Spatial regularity of the young stellar population in the ring of NGC 6217
aa r X i v : . [ a s t r o - ph . GA ] J u l Astronomy & Astrophysicsmanuscript no. 38530corr c (cid:13)
ESO 2020July 22, 2020 L etter to the E ditor Spatial regularity of the young stellar population in the ringof NGC 6217
Alexander S. Gusev ⋆ and Elena V. Shimanovskaya Sternberg Astronomical Institute, Lomonosov Moscow State University, Universitetsky pr. 13, 119234 Moscow, RussiaReceived May 29, 2020; Accepted July 15, 2020
ABSTRACT
The relative contribution of various physical processes to the spatial and temporal distribution of molecular clouds and star-formingregions in the disks of galaxies has not yet been the subject of extensive study. Investigating the spatial regularity in the distributionof the young stellar population in spiral and ring structures is a good test for studying this contribution. In this paper, we look at thephotometric properties of the ring and spiral arms in the barred spiral galaxy NGC 6217 based on an analysis using
GALEX ultraviolet,optical
UBVRI , and H α surface photometry data. The ring in the galaxy is located near the corotation area. We found evidence ofspatial regularity in the distribution of the young stellar population along the galaxy ring. The characteristic scale of spacing is about700 pc. At the same time, we did not find a similar regularity in the distribution of the young stellar population along the spiral armsof NGC 6217. The spatial regularity in the concentration of young stellar groupings along spiral arms is a quite rare phenomenon andit has never previously been seen in galactic rings. Key words.
Galaxies: individual: NGC 6217 – Galaxies: star formation – Galaxies: ISM
1. Introduction
Various physical processes, such as gravitational collapse andturbulence compression, are believed to play a main role in thecreation of molecular clouds and their further evolution intostar formation regions. The sizes of these structures and theirspatial distribution in galactic disks can be explained in termsof gravitational or magnetogravitational instability (see, e.g.,Elmegreen & Elmegreen 1983; Elmegreen 1994b). However, theinfluence of the magnetic field and shock waves on the spatialdistribution of gas clouds and young stellar groups is still notfully understood. Galactic spiral arms and rings are good labo-ratories for studying the relative e ff ect of di ff erent physical pro-cesses on the spatial distribution of star-forming regions.The regular spatial distribution of young star complexesis a rather unique phenomenon which was visually detectedby Elmegreen & Elmegreen (1983) in spiral arms of only 22grand design galaxies. We note that among these galaxies,present in only seven stellar systems, the regular chains ofstar complexes are observed in both spiral arms. The data ofElmegreen & Elmegreen (1983) remains the only published re-sult of a systematic search for regular chains of stellar complexesin spiral arms.In later years, Efremov (2009, 2010, 2011) found theregularities in the distribution of H ii superclouds in spiralarms of our Galaxy and in the distribution of star com-plexes in M31. Gusev & Efremov (2013) confirmed the resultof Elmegreen & Elmegreen (1983), finding that only one of thespiral arms in NGC 628 has a regular chain of star complexeswith a separation of ∼ . ≈
400 pc) betweenadjacent fainter star-formation regions in both spiral arms of thegalaxy. Finally, Elmegreen et al. (2018) found regularly spaced ⋆ e-mail: [email protected] Table 1.
Basic parameters of NGC 6217.
Parameter ValueType (R’L)SB(rs)bTotal apparent B magnitude ( B t ) 11.89 magAbsolute B magnitude ( M B ) a − .
45 magInclination ( i ) 33 ◦ Position angle (PA) 162 ◦ Apparent corrected radius ( D ) b D ) b d ) 20.6 MpcGalactic absorption ( A ( B ) Gal ) 0.158 mag
Notes. ( a ) Absolute magnitude of a galaxy corrected for Galactic ex-tinction and inclination e ff ect. ( b ) Isophotal diameter (25 mag arcsec − in the B -band) corrected for Galactic extinction and absorption due tothe inclination of NGC 6217. infrared peaks in the dusty spiral arms of M100 with a typicalspacing between the clumps of ∼
410 pc.Along with spiral arms, galactic rings are the product of evo-lutionary processes taking place in galactic disks. These appearto be related to specific orbital resonances with the pattern speedof a bar (Buta 2017). Unlike arms, the physical and dynamic pa-rameters of medium (gas density, temperature, pressure, rotationvelocity) along rings are likely to be constant. On the other hand,strong shock waves are not expected to be present in rings. Thesefeatures make them interesting objects for the study of the spatialdistribution of star-forming regions.Based on general ideas about physics and evolution of rings,investigating spatial regularity in the distribution of star clus-ters, as well as in arms, is not an impossible undertaking.Elmegreen (1994a) theoretically described the possibility of theexistence of regularly spaced starburst spots in galactic rings.
