Rotation Curves of Galaxies and Their Dependence on Morphology and Stellar Mass

Yongmin Yoon; Changbom Park; Haeun Chung; Kai Zhang

Yoon, Yongmin;Park, Changbom;Chung, Haeun;Zhang, Kai

2021-10-12 00:00:00

Draft version October 13, 2021 Typeset using L T X twocolumn style in AASTeX631 1 1 2 3 Yongmin Yoon , Changbom Park , Haeun Chung , and Kai Zhang School of Physics, Korea Institute for Advanced Study (KIAS), 85, Hoegiro, Dongdaemun-gu, Seoul, 02455, Republic of Korea University of Arizona, Steward Observatory, 933 N Cherry Ave, Tucson, AZ 85721, USA Lawrence Berkeley National Laboratory, 1 Cyclotron Rd, Berkeley, CA 94720, USA ABSTRACT We study how stellar rotation curves (RCs) of galaxies are correlated on average with morphology and stellar mass (M ) using the nal release of Sloan Digital Sky Survey IV MaNGA data. We use star the visually assigned T -types for the morphology indicator, and adopt a functional form for the RC that can model non- at RCs at large radii. We discover that within the radial coverage of the MaNGA data, the popularly known at rotation curve at large radii applies only to the particular classes of 10:8 galaxies, i.e., massive late types (T -type 1, M & 10 M ) and S0 types (T -type= 1 or 0, star 10:0 M & 10 M ). The RC of late-type galaxies at large radii rises more steeply as M decreases, star star 1 1 9:7 and its slope increases to about +9 km s kpc at M 10 M . By contrast, elliptical galaxies star (T -type 2) have descending RCs at large radii. Their slope becomes more negative as M star 1 1 10:2 decreases, and reaches as negative as 15 km s kpc at M 10 M . We also nd that the star 10:5 inner slope of the RC is highest for elliptical galaxies with M 10 M , and decreases as T -type star 10:5 increases or M changes away from 10 M . The velocity at the turnover radius R is higher for star t higher M , and R is larger for higher M and later T -types. We show that the inner slope of star t star the RC is coupled with the central surface stellar mass density, which implies that the gravitational potential of central regions of galaxies is dominated by baryonic matter. With the aid of simple models for matter distribution, we discuss what determines the shapes of RCs. Keywords: Galaxy rotation curves(619) | Galaxy kinematics(602) | Galaxy properties(615) | Galaxy mass distribution(606) 1. INTRODUCTION plained by visible or detectable matters, were common properties in many galaxies (Rubin et al. 1980; Persic et The rotation curve (RC) of a galaxy is the rotating al. 1996; Catinella et al. 2006). This implies that New- speed of stars or gas in the galaxy as a function of ra- tonian dynamics needs to be modi ed on the galactic dial distance from the galaxy center. The exploration to scale (Milgrom 1983a,b) or invisible matter (dark mat- investigate galaxy RCs began with nearby galaxies such ter) that we have not identi ed yet is prevalent in the as M31. Babcock (1939) studied the RC of M31 and dis- universe (Trimble 1987; de Swart et al. 2017). For now, covered nearly constant rotation velocities at the outer dark matter is an indispensable factor that describes part of the galaxy. Babcock (1939) suggested that the the current standard cosmology. In this way, RCs have mass-to-light ratio should increase at the outer part of been used as a tool for determining mass distributions the galaxy to explain the RC. Oort (1940) also arrived in galaxies. at a similar conclusion based on NGC 3115, suggesting 1 Many previous studies investigated the relation be- that substantial invisible matter should be in the outer tween the shape of RCs and various galaxy properties part of the galaxy. Radio observations for H I gas also in order to understand the correlation between dynam- detected at or nearly at RCs in M31 (van de Hulst et ical mass distributions in galaxies and their properties al. 1957) and M33 (Volders 1959). such as morphology, structure, and luminosity. For ex- With the improvement of observational instruments, ample, the shape of the RC is coupled with the galaxy it was found that at RCs, which cannot be fully ex- luminosity in the sense that rising outer RC pro les are commonly found in low-luminosity galaxies, while high-luminosity galaxies show at or even declining RCs yyoon@kias.re.kr (Casertano & van Gorkom 1991; Persic et al. 1996; So- The invisible matter was presumed to be extremely faint dwarf fue & Rubin 2001; Noordermeer et al. 2007; Swaters et stars or interstellar dust and gas in Oort (1940). arXiv:2110.06033v1 [astro-ph.GA] 12 Oct 2021 2 Yoon et al. (2021) al. 2009; Kalinova et al. 2017). Some studies showed The MaNGA project uses 17 ber-bundled IFUs with 00 00 that the shape of the RC is dependent on galaxy mor- sizes of 12 to 32 depending on the number of bers. phology (Corradi & Capaccioli 1990; Casertano & van These 17 IFUs are assigned within the 3 eld of view. Gorkom 1991; Noordermeer et al. 2007; Swaters et al. The spectrograph of the MaNGA survey is the same 2009; Erroz-Ferrer et al. 2016; Kalinova et al. 2017). as that of the Baryon Oscillation Spectroscopic Sur- Combining and summarizing these results, early-type vey (Smee et al. 2013). It covers a wavelength range galaxies generally show steeply rising inner and declin- of 3,600{10,300A with a mid-range spectral resolution ing/ at outer RC pro les, while late-type galaxies have of R 2000. The target galaxy samples of MaNGA relatively slowly rising inner and at/rising outer RC were selected by using criteria for i-band absolute mag- pro les. Some studies could not nd such a trend for nitude and redshift (and near-ultraviolet { i color for a RCs, depending on morphology (Burstein & Rubin 1985; small portion of targets) that result in 10; 000 galaxies Rubin et al. 1985). evenly distributed in the color-magnitude space and hav- The advent of integral eld unit (IFU) spectroscopy ing an uniform spectroscopic coverage up to 1.5 or 2.5 allows us to obtain spatially resolved spectra in a speci c half-light radius along the major axis (R ). See Wake region of the sky with relatively low observation costs. et al. (2017) for details about selecting target galaxy Thus, observational surveys using IFU spectroscopy are samples. capable of providing a lot of spatially resolved kinematic In this study, we use IFU data cubes from the MaNGA information for numerous galaxies. Many IFU surveys internal data release MPL-11, which is equivalent to the have been conducted in the recent decade, such as the nal release version of the MaNGA data, consisting of SAURON project (Bacon et al. 2001), the MASSIVE all observational data obtained during the survey. There survey (Ma et al. 2014), the DiskMass survey (Bershady are 10,010 galaxies with high-quality observations in this 3D et al. 2010), the ATLAS project (Cappellari et al. nal release version. They are in the redshift range of 2011), the CALIFA survey (S anchez et al. 2012, 2016), z < 0:15 with a median value of z = 0:037 and in the and the SAMI survey (Croom et al. 2021). The numbers M range of M & 10 M with a median value of star star 10:5 of observed galaxies in these surveys range from tens to M = 10 M . Only 1% of MaNGA data were star hundreds, except the SAMI survey that observed 3000 repeated observations for the same galaxy. We used one galaxies. observation among these duplicated ones by choosing The MaNGA (Mapping Nearby Galaxies at Apache the data with the largest IFU size. If the IFU sizes are Point Observatory) IFU survey (Bundy et al. 2015; identical, then we chose the data with the highest blue Drory et al. 2015; Yan et al. 2016; Wake et al. 2017) channel signal-to-noise ratio (S/N). We note that M star is a project of the fourth generation of the Sloan Digital and R of galaxies used here are based on the 2D S ersic Sky Survey (SDSS; Blanton et al. 2017) using the ARC model in the NASA-Sloan Atlas (NSA) catalog (R 2.5 m telescope (Gunn et al. 2006). More than 10; 000 was measured in the r band). galaxies were observed in the MaNGA project. This un- precedented number of observed galaxies makes it pos- 2.2. Deconvolution of IFU Data sible to conduct statistical studies of spatially resolved kinematics for a large amount of diverse galaxies that Data from ground-based telescopes are bound to be were uniformly observed. largely aected by atmospheric seeing as well as aber- Previous studies on the relation between the shape of ration from the instruments. The seeing eect is more RCs and galaxy properties are not based on as large a severe in