sjouwerman, L. O., 1998, A&A,329, 991 M., Hirabayashi, H., Morimoto, M., & Akabane, K. 1986, 10 S. Deguchi et al. [Vol. , PASJ, 38, 475 [van Langevelde et al. (1992)] vanLangevelde,H.J.,Brown,A. G. A., Lindqvist, M., Habing, H. J., & de Zeeuw, P. T. 1992, A&A,261, L17 [Vassiliadis, Wood (1993)] Vassiliadis, E.,&Wood,P.R.1993, ApJ, 413, 641 [Wood et al. (1998)] Wood, P.R., Habing, H.J., & McGregor, P. J. 1998, A&A,336, 925 [Whitelock et al. (1991)] Whitelock, P., Feast, M., & Catchpole, R. 1991, MNRAS,248, 276 [Wright et al. (2001)] Wright, M. C. H., Coil, A. L., McGary, R. S.,Ho, P. T. P., & Harris, A.I. 2001, ApJ, 551, 254 [Yusef-Zadeh, Melia (1992)] Yusef-Zadeh,F.,&Melia,F.1992, ApJ, 385, L41