Article number, page 1 of 5 & Aproofs: manuscript no. 38530corr
Fig. 1.
Composite image of NGC 6217 in the U (blue), B (green) pass-bands, and H α line (red) with the logarithmic spiral segments and thering (yellow lines) overlaid. The yellow-black-yellow circles on thecurves of the arm and ring correspond to the projected positions of thelocal maxima of brightness. The areas of flux measurements in aper-tures are bounded by magenta ellipses (segments). Examples of ellipti-cal apertures are shown in magenta. The size of the image is 120 × . North is upward and east is to the left. However, this had not been seen in observations before. OnlyArtamonov et al. (1999) noted quasi-periodic characteristic dis-tances of 1 . ± . ≈ A ( B ) Gal , taken from the NED database, and the remaining parameters are taken from theLEDA data base. The adopted value of the Hubble constant isequal to H =
75 km s − Mpc − . With the assumed distance toNGC 6217, we estimate a linear scale of 100 pc arcsec − . We http: // ned.ipac.caltech.edu / http: // leda.univ-lyon1.fr / note that the kinematic positional angle of the galaxy (287 ◦ )is significantly di ff erent from the photometric one (Font et al.2019). However, given the low inclination of NGC 6217, thisdoes not substantially a ff ect our results.
2. Observational data
The results of the
U BVRI photometry for the galaxy were pub-lished in Gusev et al. (2015). We also use the H α image, ob-tained in the GHASP survey (Epinat et al. 2008) and downloadedfrom the NED data base, in our study. The absolute calibrationof this image was carried out using the parameters of descrip-tors from the original FITS file (image) and their explanations inEpinat et al. (2008). We already used H α data for NGC 6217 inGusev et al. (2018).Ultraviolet GALEX
NUV-reduced FITS-images ofNGC 6217 were downloaded from the B. A. MiculskiArchive for space telescopes. The observations were carriedout on November 11, 2011, with a total exposure of 96 s.The description of the
GALEX mission, parameters of NUVpassband, files description, and data reduction are presented inMorrissey et al. (2005).Images of the galaxy in all bands were transformed to theface-on position using the values of i and the position angle (PA)from Table 1. Those face-on images were used for the furtheranalysis. Photometric magnitudes are corrected for Galactic ex-tinction and for projection e ff ects: m cor = m obs . A Gal cos( i ).Our image resolution is equal to 1.7 arcsec for U , 1.4 arcsec for B and V , 3.3 arcsec for H α , and 5.3 arcsec for NUV.
3. Results
For the next steps in our analysis, we used a technique developedby us in Gusev & Efremov (2013). In the first step, we define thegeometric parameters of the spiral arms and the ring by fittingthe curves of both main arms and the ring, which are clearlyoutlined in the optical bands (see Fig. 1). The arms and the ringare defined by a visual selection of pixels in the part of theseimages that belongs to the center lines of the spiral arms or thering. Pixels in this region are then fitted with a logarithmic spiralfor the arms or a circle for the ring using linear least-squares.Here we assume that the pitch angle µ is constant along the armand equal for both spirals.Using the classic equation for the logarithmic spiral r = r exp ( k ( θ − θ )), where k = tan µ , we obtained the followingparameters for the spiral arms, r = .
65 arcsec (960 pc) and k = . ± .