the data from ber-based IFUs due to large number of uniformly observed galaxies as the MaNGA physical gaps between sampling elements (IFU bers). data have. Now, a comprehensive understanding of Minimizing such an eect allow us to extract more ro- shapes of RCs across various types of galaxies is pos- bust spatially resolved kinematics, particularly in the sible with the aid of the MaNGA data. Here, we in- central parts of galaxies. As in Chung et al. (2020), we vestigate the relation between shapes of stellar RCs for deconvolved all of the IFU data cubes of 10,010 galaxies MaNGA galaxies and galaxy properties such as stellar using the Lucy{Richardson (LR) algorithm (Richard- mass (M ) and morphology, and we discuss how the star son 1972; Lucy 1974) in order to mitigate the seeing dark matter and baryon mass distributions in galaxies eect. The LR deconvolution algorithm is an iterative are coupled with morphology and M of galaxies. star procedure to recover an original image that has been Throughout this study, we use H = 70 km s convolved by a point spread function (PSF). This algo- Mpc , = 0:7, and = 0:3 as cosmological pa- rithm requires only a few parameters to perform, so that rameters. it is suitable for deconvolution of very large data. De- tails about the algorithm and its application to MaNGA 2. SAMPLE AND ANALYSIS This is a base catalog for the selection of target galaxies in the 2.1. SDSS-IV MaNGA MaNGA survey (Wake et al. 2017). Rotation Curves of SDSS-IV MaNGA Galaxies 3 data cubes are in Chung et al. (2020). Therefore, in this less than 15% between galaxies in dierent M bins. star paper, we only brie y describe the implementation of the LR algorithm on the IFU data. 2.3. Extracting Line-of-sight Velocities The algorithm can be described by a simple equation (Shepp & Vardi 1982; Chung et al. 2020), To extract line-of-sight stellar velocities from the IFU data of 10,010 galaxies, we used the code Penalized n+1 n Pixel-Fitting (pPXF; Cappellari & Emsellem 2004; Cap- u = u p ; (1) pellari 2017). This code extracts the stellar kinemat- ics and stellar population through full spectrum t- where u is nth estimation of the maximum likeli- ting on galaxy spectra, using the maximum penalized hood solution and d is an original PSF-convolved im- likelihood formalism. For model templates used in age (hence u = d). Parameter p is a two-dimensional the pPXF, we used MILES single stellar population (2D) PSF and means 2D convolution. This LR algo- models (S anchez-Bl azquez et al. 2006; Vazdekis et al. rithm was applied to the MaNGA cube data in a sense 2010; Falc on-Barroso et al. 2011) with the Padova+00 that the 2D image slice at each wavelength bin was de- isochrone (Girardi et al. 2000) and the initial mass func- convolved separately. MaNGA data cubes contain PSF tion of Chabrier (2003). The model templates have six full-width-half-maximum (FWHM) values at g, r, i, and metallicity values in a range of 1:71 [M=H] 0:22 z bands. We t a linear function to these FWHM values and 10 ages from 0.07 Gyr to 12.59 Gyr, so that 60 model at the wavelengths of the four bands. The FWHM of the templates were used in total. As in Bel ore et al. (2019), PSF at each wavelength bin in the data cubes was de- we also included an eighth-order additive Legendre poly- termined from this linear function, and a 2D Gaussian nomial during the t, in order to improve the quality of function with that FWHM was used in the deconvolu- the derived stellar kinematics (see also Emsellem et al. tion process. The number of iterations (N ) in the iter 2004). LR algorithm was set to 20. This is because N = 20 is iter As in the method of Cappellari (2017), the model the optimum iteration number as tested in Chung et al. templates were convolved with a Gaussian function to (2020): beyond N = 20, the quality of deconvolution iter match the resolution of the MaNGA spectra and the is not signi cantly improved further, while additional spectra were shifted to the rest frame before extract- artifacts can occur in the image with ampli ed noise. ing stellar kinematics. The tting was conducted after Chung et al. (2020) intensively tested the LR decon- masking pixels around known emission lines and bad volution algorithm using MaNGA IFU data and mock pixels (pixels with low coverage depth, dead bers, or IFU data. Speci cally, Chung et al. (2020) generated contamination from foreground stars, etc.) agged in a great deal of mock data of which the properties are the MaNGA data reduction pipeline (Law et al. 2016). identical to MaNGA IFU data in terms of the eld of The tting range was limited in a wavelength range from view, radial coverage, S/N, PSF eects, etc. To mimic 3700 to 7400A in consideration of the wavelength cov- realistic cases, galaxies with various surface brightness erage of the model templates (3540{7410A). The second pro les (S ersic indices) and diverse RC model parame- rows in Figures 1 and 2 show extracted 2D line-of-sight ters (Equation 2) were assigned to the mock IFU data. velocity maps for several examples. Then, they deconvolved the mock IFU data and de- rived RC parameters from the data. By comparing the derived RC parameters and the true values, they 2.4. RC Model Fitting demonstrated that the LR algorithm eectively recovers Here, we describe the method to t 2D RC models to the true stellar kinematics of galaxies. For example, the the line-of-sight velocity maps extracted in Section 2.3. inner slope of the RC (Equation 4), which is expected to Since our aim is to measure quantitative shapes of RCs be largely in uenced by PSF eects, can be recovered for galaxies with diverse properties and to compare them with an underestimation less than 25%. We note with each other, we used an empirical RC model func- that the underestimation of the inner slope of the RC tion that can well quantify shapes of RCs across various due to the limitation of the deconvolution process varies types of galaxies. To quantify shapes of RCs, previous studies used several functional forms such as arctangent, exponential, and hyperbolic tangent (Puech et al. 2008; The dierence between FWHM values of the four bands recorded Feng & Gallo 2011; Andersen & Bershady 2013; Bouch e in MaNGA data cubes and those from the tted linear function is negligible: the average absolute dierence is 0:007 (Chung et al. 2015). RCs based on these models show a rising et al. 2020), since the FWHM value in MaNGA data is only a curve in the inner part and converge to a constant veloc- weak function of wavelength (Law et al. 2016; Chung et al. 2020). ity in the outer part. However, many previous studies Thus, the FWHM value derived from the linear function is a good showed that RCs of galaxies have various slopes in the approximation at each wavelength bin. outer parts depending on galaxy properties (Corradi & We note that PSFs of MaNGA data are well described by a single 2D Gaussian function and the FWHM varies less than 10% across Capaccioli 1990; Casertano & van Gorkom 1991; Per- a given IFU (Law et al. 2015, 2016). sic et al. 1996; Sofue & Rubin 2001; Noordermeer et al. 4 Yoon et al. (2021) 7957-3701 8078-1902 10218-6102 8329-6103 T=-3 T=-3 T=-1 T=1 -100 -50 0 50 100 -60 -40 -20 0 20 40 60 -100 0 100 -150-100 -50 0 50 100 150 -1 -1 -1 -1 V [km s ] V [km s ] V [km s ] V [km s ] LoS LoS LoS LoS 0 0 0 0 -2 -5 -5 -5 -4 -10 -10 -6 5 5 0 0 0 0 -2 -5 -5 -5 -4 -10 -10 -6 -5 0 5 -6 -4 -2 0 2 4 6 -10 -5 0 5 10 -10 -5 0 5 10 Δα [arcsec] Δα [arcsec] Δα [arcsec] Δα [arcsec] 0 0 V = 168 V = 271 V = 238 V = 232 c c c c -100 R = 2.2 R = 1.5 R = 1.2 R = 2.2 t t t t -100 -100 s = 1.9 s = -6.3 -100 s = -3.8 out out out s = -31.0 -200 out i = 59.9 i = 33.9 i = 61.2 i = 26.7 -200 -200 φ = 102 φ = 297 φ = 186 0 0 0 φ = 342 -300 0 -200 -5 0 5 10 -6 -4 -2 0 2 4 6 -10 -5 0 5 10 -10 -5 0 5 10 15 r [arcsec] r [arcsec] r [arcsec] r [arcsec] Figure 1. Examples of galaxies used in this study. The rst row: color images in SDSS. The Plate ID (e.g., 7957) and IFU design ID (e.g., 3701) are on top of the color images. The hexagon represents the MaNGA IFU eld of view. T -type values of the galaxies are in the lower left corner of the images. The second row: extracted 2D line-of-sight velocity maps in which systematic line-of-sight velocities at the kinematic center are set to 0 km s . See the color bars on top of the velocity maps for the color-coded line-of-sight velocity scales. The x and y axes are relative R.A. and decl., respectively. The third row: the best- t 2D models (Equation 3) for the observed velocity maps. The bottom rows: observed 1D velocity pro les. The circles are line-of-sight velocities at r (distance from the kinematic center) after correcting for the geometric term cos( ) and the kinematic inclination sin i (shown here are data points derived from spectra with median S/N 10 and located within 60 from the kinematic major axis). The solid lines are the best- t rotation curve models of Equation 2. In each panel of the bottom 1 1 1 rows, we provide the tted model parameters of Equations 2 and 3 (V in km s , R in arcseconds, s in km s arcsec , c t out and i, in degrees). These examples are galaxies with T -type 1. The corresponding section: Section 2. -1 V [km s ] Δδ [arcsec] Δδ [arcsec] -1 V [km s ] Δδ [arcsec] Δδ [arcsec] -1 V [km s ] Δδ [arcsec] Δδ [arcsec] -1 V [km s ] Δδ [arcsec] Δδ [arcsec] Rotation Curves of SDSS-IV MaNGA Galaxies 5 8723-12703 8723-12705 7495-12704 10220-12705 T=3 T=3 T=4 T=5 -100 0 100 -150 -100 -50 0 50 100 150 -150-100 -50 0 50 100 150 -150-100 -50 0 50 100 150 -1 -1 -1 -1 V [km s ] V [km s ] V [km s ] V [km s ] LoS LoS LoS LoS 15 15 10 10 5 5 0 0 0 0 -5 -5 -5 -10 -10 -10 -10 -15 -15 -15 15 15 10 10 10 5 5 5 0 0 0 0 -5 -5 -5 -10 -10 -10 -10 -15 -15 -15 -15 -10 -5 0 5 10 15 -15 -10 -5 0 5 10 15 -10 0 10 -15 -10 -5 0 5 10 15 Δα [arcsec] Δα [arcsec] Δα [arcsec] Δα [arcsec] 300 300 100 100 0 0 0 V = 173 V = 143 V = 74 V = 151 c c c c R = 2.4 R = 2.8 R = 1.2 R = 2.9 -100 -100 -100 t -100 t t t s = 2.9 s = 5.1 s = 4.3 s = 8.7 out out out out i = 58.2 i = 44.0 i = 58.1 i = 55.6 -200 -200 -200 -200 φ = 274 φ = 18 φ = 83 φ = 305 0 0 0 0 -300 -300 -10 0 10 -20 -10 0 10 -20 -10 0 10 20 -20 -10 0 10 20 r [arcsec] r [arcsec] r [arcsec] r [arcsec] Figure 2. Examples of galaxies used in this study. These examples are galaxies with T -type 3. Other descriptions about the gure are identical to those in Figure 1. The corresponding section: Section 2. -1 V [km s ] Δδ [arcsec] Δδ [arcsec] -1 V [km s ] Δδ [arcsec] Δδ [arcsec] -1 V [km s ] Δδ [arcsec] Δδ [arcsec] -1 V [km s ] Δδ [arcsec] Δδ [arcsec] 6 Yoon et al. (2021) follows median RC pro les for galaxies in diverse M star s =6 and morphology bins (Figure 7). The fact that the same out functional form can be applied to both early- and late- type galaxies is consistent with recent studies of stellar kinematics of galaxies that revealed that a large frac- s =0 out tion of elliptical galaxies, even those with apparently round and nondisky shapes, contain substantial rotat- ing disk components like late-type galaxies (Cappellari s =-6 out 2016; Graham et al. 2018). Furthermore, RCs of simple models consisting of a baryonic disk and a dark mat- ter halo can also be described by this functional form (Figures 12, 13, 14, and 15). The radial coverage of RCs for MaNGA galaxies in the nal sample used here is 5R on average. Thus, the functional form of Equation 2 can be regarded as a good RC model within this range, although it may not be suitable at very large radii (Noordermeer et al. 2007), for which extraction of stellar kinematics using MaNGA data is nearly impossible. 0 2 4 6 8 To t our RC model on the 2D line-of-sight veloc- r [kpc] ity maps extracted in Section 2.3, we used a 2D model (Beckman et al. 2004; Chung et al. 2020) such as Figure 3. Examples of three rotation curves in the form of 0 0 Equation 2. All of the rotation curves have V = 150 km V (r ; ) = V + V (r) sin i cos( ): (3) c obs sys 0 s and R = 2 kpc but have dierent s (6, 0, and 6 km t out 0 0 In this equation, V (r ; ) is the observed 2D line-of- obs 1 1 s kpc , respectively). The vertical dashed line denotes sight velocity map where r is the radial distance from R = 2 kpc. We note that the rotation curve with s = 6 t out the kinematic center of a galaxy to each pixel on the 1 1 1 km s kpc does not reach 150 km s due to the negative plane of the sky, and is the angle of the pixel. Pa- linear term in Equation 2, even though V is 150 km s . rameters r and are the radial coordinate and angle The corresponding section: Section 2.4. in the plane of the galaxy (deprojected plane), respec- tively. V (r) in this equation is the adopted RC function 2007; Kalinova et al. 2017). Re ecting on this fact in of Equation 2. Parameter i is the kinematic inclina- this study and Chung et al. (2020), we devised a new tion angle and is the kinematic position angle of the functional form for RCs of MaNGA galaxies that is a galaxy. V is the systematic line-of-sight velocity at the sys combination of the hyperbolic tangent function and a kinematic center. Including the kinematic center posi- linear term, tion (x and y ) in the plane of the sky, eight center center tting parameters are used in the tting process (x , center V (r) = V tanh(r=R ) + s r; (2) c t out y , V , i, , V , R , and s ). center sys 0 c t out We used the minimum method for the 2D tting, where s is the slope of the RC at large radii r out nding a set of parameters that minimizes between R and R is the turnover radius where the hyperbolic t t the observed line-of-sight map and the 2D model of tangent term begins to be at. The coecient of the Equation 3. As in Section 2.3, pixels contaminated by hyperbolic tangent term V is equal to the maximum foreground stars were masked in the tting. In the t- V (r) when s is 0. Figure 3 shows examples of three out ting, we only used line-of-sight velocities derived from RCs from Equation 2 with dierent s . out spectra with a median S/N larger than or equal to 5. We note that RCs of MaNGA galaxies are well rep- The third rows in Figures 1 and 2 show several ex- resented by the functional form of Equation 2 through- amples of best- t 2D models (Equation 3) for observed out this study. For example, individual RCs for vari- velocity maps that are in the second rows in the gures. ous types of galaxies (both early and late types) prop- The bottom rows in the same gures show observed one- erly t into the functional form, as shown in Figures dimensional velocity pro les with the best- t RC models 1 and 2. The functional form of Equation 2 also well of Equation 2. Through intensive tests with mock IFU data, Chung et al. (2020) found that the maximum radial coverage The concept of the functional form for the RC used in Wiegert & of the 2D velocity maps along the major axis (R ) max English (2014) is similar to ours in the sense that it is a combi- should be at least larger than 2:5R (or conservatively nation of the hyperbolic tangent function and a linear function, t although they used a dierent formula. See Courteau (1997) for 3:0R ) in order to robustly determine the outer slope other functional forms that are more complex. of the RC (s ). For this reason, we only used galaxies out -1 V(r) [km s ] Rotation Curves of SDSS-IV MaNGA Galaxies 7 E S0 Sa Sb Sc Sd Sm angles in this study were not measured erroneously. 400 400 2.5. Morphology Classi cation 300 300 We assigned T -type numbers to galaxies according to their morphological classes. The T -type system is a widely used morphology classi cation scheme (Nair & Abraham 2010; Ann et al. 2015; Dom nguez S anchez et 200 200 al. 2018). We used 14 morphological classes and hence 14 T -type values from 3 to 10 as shown in Table 1. Galaxies with T -type < 0 are early types (elliptical and lenticular galaxies), while those with T -type > 0 corre- 100 100 spond to late types (from spiral to irregular galaxies). Major morphological classes (e.g., E, S0, Sa, Sb, etc.) are odd numbers in our T -type scheme, whereas inter- mediate classes between major ones (e.g., S0/a, Sab, Sbc, etc.) are even numbers. -2 -2 00 22 44 66 88 10 10 We classi ed T -types of galaxies by visually inspecting T-type combined color images of the g, r, and i bands in SDSS (see the rst rows of Figures 1 and 2). A morphologi- Figure 4. T -type distribution for the nal sample. The cal characteristic of elliptical galaxies (E, T -type = 3) corresponding section: Section 2.5. is a smooth, featureless light distribution with an ex- tended stellar halo in the outside of galaxies. Lenticular with R =R 3. To exclude IFU data that cover max t galaxies (S0, T -type = 1) have a large and bright bulge only a small (or central) part of galaxies, we ruled out surrounded by a disk-like structure with no visible spiral galaxies whose R is smaller than R . Applying these max e arms (sometimes they have a ring structure). Sa galax- radial coverage criteria, the number of MaNGA galaxies ies (T -type = 1) show a bright bulge and very tightly is 2829. wound smooth