07, which corresponds to a pitch angle µ = . ± . ◦ .The zero point angle θ = ◦ for the eastern arm and 210 ◦ forthe western arm (where θ is counted from north towards east, thesame as the PA; see Fig. 1). As a result of the fitting, we alsoobtained a radius of the ring r ring = ± . ± . r and PA) are close to those ob-tained by Comerón et al. (2014). The resulting spiral arms andthe ring are shown in Fig. 1 in projection to the apparent planeof the galaxy.We note the extremely high pitch angle, which is ≈ http: // galex.stsci.edu / GR6 / ; files AIS_21_sg76-nd-*.fitsArticle number, page 2 of 5usev and Shimanovskaya: Regularity of the young stellar population in the ring Fig. 2.
Photometric profiles in the U (blue), B (green), NUV (black)bands, and H α line (red) along the ring (thin curves) and the spiral arms(thick curves). The ordinate units are magnitudes and logarithm of H α flux in units of erg s − cm − within the aperture. Positions of local max-ima of brightness on the profiles in the U , NUV bands, and H α line areindicated by circles for the ring (large symbols) and for the arms (smallsymbols). Fig. 3.
Separations l between adjacent local maxima of brightness alongthe ring (black circles), Arm E (blue), and Arm W (red). The error barsare shown. Positioning errors, ∆ s , are smaller than the circles size. Seethe text for details. nucleus in a wide range of wavelengths (see Williams et al.2019, and references therein), a high H i central surface density(van Driel & Buta 1991), and it exhibits signs of nuclear activity(Véron-Cetty & Véron 2010).To study variations in brightness along the ring and spiralarms, we obtained photometric profiles along these structures.We used the elliptical aperture (20 × ), with a minoraxis along the spiral arm and ring (see Fig. 1) with a step of 1 ◦ by PA This step corresponds to a linear distance of ≈
67 pc forthe ring and varies from ≈
20 to ≈
80 pc for the spiral arms.The strip of 20 arcsec wide captures all star-forming regions inthe ring (spiral arms), but does not include areas outside them.The exception is the large star complex in the northwestern partof the bar (Fig. 1), however, in the analysis it is superimposedon the young stellar region in the ring having the same PA. Theobtained photometric profiles in NUV, B , V , and H α along thering and arms are presented in Fig. 2. Longitudinal displacement along the ring, denoted as s , is equalto r ring ( θ − θ ), where θ is in radians. For the logarithmic spirals, s = (sin µ ) − r (exp ( k ( θ − θ )) − s =
0, corresponding to PA =
0. It increases from north toward east, the same as the PA. Thedisplacement values along the spiral arms were combined withthe displacements along the ring: at the points of intersectionbetween the spiral arms and the ring at the outer ends of the arms,their displacement was taken equal to the displacement along thering, s arm = s ring .Using mostly profiles in the U band and H α line, and in-volving profiles in the NUV band, we found the local brightnessmaxima on the profiles. We prefer to use the U band image, sinceit has su ffi cient resolution and is sensitive to the presence of ayoung stellar population. In addition, the H α -image was used toidentify H ii regions those are weakly visible in the U passband.The low-resolution NUV image, as well as the image in the B band, were used for control.The local maxima of brightness were determined as thelower extrema of the functions m U ( s ), m NUV ( s ), − log F (H α )( s )for the ring and both spiral arms. To locate them, we looked forpoints with the first derivative of the function, dm / ds = d (log F (H α )) / ds =
0, and the second derivative, d m U / ds > d ( − log F (H α )) / ds = , on the profiles. We selected peakswhose widths exceed three measurement points (correspondingto the angular resolution of the images in optical bands) andwhose amplitude exceeds 0.043 mag for the U , 0.017 dex forH α and 0.24 mag for the NUV (which corresponds to the thresh-old 3 σ above the average background level in the correspondingimages within the used aperture, 20 × ).As a result, we obtained 23 local maxima of brightness in thering, with 6 maxima in the eastern arm (Arm E), and 6 maximain the western arm (Arm W). The positions of these maximaare shown in Fig. 2. Some local maxima of brightness do notappear in U or in H α (see Fig. 2), which is a reflection of thephotometric evolution of young star clusters with an age from ∼ ∼
300 Myr (Whitmore et al. 2011). The typical errorsof maximum positions are ∆ s = ± ±
70 pc for the ring; the maximum ∆ s does not exceed 140 pc (see Fig. 3).In the next step, we measure separations, l , between adjacentlocal maxima of brightness along the ring and arms (Fig. 3). Theseparation between the n th and ( n + st maxima is defined as l n = s n + − s n . For the ring, we looped the profile so that l = s − s .Figure 3 shows a rather regular separation between adjacent localmaxima of brightness along the ring.We built a histogram of the distribution of separations l inthe ring and in the spiral arms (Fig. 4). We note that the armdistributions do not show noticeable regularity. For the ring, thehistogram shows a well-pronounced main peak in the distribu-tion at l ≈
700 pc and several secondary peaks.We estimated the mean and median separations between lo-cal maxima of brightness in the ring of the galaxy for four sepa-rate subsets of objects, numbered in Fig. 4. Our results are pre-sented in Table 2, where the confidence intervals correspondto a 95% confidence probability (0.05 significance level in theframework of Student’s t -test). The errors in the table do not in-clude the intrinsic measurement errors in pairs, that is, equal to ≈ . ffi liation of pairs to aparticular subset. Article number, page 3 of 5 & Aproofs: manuscript no. 38530corr
Fig. 4.