arms. Sb galaxies (T -type = 3) have a RCs for galaxies with low kinematic inclinations (close fainter bulge and less tight spiral arms than Sa. Their to face-on) can be uncertain due to the strong coupling spiral arms are usually blue and prominent. Sc galax- between the sin i and V terms in Equations 2 and 3 ies (T -type = 5) are marked by a small and/or faint (Noordermeer et al. 2007; Chung et al. 2020). A test bulge with loosely wound blue arms. Resolved stellar in Chung et al. (2020) reveals that RCs can be reliably clusters and nebulae can be seen in some Sc galaxies. determined for kinematic inclinations of i & 20 { 25 . Sd galaxies (T -type = 7) have a very faint bulge with a Therefore, we excluded galaxies with i < 20 in this blue disk in which spiral arms are generally fragmented study. The number of MaNGA galaxies after this cut is into stellar clusters and nebulae. Galaxies with irregular 2498. As mentioned in Section 2.5, we excluded edge-on morphologies are classi ed into Sm (T -type = 9) or Im galaxies, so that 99% of galaxies in the nal sample have (T -type = 10). Among them, Sm galaxies show weak i < 75 . spiral structures. For a test, we converted the kinematic inclination In the visual inspection, we found that 515 galaxies angles of galaxies in the nal sample into axis ratios of are too inclined to determine their morphology correctly. rotating disks (q ), assuming that the rotation compo- We also visually inspected results of the RC model t- nents in galaxies are thin disks with intrinsic axis ratios ting (Section 2.4) and found that 43 galaxies have bad of 1, and then compared them with photometric axis tting qualities. We excluded those edge-on (too in- ratios (q ) estimated from a 2D S ersic model in the NSA clined) galaxies and galaxies with bad tting qualities catalog. By doing so, we found that q is very consistent in the nal sample. Thus, the number of galaxies in with q , so that the average dierence between them is our nal sample is 1940. We note that the number of only 0.006 and the standard deviation of the dierence early-type galaxies with T -type < 0 is 830, while that of is 0.15, which implies that the kinematic inclination late-type galaxies with T -type > 0 is 956 (the number of galaxies with T = 0 is 154). Most of galaxies in the nal sample (99:5%) have T -type 6. See Figure 4 for In this study, R was calculated using pixels in the velocity max the T -type distribution for our nal sample. map that are within 5 from the major axis and correspond to spectra with a median S/N 5. In order to check the reliability of our T -type classi- cations, we compared our classi cations with those in two other catalogs. The rst catalog is a value-added Count -2 8 Yoon et al. (2021) Table 1. T -type Classi cation Scheme Class E E/S0 S0 S0/a Sa Sab Sb Sbc Sc Scd Sd Sdm Sm Im T -type -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 E S0 Sa Sb Sc E S0 Sa Sb Sc 11.5 11.5 11.0 11.0 10.5 10.5 10.0 10.0 9.5 9.5 -10 -5 0 5 60 80 100 120 140 160 s s -1 -1 -1 -1 out in s [km s kpc ] s [km s kpc ] out in 9.0 9.0 -2 0 2 4 6 -2 0 2 4 6 T-type T-type Figure 5. Inner and outer slopes of rotation curves (s and s ) in the T -type versus log M plane. The left panel is for in out star s and the right panel is for s . The gray dots are individual galaxies. Slopes s and s are represented by colors (see the out in out in color bars for the color-coded s and s ) and contours (values in the middle of contour lines indicate s or s ). To construct out in out in color maps and contours for s and s , we used a grid in which block sizes in the T -type and log M axes are 1 and 0.14, out in star respectively. At each point in the grid, we calculated the median slopes for galaxies within a rectangular bin whose sizes in the T -type and log M sides are 2 and 0.7, respectively. We only used the 2D bins in which the number of galaxies is larger than star 20. The corresponding section: Section 3. catalog of SDSS containing visually classi ed morphol- mean dierence is reduced to 0:01 0:55 for galaxies ogy information for all MaNGA galaxies in SDSS Data with T -type 3 in our sample . Release (DR) 15 (4614 galaxies). The visual inspec- The second catalog for comparison is a catalog of tion of morphology was conducted based on images of Dom nguez S anchez et al. (2018) that has morphology SDSS and the DESI (Dark Energy Spectroscopic Instru- information for SDSS DR7 galaxies. The classi cations ment) Legacy Imaging Survey (Dey et al. 2019) following of the catalog were obtained with deep learning algo- the methods in Hern andez-Toledo et al. (2010). A total rithms using convolutional neural networks. For the of 846 galaxies in our sample are also in that catalog. training of the T -type classi cation, they used a mor- Since the T -type scheme for early-type galaxies (T -type phology catalog of Nair & Abraham (2010), in which < 0) in the catalog is slightly dierent from ours, we T -type classi cations were determined for SDSS DR4 only conducted a comparison using our late-type galax- galaxies based on visual inspection. The catalog of ies with T -type > 0 and found that the mean dierence Dom nguez S anchez et al. (2018) uses the same T -type of T -type is 0:19 0:71 after excluding 3 outliers scheme as ours, except that the T -type values in their (our T -type values are slightly smaller on average). The catalog are non-integer numbers (e.g., T -type = 2:3) unlike our T -type values that are all integers. A total of 1646 galaxies in our sample are also in this catalog. T -type = 5 for E and T -type = 2 for S0, with essentially no We found that the mean dierence in T -type between intermediate type between them. our study and the catalog of Dom nguez S anchez et al. The error is the standard deviation of the dierence. -4 -6 log[M /M ] star O log[M /M ] star O Rotation Curves of SDSS-IV MaNGA Galaxies 9 E S0 Sa Sb Sc E S0 Sa Sb Sc E S0 Sa Sb Sc E S0 Sa Sb Sc (a) (c) (e) (g) logM≥10.6 logM≥10.6 logM≥10.6 5 160 2.5 logM<10.6 logM<10.6 logM<10.6 180 0 160 2.0 -5 1.5 -10 1.0 logM≥10.6 -15 logM<10.6 60 0.5 -2 0 2 4 6 -2 0 2 4 6 -2 0 2 4 6 -2 0 2 4 6 T-type T-type T-type T-type (b) (d) (f) (h) T-type≥1 T-type≥4 T-type≥-1 T-type≥1 -1≤T-type≤0 0≤T-type≤3 T-type≤-2 2.5 T-type≤0 T-type≤-2 T-type≤-1 2.0 -5 1.5 -10 1.0 -15 60 0.5 9.5 10.0 10.5 11.0 11.5 9.5 10.0 10.5 11.0 11.5 9.5 10.0 10.5 11.0 11.5 9.5 10.0 10.5 11.0 11.5 log[M /M ] log[M /M ] log[M /M ] log[M /M ] star O • star O • star O • star O • Figure 6. Rotation curve parameters as a function of T -type (the upper panels) and log M (the lower panels). Panels (a) star and (b): s ; (c) and (d): s ; (e) and (f ): R ; (g) and (h): V (R ). The square denotes the median value within a bin whose out in t t size is 2 and 0.4 in the upper and lower panels, respectively. The error bar is the standard deviation of the median values from 1000 bootstrap resamplings. Galaxies are divided into two identical mass bins in the upper panels. By contrast, in the lower panels, we divide galaxies into two or three dierent T -type bins, in order to clearly display dierent trends in dierent T -type classes. The corresponding section: Section 3. (2018) is 0:07 1:11 after excluding 3 outliers. For Here, we present how the RC parameters in Equation galaxies with T -type < 0 in our sample, the mean dif- 2 depend on M and T -type. Firstly, we examined the star ference is 0:13 0:91, while it is 0:04 1:15 for galaxies inner and outer slopes of RCs. The outer slope is s out with T -type > 0. Through the comparison of the two in Equation 2. The inner slope is de ned as catalogs, we conclude that our T -type classi cations are V (r) consistent with those in the other two catalogs within s = lim ; (4) in r!0 about 1. r We selected 1940 galaxies for the nal sample with where V (r) is the RC model function of Equation 2. various criteria described above. The selection work Thus, s is a slope of the tangent line at the center of in excludes galaxies evenly in the IFU bundle size and the RC (Lelli et al. 2013; Erroz-Ferrer et al. 2016). (near-ultraviolet { i) color distributions of the whole The left panel of Figure 5 shows s in the T -type out MaNGA galaxy sample without any bias. The selection versus log M plane, and panels (a) and (b) of Fig- star of the nal sample cuts low-mass galaxies a little bit ure 6 display median s as a function of T -type and out more, so that the median M of the nal sample is star 10:7 log M , respectively. For late-type galaxies with T - star 10 M , while that of