The number distribution histograms of local maxima of bright-ness by separation between adjacent objects along the ring (thick blackline), Arm E (shaded in blue), and Arm W (filled red). Roman numeralsdenote subsets of the objects with close characteristic separations.
Table 2.
Characteristic separations l of local maxima of brightness inthe ring. Subset N Mean (kpc) Median (kpc)I 8 0 . ± .
04 0 . . ± .
07 1 . . ± .
11 1 . . ± .
09 1 . ffi li-ation of pairs of local maxima of brightness to di ff erent subsetsand the accuracy of measurement of their positions and peak am-plitudes.Thus, the distribution of separations can be multimodal.Given the sample size is too small to make a robust multimodal-ity test, we calculated distances D between the peaks relative totheir widths according to the definition by Ashman et al. (1994): D = | µ a − µ b | / q ( σ a + σ b ) /
2, where µ a , µ b are the means ofsubsets a and b (I and II or II and III), σ a , σ b are the standarddeviations. For both pairs of subsets D >
4, as well as for a mix-ture of two Gaussian distributions, D > σ tot (where σ tot comprises both statistical errors and measurement errors for apair of subsets): ∆ µ I − II = .
31 kpc, whereas 3 σ I − II = .
29 kpc;and ∆ µ II − III = .
44 kpc, whereas 3 σ II − III = .
42 kpc.The most important result obtained from the analysis of thedistribution over l is the multiplicity of secondary peaks relativeto the main one. The characteristic (mean) separations in subsetII are about one and a half times greater (1 . ± .
25, consideringthe internal positioning errors of pairs), while in subset III, theyare about two times greater (2 . ± . . ± .
49) than the separationin subset I. We do not discuss subset IV as it contains only twomeasurements, which is insu ffi cient for a statistic analysis.Additionally, to estimate the power spectrum of our data, wecomputed the Lomb-Scargle periodograms for the U -band pro-files along the ring shown in Fig. 2, and for the function p ( s ).The function p ( s ) is a collection of Gaussians, centered at pointsof local maxima of brightness on the profiles, with σ equal tothe peak positioning error. The Gaussian amplitudes are assumed Fig. 5.
Normalized power spectral density of the U profile data fromFig. 2 (thick curves) and the function p ( s ) (thin curves) for the ring.Dashed lines are the FAP levels of 25 and 50%. See the text for moreexplanation. to be equal to the measured peak amplitudes, normalized to themaximum amplitude. If significant peaks were recorded in sev-eral bands ( U , NUV, H α ), the largest normalized amplitude wasadopted. Outside of the Gaussians, the function p ( s ) = U profiles and for functions p ( s ), complementeach other to support the presence of the spatial regularity oflocal maxima of brightness. The former shows characteristiclow frequencies ( l > l < <
25% (FAP; see Horne & Baliunas 1986, for de-tails) of l ≈ .
72 and ≈ .