the whole MaNGA galaxy sam- 10:5 type 1, galaxies with lower M have positively star ple is 10 M . In conclusion, the nal sample used steeper s , so that s changes from 0 to 9 km out out in this study still represents the whole MaNGA galaxy 1 1 1 1 s kpc (s = 9 km s kpc ) when M goes out star sample in terms of (near-ultraviolet { i) color, M , star 11:5 9:5 from 10 to 10 M . By contrast, s only changes out and the IFU bundle size. 1 1 less than 1 km s kpc at a given M in a range of star 1 T -type 6. Early-type galaxies with T -type 0 have smaller s out 3. RESULTS at smaller T -type. This trend is more severe for the lower-mass range of log(M =M ) < 10:7 (s > star out 1 1 17 km s kpc ) than for the higher-mass range of 1 1 log(M =M ) 10:7 (s = 10 km s kpc ). Con- Galaxies with T= 3 and T= 1 in our classi cation have on star out average T= 2:4 and T= 1:5, respectively, in the catalog of sequently, elliptical galaxies with T -type 2 have neg- Dom nguez S anchez et al. (2018). ative s and their s falls down as M decreases out out star -1 -1 -1 -1 s [km s kpc ] s [km s kpc ] out out -1 -1 -1 -1 s [km s kpc ] s [km s kpc ] in in R [kpc] R [kpc] t t -1 -1 V(R ) [km s ] V(R ) [km s ] t t 10 Yoon et al. (2021) T-type ≤-2 -1≤ T-type ≤1 T-type ≥2 2.0 2.0 2.0 log[M /M ] ≥10.8 log[M /M ] ≥10.8 log[M /M ] ≥10.8 • • • star O star O star O 1.5 1.5 1.5 1.0 1.0 1.0 0.5 0.5 0.5 8 15 30 61 122 9 19 37 74 148 16 32 64 128 256 0.0 0.0 0.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 r/R r/R r/R t t t T-type ≤-2 -1≤ T-type ≤1 T-type ≥2 2.0 2.0 2.0 log[M /M ] <10.8 log[M /M ] <10.8 log[M /M ] <10.8 • • • star O star O star O 1.5 1.5 1.5 1.0 1.0 1.0 0.5 0.5 0.5 3 6 13 26 52 12 25 49 99 198 26 51 103 205 411 0.0 0.0 0.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 r/R r/R r/R t t t Figure 7. Stacked one-dimensional velocity pro les for galaxies in several M and T -type bins. In the diagrams, the distances star from the kinematic center r and V (r) are normalized at R and V (R ) of individual galaxies, respectively. In this gure, we t t only used the velocities that were derived from spectra with a median S/N 10 and located within 60 from the kinematic major axes. Each velocity pro le of a galaxy was folded around r = 0 and V (r) = 0 before stacking. The squares are median values at each bin of r=R , while the dotted lines indicate 84th and 16th percentiles of normalized velocities at the bins. The solid lines are the best- t rotation curve models of Equation 2 tted to the median values. The numbers at the bottom of each panel indicate the number of data points in 2D bins of 0.1 (in the x-axis) 0.06 (in the y-axis) used for construction of the ve levels of the contours. The scale of the contour levels in each panel is adjusted according to the total number of data points used in each panel. The corresponding section: Section 3. 1 1 10:0 1 1 10:0 (s = 7 km s kpc in 10 < M =M < with M (from 70 to 90 km s kpc in 10 < out star star 11:6 11:5 10 ), which is the opposite trend to s for T -type M =M < 10 ). By contrast, for T -type 1, the out star 1. maximum s appears at log(M =M ) 10:5 ( 150 in star 1 1 The results for s can also be checked in Figure 7 km s kpc ) and less massive and more massive galax- out showing stacked one-dimensional velocity pro les (nor- ies than log(M =M ) 10:5 have a smaller s . The star in malized at R and V (R )) for galaxies in several M most massive early-type galaxies with T -type 1 and t t star 1 1 and T -type bins. In Figure 8, we also show thumbnail log(M =M ) & 11:3 have s < 100 km s kpc . star in images and one-dimensional velocity pro les for typical Next, we investigated R and the rotation velocity at galaxies that are superposed on the color map of s in R (V (R )), which can be regarded as the typical rota- out t t the T -type versus log M plane. tion velocity of a galaxy. The left panel of Figure 9 shows star The right panel of Figure 5 shows s in the T -type R in the T -type versus log M plane, and panels (e) in t star versus log M plane, and panels (c) and (d) of Fig- and (f ) of Figure 6 display median R as a function of star t ure 6 display median s as a function of T -type and T -type and log M , respectively. R is well correlated in star t 9:5 11:5 log M , respectively. In general, s is a strong func- with M (R 1:7 kpc from 10 to 10 M ) in star in star t tion of T -type in such a way that galaxies with a smaller the sense that galaxies with a higher M have a larger star 1 1 T -type have a steeper s ( 130 km s kpc at T -type R . The linear Pearson correlation coecient (C ) of R in t P t 2) than those with a larger T -type (s 70 km and log M is 0:53. R also depends on T -type in the in star t 1 1 s kpc at T -type 5). The M dependence is also sense that galaxies with a smaller T -type have a smaller star seen in s . For T -type 3, s tends to mildly increase R at a given M (R 0:8 kpc from T -type = 3 in in t star t V(r)/V(R ) V(r)/V(R ) t t V(r)/V(R ) V(r)/V(R ) t t V(r)/V(R ) V(r)/V(R ) t t Rotation Curves of SDSS-IV MaNGA Galaxies 11 E S0 Sa Sb Sc constructed by taking median values of RC parameters 11.5 in each bin. T=-3 logM =11.0 T=3 logM =11.0 * 0 * 0 11.0 4. DISCUSSION: WHAT DETERMINES THE 0 2 4 6 8 10 0 5 10 15 SHAPES OF RCs? Previous studies suggested that baryon mass (or lumi- 10.5 50 nous matter) dominates the gravitational potentials of T=-1 logM =10.7 0 2 4 6 8 10 the inner regions of galaxies so that inner slopes of RCs are largely coupled with the central surface brightness 100 or central surface stellar mass density ( ) of galaxies 10.0 50 50 (Corradi & Capaccioli 1990; Noordermeer et al. 2007; T=-3 logM =10.4 T=5 logM =10.3 * * 0 0 0 2 4 6 0 5 10 15 20 Swaters et al. 2009; Lelli et al. 2013; Erroz-Ferrer et al. 2016). Thus, we calculated for our galaxies to ex- 9.5 amine whether can explain the trends of s in the in T -type versus log M plane. Here, is half the stellar star -10 -5 0 5 s mass divided by the surface area of the half-light ellipse -1 -1 out s [km s kpc ] out 9.0 0:5M star -2 0 2 4 6 = ; (5) (R q ) e p T-type Figure 8. A color map showing s in the T -type versus where q is the axis ratio of each galaxy estimated from out p the 2D S ersic model in the NSA catalog. log M plane, which is a smoothed version of that in the left star Figure 11 shows log in the T -type versus log M panel of Figure 5. Thumbnail images and one-dimensional star plane. It is obvious that the variation of log in the velocity pro les for ve typical examples are superposed. gure is very similar to that of s (the right panel of in The description for the one-dimensional velocity pro les and Figure 5) in the sense that (1) log is mainly depen- the best- t rotation curve models (the white solid lines) are dent on T -type (earlier-type galaxies have a larger ); the same as that in Figure 1, except that they are folded (2) log tends to slightly increase with M for T - star around r = 0 and V (r) = 0 here. The superposed gures type 3; (3) For T -type . 0, low-mass and the most are placed in the map according to the T -type and M star massive galaxies have smaller log than galaxies with values of the galaxies, which are marked in the bottoms of log(M =M ) 10:5. This indicates that s is cou- star in the thumbnail images. The corresponding section: Section pled with in our galaxies in the sense that galax- ies with a high have steep inner slopes in their RCs (C = 0:47 for log and s ), which implies that the P in to 5). The reason for the T -type dependence is that R gravitational potentials of the inner regions of galaxies also depends on R (see Section 4) in such a way that are dominated by luminous matter in the whole range galaxies with a smaller R have a smaller R (C = 0:64 e t P of M and T -type we probed. star for R and R ) and earlier-type galaxies (smaller T -type) t e In order to understand the dependence of s (and in have a smaller R at a given M (Shen et al. 2003). e star also other parameters: s , R , and V (R )) on more out t t The right panel of Figure 9 shows V (R ) in the T - elaborately, we used simple models of a rotating bary- type versus log M plane, and panels (g) and (h) of star onic disk within a dark matter halo and examined how Figure 6 display median V (R ) as a function of T -type their RCs depend on the size and mass of the baryonic and log M , respectively. V (R ) mainly correlates star t disk. For the rotating disk model, we adopted an ex- with M so that galaxies with a higher M have a star star ponential disk model with a thickness of 0.1. The RC larger V (R ) (C = 0:59 for log M and V (R )). We t P star t of the baryonic disk model (V ) was derived by a baryon note that massive elliptical galaxies with T -type 2 method in Noordermeer (2008). and log(M =M ) 11:0 have a noticeably smaller star V (R ) ( 140 km s ) than galaxies with T -type > 1 with similar M (V (R ) 190 km s ). This is be- star t In the standard dark matter cosmological models, massive early-type galaxies with log(M =M ) & 11:2 have experienced cause such massive elliptical galaxies are generally not star multiple (dry) mergers in their formation histories, which can rotation-supported systems, in contrast to disk galaxies pu up their sizes (Oogi & Habe 2013; Yoon et al. 2017) and (Emsellem et al. 2011; Penoyre et al. 2017; Veale et al. make their relatively small compared to that of lower-mass 2017; Graham et al. 2018; Lagos et al. 2018). galaxies with log(M =M ) 10:5. star The results for s , s , R , and V (R ) can also be out in t t Here, V can be assumed to be the RC of the stellar com- baryon ponent in the disk. We did not consider the gas component sep- identi ed in Figure 10. Figure 10 shows typical RCs of arately. galaxies in various T -type and M bins, which were star log[M /M ] star O 1.1 1.5 1.1 1.5 1.9 2.3 12 Yoon et al. (2021) E S0 Sa Sb Sc E S0 Sa Sb Sc 11.5 11.5 11.0 11.0 10.5 10.5 10.0 10.0 9.5 9.5 0.5 1.0 1.5 2.0 2.5 50 100 150 200 R V(R ) -1 t t R [kpc] V(R ) [km s ] t t 9.0 9.0 -2 0 2 4 6 -2 0 2 4 6 T-type T-type Figure 9. R and V (R ) in the T -type versus log M plane. The left panel is for R and the right panel is for V (R ). R t t star t t t and V (R ) are represented by colors (see the color bars for the color-coded R and V (R )) and contours (values in the middle t t t of contour lines indicate R and V (R )). Other descriptions about the construction of the color maps and contours are identical t t to those in Figure 5. The corresponding section: Section 3. For the density pro le of the dark matter halo model, we adopted the Navarro{Frenk{White (NFW) pro le (Navarro et al. 1997): (r) = ; (6) (r=r )(1 + r=r ) s s 150 where r is the scale radius and is the characteristic s s density related to the critical density of the universe. Then, the RC of the dark matter halo model (Navarro et al. 1997; Lang et al. 2017) is V ln(1 + cx) (cx)=(1 + cx) 2 200 T-type ≥ 2 V (r) = ; (7) DM x ln(1 + c) c=(1 + c) -1 ≤ T-type ≤ 1 T-type ≤ -2 where x = r=r and Parameter c is the concentra- tion parameter of the NFW pro le. c is related to r log[M /M ] ≥ 10.8 s star O log[M /M ] < 10.8 through r = r =c. Here, r is the radius within • s 200 200 star O which the mean density is 200 times the critical density of the universe and V is the circular speed at r , so 200 200 0 2 4 6 8 that r = V =(10H (z)). V is connected with M 200 200 200 200 r [kpc] through V = 10GH (z)M , where G is the gravita- tional constant and M is the halo mass within r . 200 200 Figure 10. Typical rotation curves of galaxies in various T - The Hubble parameter H (z) (Lang et al. 2017) is type and M bins. These rotation curves were constructed star by taking median values of rotation curve parameters (Equa- H (z) = H (1 + z) + ; (8) 0 m tion 2) in each bin. The rotation curve pro les are displayed up to 5R . The corresponding section: Section 3. in which we use the typical redshift of our sample (z = 0:04). Since numerical simulations demonstrate that c is correlated with halo mass in such a way that low- mass halos are more concentrated (Bullock et al. 2001; 0.7 -1 V(r) [km s ] log[M /M ] star O log[M /M ] star O 8.5 8.8 Rotation Curves of SDSS-IV MaNGA Galaxies 13 E S0 Sa Sb Sc Model fit 11.5 Total Disk DM R =1.50kpc 0 e -5 0 2 4 6 8 11.0 r [kpc] 120 1.4 1.6 1.8 2.0 2.2 2.4 100 R [kpc] 10.5 R =1.75kpc 0 2 4 6 8 r [kpc] 10.0 100 60 1.4 1.6 1.8 2.0 2.2 2.4 60 R [kpc] 1.3 9.5 20 R =2.00kpc 0 e 1.2 0 2 4 6 8 1.1 r [kpc] 8.0 8.2 8.4 8.6 8.8 9.0 120 1.0 -2 log[μ /(M kpc )] * O • 100 9.0 0.9 0.8 -2 0 2 4 6 1.4 1.6 1.8 2.0 2.2 2.4 T-type R [kpc] R =2.25kpc e 0 e 0 2 4 6 8 r [kpc] Figure 11. Central surface stellar mass density log in the T -type versus log M plane. Here, log is represented star 100 by colors (see the color bars for the color-coded log ) and contours (values in the middle of the contour lines indicate 40 R =2.50kpc log ). Other descriptions about the construction of the e 1.4 1.6 1.8 2.0 2.2 2.4 0 2 4 6 8 color map and contours are identical to those in Figure 5. R [kpc] r [kpc] The variation of log is very similar to that of s shown in Figure 12. Dependence of the rotation curve shape and in the right panel of Figure 5. The corresponding section: parameters (s , s , R , and V (R )) on R of the bary- out in t t e Section 4. onic rotating disk. Five values of R (1.50, 1.75, 2.00, 2.25, and 2.50 kpc) are used here, while M , , and M baryon 0 200 Wechsler et al. 2002), we used a relation between c and 9 1 11:5 are xed to 7:5 10 M , 20 km s , and 10 M , respec- M in Duy et al. (2008), which is tively. The black solid line means a total observed rotation 0:097 curve of the model (Equation 11). The contributions to the c = 5:74 : (9) total rotation curve from the dark matter halo (Equation 7) 12 1 2 10 h M and the baryonic disk (Noordermeer 2008) are indicated by the dashed line and the blue solid line, respectively. The Therefore, V (r) is determined by a single parameter DM thick gray line is the tted functional form of Equation 2. M in this simple dark matter halo model. The total The rotation curve pro les are displayed up to 3R . The circular velocity of the baryonic disk and the dark mat- corresponding section: Section 4. ter halo model (Burkert et al. 2010, 2016; Lang et al. 2017; Genzel et al. 2020) is calculated by in which R is the exponential scale length, which is 2 2 2 related to R by R = 1:68R . Since V (r) is deter- V (r) = V (r) + V (r): (10) e e d baryon circ baryon DM mined by R of the exponential pro le and the total mass of the baryonic disk model (M ), V (r) of The velocity dispersion in the stellar components of baryon obs the disk can generate an outward pressure gradient that the simple model is a function of the four parameters (R , M , , and M ). reduces the centripetal force. Thus, we should take into e baryon 0 200 account the eect of the pressure support in RCs. As- Figures 12 and 13 show how the observed RC shape is dependent on R (Figure 12) or M (Figure 13) suming the velocity dispersion is constant in the ex- 0 e baryon ponential disk model in hydrostatic equilibrium, the ob- of the baryonic disk when the other three parameters served RC (Burkert et al. 2010, 2016; Lang et al. 2017; Genzel et al. 2020; Kretschmer et al. 2021) is described In this model, R is equivalent to the half-mass radius. by 2 2 2 V (r) = V (r) 2 ; (11) obs circ 0 8.85 8.95 8.2 8.9 8.85 8.1 8.4 8.3 8.7 8.6 log[M /M ] star O -1 -1 -1 -1 -1 V(r) [km s ] V(r) [km s ] V(r) [km s ] V(r) [km s ] V(r) [km s ] -1 -1 -1 V(R ) [km s ] s [km s kpc ] t in R [kpc] -1 -1 s [km s kpc ] out 14 Yoon et al. (2021) Model fit Total 5 decreases and/or M increases), though it is natu- Disk baryon 150 DM rally expected by the Newton's law. Moreover, the ef- fects of R and M on R and V (R ) in the models e baryon t t are consistent with Figure 9 and the explanations about log[M /M ]= 9.7 baryon O • 0 -5 0 2 4 6 8 it. r [kpc] In Figures 12 and 13, the simple RC models also show 9.8 10.0 10.2 10.4 log[M /M ] baryon O that galaxies with a larger (smaller R and/or larger 150 e M ) have smaller s . This is due to the fact that baryon out concentrated matter in the central parts causes a more log[M /M ]= 9.9 15 baryon O 0 rapid decline of circular speed in the outer parts. Fur- 0 2 4 6 8 r [kpc] thermore, this is also attributable to the fact that galax- ies with a larger are less aected by dark matter halos that make RCs rise in the outer regions unlike baryonic 9.8 10.0 10.2 10.4 log[M /M ] baryon O disks. We note that this is consistent with previous stud- 1.4 ies that suggested that outer RC pro les are related to 1.3 log[M /M ]=10.1 baryon O • light distributions or mass densities of galaxies (Caser- 0 2 4 6 8 1.2 r [kpc] tano & van Gorkom 1991; Kalinova et al. 2017; Tiley 1.1 1.0 et al. 2019). Our observational data also support the 0.9 correlation between and s (C = 0:37 for log out P 0.8 and s ). out 9.8 10.0 10.2 10.4 log[M /M ] log[M /M ]=10.3 • We also conducted a test to examine how the observed baryon O • baryon O 0 2 4 6 8 RC pro les depend on M when the other three pa- r [kpc] 200 rameters are xed, which is shown in Figure 14. The test in Figure 14 shows that the change of M mainly 100 has an eect on s in the sense that a higher fraction of out 50 dark matter makes a positively steeper slope in the outer log[M /M •]=10.5 baryon O RC pro le. Therefore, the strong M dependence of star 9.8 10.0 10.2 10.4 0 2 4 6 8 log[M /M ] r [kpc] • s in late-type galaxies can be explained if lower-mass baryon O out late-type galaxies are more dominated by dark matter Figure 13. Dependence of the rotation curve shape and pa- than are higher-mass ones, which is supported by obser- rameters (s , s , R , and V (R )) on M of the bary- out in t t baryon vational evidence from previous studies (Casertano & 9:7 9:9 10:1 onic rotating disk. Five values of M (10 , 10 , 10 , baryon van Gorkom 1991; Persic et al. 1996; Lelli et al. 2013; 10:3 10:5 10 , and 10 M ) are used here, while R , , and M e 0 200 Martinsson et al. 2013; Courteau & Dutton 2015). 