4. Discussion
Earlier in this paper, we use NGC 628 as an example to dis-cuss the e ff ect of a shock wave on the formation of regular-ity in the distribution of young stellar groupings, along with apossible anti-correlation between shock wave signatures and thepresence of chains of star complexes (Gusev & Efremov 2013).This galaxy has two primary spiral arms, one of them showingan absence of shock wave and hosting the regular chain of starcomplexes, while the other does not. Nevertheless, both spiralarms in NGC 628 have the same regular characteristic separa-tion between fainter star-formation regions that are adjacent toone another. Article number, page 4 of 5usev and Shimanovskaya: Regularity of the young stellar population in the ring
NGC 6217 is an example of a galaxy in which a similar regu-larity in the distribution of young stellar groups occurs in the ringlocated near the corotation radius. No shock waves are expectedto be observed in this area. The detection of regularity here isa rather unusual result. In the corotation region, H i / H clouds,the progenitors of H ii regions and young stellar groups are notexpected to experience any significant mutual influence and theydo not show a tendency towards concentration. A possible rea-son for the structural features observed in the ring of NGC 6217may be that it may be, in fact, a pseudo-ring. A prominent barcan act as the driver of the Jeans instability wave. We note thatthe largest distances between adjacent maxima of brightness areobserved directly in front of the bar (relative to the direction ofrotation of the galaxy; see Fig. 1).We obtained the characteristic distance, Λ ∼
700 pc, whichseems to correspond to a double wavelength of spacing regu-larity. Elmegreen et al. (2018) in M100 and Gusev & Efremov(2013) in M74 (NGC 628) obtained the characteristic spacingsof ∼
400 pc, half as large. Our subset II corresponds to (cid:14) Λ ,which may indicate the existence of a wavelength of (cid:14) Λ . Thepower spectral density of the U profile has the largest peak at l = .
45 kpc (Fig. 5) that is close to (cid:14) Λ . Our linear resolutionis insu ffi cient to look for spatial regularities on the scale of (cid:14) Λ .The characteristic spacing wavelength for hydrogen cloudsis λ = c ( G σ ) − , where c is the sound speed and σ is the masscolumn density of the gas (Elmegreen & Elmegreen 1983). Spa-tial regularity in the ring of NGC 6217 indicates the constancy ofparameters of the gas medium along the ring. The mass columndensity of H i near the ring is σ ≈ M ⊙ pc − (van Driel & Buta1991). The sound speed c is a few km s − for both λ = Λ and λ = (cid:14) Λ , which is a typical speed in a di ff use H i medium.
5. Conclusions
For the first time, a spatial regularity was found in the distribu-tion of the young stellar population in the galaxy ring locatednear the corotation region. This refines our understanding of thephysical processes that contribute to the formation of a regularwave of star formation in rings and arms. The presence or ab-sence of shock waves does not appear to a ff ect the genesis ofregularity in regions of star formation.The obtained characteristic separation, Λ , between adjacentyoung stellar groupings in the ring of NGC 6217 is equal to ≈
670 pc, that is twice as much as the characteristic separationobtained for young stellar groups in the spiral arms of M100and M74. However, we suspect a characteristic separation witha scale of (cid:14) Λ along the ring of NGC 6217, the same as wasfound earlier in spiral arms of M100 and M74. This may indi-cate a unified mechanism for the formation of such regularitiesin spiral arms and in the rings of galaxies. Achieving greaterclarity around the characteristic separations in the rings requiresthe search and analysis of regularities in nearer ringed galaxieswith improved spatial resolution. Acknowledgements.
We are grateful to the referee for their constructive com-ments. We would like to honor Prof. Yu. N. Efremov, who was the initiator of thisresearch program, and who left us in August 2019. The authors acknowledge theuse of the HyperLeda data base (http: // leda.univ-lyon1.fr), the NASA / IPAC Ex-tragalactic Database (http: // ned.ipac.caltech.edu), Barbara A. Miculski Archivefor space telescopes (https: // archive.stsci.edu), and the IDL Astronomy User’sLibrary (https: // idlastro.gsfc.nasa.gov). This study was supported by the RussianFoundation for Basic Research (project no. 20-02-00080). A.G. acknowledgesthe support from the Lomonosov Moscow State University Development Pro-gram (Leading Scientific School ”Physics of stars, relativistic objects and galax-ies”). References