1 12 are xed to 2:5 kpc, 20 km s , and 10 M , respectively. For a further test, we investigated how the observed Other descriptions about the gure are identical to those in RC pro les depend on when the other three param- Figure 12. The corresponding section: Section 4. eters of the model are xed. This test, the result of which is shown in Figure 15, demonstrates that the change in largely aects s so that a model galaxy are xed in each case. So, both of the gures also 0 out with a large has a negatively steep slope in the outer show a dependence of the RC shape on if we assume 0 RC pro le. This can be the reason for our result that that M is similar to M . The RC shape param- baryon star elliptical galaxies for which stellar velocity dispersions eters such as s , s , R , and V (R ) were derived by out in t t are high have descending RCs at large radii. tting the functional form (Equation 2) to the RCs of It is known that cluster environments can strip dark the simple models. The change of R has a large eect matter halos of satellite galaxies up to 80% (Whitmore on R and s in such a way that R increases and s t in t in et al. 1988; Whitmore & Forbes 1989; Smith et al. 2016; decreases when R goes up. The variation of M e baryon Rhee et al. 2017; Niemiec et al. 2019). Thus, this en- largely aects V (R ) and s so that both V (R ) and t in t vironmental eect can contribute in part to the small s increase when M grows (see Bekerait e et al. in baryon values of s in elliptical galaxies of T -type 2, 2016 for the absolute magnitude{rotation velocity re- out since they usually reside in dense environments accord- lation for rotating galaxies). As a result, our simple model veri es that s increases when increases (R in e In the extreme case where most of the matter is concentrated in a very tiny region, the RC declines with distance just as the See the captions of Figures 12 and 13 for parameters used in the velocity of a planetary system does following the Kepler's laws models. We selected these parameter values to make model RCs 0:5 (V (r) / r ). broadly consistent with observational ones. 14 See the caption of Figure 14 for parameters used in the test. We note that the eect of the decrease in R in the model is See the caption of Figure 15 for parameters used in the test. similar to that of the increase in the S ersic index. -1 -1 -1 -1 -1 V(r) [km s ] V(r) [km s ] V(r) [km s ] V(r) [km s ] V(r) [km s ] -1 -1 -1 V(R ) [km s ] s [km s kpc ] t in R [kpc] -1 -1 s [km s kpc ] out Rotation Curves of SDSS-IV MaNGA Galaxies 15 Model fit Total Model fit Total 0 150 Disk 5 Disk DM 300 DM -5 -10 -1 log[M /M •]=11.0 σ = 0 km s 200 O 0 0 -5 0 0 2 4 6 8 0 2 4 6 8 -15 r [kpc] r [kpc] 11.0 11.5 12.0 12.5 13.0 0 20 40 60 150 -1 log[M /M ] σ [km s ] 200 O 0 -1 log[M /M ]=11.5 σ = 15 km s 200 O • 0 160 0 0 0 2 4 6 8 0 2 4 6 8 r [kpc] r [kpc] 80 400 11.0 11.5 12.0 12.5 13.0 0 20 40 60 -1 log[M /M ] σ [km s ] 200 O 0 50 1.4 1.5 -1 1.3 1.4 log[M /M ]=12.0 • σ = 30 km s 200 O 0 0 0 0 2 4 6 8 1.2 0 2 4 6 8 1.3 r [kpc] r [kpc] 1.1 1.2 1.0 1.1 0.9 1.0 0.8 0.9 11.0 11.5 12.0 12.5 13.0 0 20 40 60 -1 -1 log[M /M ] σ [km s ] log[M /M ]=12.5 • σ = 45 km s 200 O • 200 O 0 0 0 0 160 200 0 2 4 6 8 0 2 4 6 8 r [kpc] r [kpc] 140 180 150 120 160 100 140 80 120 -1 60 100 log[M /M ]=13.0 200 O • σ = 60 km s 0 0 11.0 11.5 12.0 12.5 13.0 0 20 40 60 0 2 4 6 8 0 2 4 6 8 -1 log[M /M ] σ [km s ] r [kpc] • r [kpc] 200 O 0 Figure 14. Dependence of the rotation curve shape and pa- Figure 15. Dependence of the rotation curve shape and rameters (s , s , R , and V (R )) on M of the dark mat- parameters (s , s , R , and V (R )) on of the baryonic out in t t 200 out in t t 0 11:0 11:5 12:0 12:5 1 ter halo. Five values of M (10 , 10 , 10 , 10 , disk. Five values of (0, 15, 30, 45, and 60 km s ) are 200 0 13:0 and 10 M ) are used here, while R , M , and are used here, while R , M , and M are xed to 2:5 kpc, e baryon 0 e baryon 200 10 1 10 13:5 xed to 2:5 kpc, 10 M , and 20 km s , respectively. Other 5 10 M and 10 M , respectively. Other descriptions descriptions about the gure are identical to those in Figure about the gure are identical to those in Figure 12. The 12. The corresponding section: Section 4. corresponding section: Section 4. ing to the morphology{density relation (Dressler 1980; trum tting on galaxy spectra. To quantify shapes of Springel et al. 2001; Thomas & Katgert 2006). RCs, we used a functional form of the RC (Equation 2) Previous studies showed that some elliptical galaxies that can model non- at RCs at large radii and well rep- have irregular rotating features such as kinematically resents RCs of MaNGA galaxies. Using this functional distinct cores and counter-rotating disks (Cappellari form, we t 2D RC models (Equation 3) to the line-of- 2016; Graham et al. 2018). Since their s is small sight velocity maps extracted using pPXF. We classi ed out compared with that of regularly rotating galaxies, the the morphologies of our galaxies and assigned T -type statistically small s in galaxies with T -type 2 numbers. The number of galaxies in the nal sample out can be partly due to the existence of such galaxies with is 1940. The main conclusions about RCs of MaNGA irregular rotations. galaxies are as follows. 1. Within the radial coverage of the MaNGA data, the at RC at large radii applies only 5. SUMMARY to massive late-type galaxies (T -type 1 and We investigated stellar RCs of SDSS-IV MaNGA log(M =M ) & 10:8) and S0 galaxies (T -type= star galaxies and their dependence on stellar mass and mor- 1 or 0, and log(M =M ) & 10:0). star phology (T -type). Deconvolution was conducted on the MaNGA IFU data using the LR algorithm in order to 2. For late-type galaxies with T -type 1, galaxies mitigate the seeing eect as in Chung et al. (2020). Line- with a lower M have a positively steeper outer star of-sight stellar velocities were extracted from the IFU slope (s ) so that s increases to about +9 km out out 1 1 data using the pPXF code that implements full spec- s kpc at log(M =M ) 9:7. star -1 -1 -1 -1 -1 V(r) [km s ] V(r) [km s ] V(r) [km s ] V(r) [km s ] V(r) [km s ] -1 -1 -1 V(R ) [km s ] s [km s kpc ] t in R [kpc] -1 -1 s [km s kpc ] out -1 -1 -1 -1 -1 V(r) [km s ] V(r) [km s ] V(r) [km s ] V(r) [km s ] V(r) [km s ] -1 -1 -1 -1 -1 V(R ) [km s ] s [km s kpc ] t in s [km s kpc ] R [kpc] t out 16 Yoon et al. (2021) 3. By contrast, elliptical galaxies (T -type 2) have descending RCs at large radii. The s of ellipti- This work was supported by a KIAS Individual Grant out cal galaxies falls down as M decreases so that PG076301 at the Korea Institute for Advanced Study. star 1 1 s becomes as negative as 15 km s kpc at Funding for the Sloan Digital Sky Survey IV has been out log(M =M ) 10:2. provided by the Alfred P. Sloan Foundation, the U.S. star Department of Energy Oce of Science, and the Partic- ipating Institutions. SDSS-IV acknowledges support and resources from 4. The inner slope of the RC (s ) is highest in the Center for High Performance Computing at the Uni- for elliptical galaxies with T -type . 2 and versity of Utah. The SDSS website is www.sdss.org. log(M =M ) 10:5, and decreases as T - star SDSS-IV is managed by the Astrophysical Research type increases or M changes away from star Consortium for the Participating Institutions of the log(M =M ) 10:5. star SDSS Collaboration including the Brazilian Partici- pation Group, the Carnegie Institution for Science, Carnegie Mellon University, Center for Astrophysics 5. The velocity at the turnover radius (R ) is higher | Harvard & Smithsonian, the Chilean Participation for a higher M , and R is larger for a higher Group, the French Participation Group, Instituto de star t M and later T -types. Astrof sica de Canarias, The Johns Hopkins Univer- star sity, Kavli Institute for the Physics and Mathematics of the Universe (IPMU) / University of Tokyo, the Ko- rean Participation Group, Lawrence Berkeley National 6. Our observational data and the simple model of a Laboratory, Leibniz Institut fur Astrophysik Potsdam rotating baryonic disk within a dark matter halo (AIP), Max-Planck-Institut fur Astronomie (MPIA Hei- suggest that s is coupled with the central surface in delberg), Max-Planck-Institut fur Astrophysik (MPA stellar mass density : galaxies with a higher Garching), Max-Planck-Institut fur Extraterrestrische have a steeper s , which implies that the gravita- in Physik (MPE), National Astronomical Observatories of tional potentials of the inner regions of galaxies are China, New Mexico State University, New York Uni- dominated by luminous matter (baryonic matter). versity, University of Notre Dame, Observat ario Na- cional / MCTI, The Ohio State University, Pennsylva- nia State University, Shanghai Astronomical Observa- We plan to conduct a further study using MaNGA tory, United Kingdom Participation Group, Universidad data to investigate the eects of environments on RCs, Nacional Aut onoma de M exico, University of Arizona, since the environmental eect can also be an important University of Colorado Boulder, University of Oxford, factor for the shapes of RCs (Whitmore et al. 1988; University of Portsmouth, University of Utah, Univer- Whitmore & Forbes 1989; Smith et al. 2016; Rhee et al. sity of Virginia, University of Washington, University of 2017; Niemiec et al. 2019). Wisconsin, Vanderbilt University, and Yale University. REFERENCES Andersen, D. R. & Bershady, M. A. 2013, ApJ, 768, 41. Bershady, M. A., Verheijen, M. A. W., Swaters, R. A., et al. doi:10.1088/0004-637X/768/1/41 2010, ApJ, 716, 198. doi:10.1088/0004-637X/716/1/198 Ann, H. B., Seo, M., & Ha, D. K. 2015, ApJS, 217, 27. Blanton, M. R., Bershady, M. A., Abolfathi, B., et al. 2017, doi:10.1088/0067-0049/217/2/27 AJ, 154, 28. doi:10.3847/1538-3881/aa7567 Babcock, H. W. 1939, Lick Observatory Bulletin, 498, 41. Bouch e, N., Carfantan, H., Schroetter, I., et al. 2015, AJ, doi:10.5479/ADS/bib/1939LicOB.19.41B 150, 92. doi:10.1088/0004-6256/150/3/92 Bacon, R., Copin, Y., Monnet, G., et al. 2001, MNRAS, Bullock, J. S., Kolatt, T. S., Sigad, Y., et al. 2001, 326, 23. doi:10.1046/j.1365-8711.2001.04612.x MNRAS, 321, 559. doi:10.1046/j.1365-8711.2001.04068.x Beckman, J. E., Zurita, A., & Vega Beltr an, J. C. 2004, Bundy, K., Bershady, M. A., Law, D. R., et al. 2015, ApJ, Lecture Notes and Essays in Astrophysics, 43 798, 7. doi:10.1088/0004-637X/798/1/7 Bekerait e, S., Walcher, C. J., Falc on-Barroso, J., et al. Burkert, A., F orster Schreiber, N. M., Genzel, R., et al. 2016, A&A, 593, A114. doi:10.1051/0004-6361/201527405 2016, ApJ, 826, 214. doi:10.3847/0004-637X/826/2/214 Bel ore, F., Westfall, K. B., Schaefer, A., et al. 2019, AJ, Burkert, A., Genzel, R., Bouch e, N., et al. 2010, ApJ, 725, 158, 160. doi:10.3847/1538-3881/ab3e4e 2324. doi:10.1088/0004-637X/725/2/2324 Rotation Curves of SDSS-IV MaNGA Galaxies 17 Burstein, D. & Rubin, V. C. 1985, ApJ, 297, 423. Graham, M. T., Cappellari, M., Li, H., et al. 2018, doi:10.1086/163541 MNRAS, 477, 4711. doi:10.1093/mnras/sty504 Cappellari, M. 2016, ARA&A, 54, 597. Gunn, J. E., Siegmund, W. A., Mannery, E. J., et al. 2006, doi:10.1146/annurev-astro-082214-122432 AJ, 131, 2332. doi:10.1086/500975 Cappellari, M. 2017, MNRAS, 466, 798. Hern andez-Toledo, H. M., V azquez-Mata, J. A., doi:10.1093/mnras/stw3020 Mart nez-V azquez, L. A., et al. 2010, AJ, 139, 2525. Cappellari, M. & Emsellem, E. 2004, PASP, 116, 138. doi:10.1088/0004-6256/139/6/2525 doi:10.1086/381875 Kalinova, V., Colombo, D., Rosolowsky, E., et al. 2017, Cappellari, M., Emsellem, E., Krajnovi c, D., et al. 2011, MNRAS, 469, 2539. doi:10.1093/mnras/stx901 MNRAS, 416, 1680. doi:10.1111/j.1365-2966.2011.18600.x Kretschmer, M., Dekel, A., Freundlich, J., et al. 2021, MNRAS, 503, 5238. doi:10.1093/mnras/stab833 Casertano, S. & van Gorkom, J. H. 1991, AJ, 101, 1231. doi:10.1086/115759 Lagos, C. del P., Schaye, J., Bah e, Y., et al. 2018, MNRAS, Catinella, B., Giovanelli, R., & Haynes, M. P. 2006, ApJ, 476, 4327. doi:10.1093/mnras/sty489 640, 751. doi:10.1086/500171 Lang, P., F orster Schreiber, N. M., Genzel, R., et al. 2017, Chabrier, G. 2003, PASP, 115, 763. doi:10.1086/376392 ApJ, 840, 92. doi:10.3847/1538-4357/aa6d82 Chung, H., Park, C., & Park, Y.-S. 2020, arXiv:2008.04313 Law, D. R., Cherinka, B., Yan, R., et al. 2016, AJ, 152, 83. doi:10.3847/0004-6256/152/4/83 Corradi, R. L. M. & Capaccioli, M. 1990, A&A, 237, 36 Courteau, S. 1997, AJ, 114, 2402. doi:10.1086/118656 Law, D. R., Yan, R., Bershady, M. A., et al. 2015, AJ, 150, 19. doi:10.1088/0004-6256/150/1/19 Courteau, S. & Dutton, A. A. 2015, ApJL, 801, L20. doi:10.1088/2041-8205/801/2/L20 Lelli, F., Fraternali, F., & Verheijen, M. 2013, MNRAS, 433, L30. doi:10.1093/mnrasl/slt053 Croom, S. M., Owers, M. S., Scott, N., et al. 2021, Lucy, L. B. 1974, AJ, 79, 745. doi:10.1086/111605 MNRAS, 505, 991. doi:10.1093/mnras/stab229 de Swart, J. G., Bertone, G., & van Dongen, J. 2017, Ma, C.-P., Greene, J. E., McConnell, N., et al. 2014, ApJ, Nature Astronomy, 1, 0059. doi:10.1038/s41550-017-0059 795, 158. doi:10.1088/0004-637X/795/2/158 Dey, A., Schlegel, D. J., Lang, D., et al. 2019, AJ, 157, 168. Martinsson, T. P. K., Verheijen, M. A. W., Westfall, K. B., doi:10.3847/1538-3881/ab089d et al. 2013, A&A, 557, A131. doi:10.1051/0004-6361/201321390 Dom nguez S anchez, H., Huertas-Company, M., Bernardi, Milgrom, M. 1983a, ApJ, 270, 365. doi:10.1086/161130 M., et al. 2018, MNRAS, 476, 3661. doi:10.1093/mnras/sty338 Milgrom, M. 1983b, ApJ, 270, 371. doi:10.1086/161131 Nair, P. B. & Abraham, R. G. 2010, ApJS, 186, 427. Dressler, A. 1980, ApJ, 236, 351. doi:10.1086/157753 Drory, N., MacDonald, N., Bershady, M. A., et al. 2015, doi:10.1088/0067-0049/186/2/427 AJ, 149, 77. doi:10.1088/0004-6256/149/2/77 Navarro, J. F., Frenk, C. S., & White, S. D. M. 1997, ApJ, 490, 493. doi:10.1086/304888 Duy, A. R., Schaye, J., Kay, S. T., et al. 2008, MNRAS, 390, L64. doi:10.1111/j.1745-3933.2008.00537.x Niemiec, A., Jullo, E., Giocoli, C., et al. 2019, MNRAS, Emsellem, E., Cappellari, M., Krajnovi c, D., et al. 2011, 487, 653. doi:10.1093/mnras/stz1318 MNRAS, 414, 888. doi:10.1111/j.1365-2966.2011.18496.x Noordermeer, E. 2008, MNRAS, 385, 1359. Emsellem, E., Cappellari, M., Peletier, R. F., et al. 2004, doi:10.1111/j.1365-2966.2008.12837.x MNRAS, 352, 721. doi:10.1111/j.1365-2966.2004.07948.x Noordermeer, E., van der Hulst, J. M., Sancisi, R., et al. 2007, MNRAS, 376, 1513. Erroz-Ferrer, S., Knapen, J. H., Leaman, R., et al. 2016, MNRAS, 458, 1199. doi:10.1093/mnras/stw300 doi:10.1111/j.1365-2966.2007.11533.x Falc on-Barroso, J., S anchez-Bl azquez, P., Vazdekis, A., et Oogi, T. & Habe, A. 2013, MNRAS, 428, 641. doi:10.1093/mnras/sts047 al. 2011, A&A, 532, A95. doi:10.1051/0004-6361/201116842 Oort, J. H. 1940, ApJ, 91, 273. doi:10.1086/144167 Feng, J. Q. & Gallo, C. F. 2011, Research in Astronomy and Penoyre, Z., Moster, B. P., Sijacki, D., et al. 2017, MNRAS, Astrophysics, 11, 1429. doi:10.1088/1674-4527/11/12/005 468, 3883. doi:10.1093/mnras/stx762 Genzel, R., Price, S. H., Ubler, H., et al. 2020, ApJ, 902, Persic, M., Salucci, P., & Stel, F. 1996, MNRAS, 281, 27. 98. doi:10.3847/1538-4357/abb0ea doi:10.1093/mnras/278.1.27 Girardi, L., Bressan, A., Bertelli, G., et al. 2000, A&AS, Puech, M., Flores, H., Hammer, F., et al. 2008, A&A, 484, 141, 371. doi:10.1051/aas:2000126 173. doi:10.1051/0004-6361:20079313 18 Yoon et al. (2021) Rhee, J., Smith, R., Choi, H., et al. 2017, ApJ, 843, 128. Thomas, T. & Katgert, P. 2006, A&A, 446, 31. doi:10.3847/1538-4357/aa6d6c doi:10.1051/0004-6361:20053661 Richardson, W. H. 1972, Journal of the Optical Society of Tiley, A. L., Swinbank, A. M., Harrison, C. M., et al. 2019, America (1917-1983), 62, 55 MNRAS, 485, 934. doi:10.1093/mnras/stz428 Rubin, V. C., Burstein, D., Ford, W. K., et al. 1985, ApJ, Trimble, V. 1987, ARA&A, 25, 425. 289, 81. doi:10.1086/162866 doi:10.1146/annurev.aa.25.090187.002233 Rubin, V. C., Ford, W. K., & Thonnard, N. 1980, ApJ, van de Hulst, H. C., Raimond, E., & van Woerden, H. 1957, 238, 471. doi:10.1086/158003 BAN, 14, 1 S anchez, S. F., Garc a-Benito, R., Zibetti, S., et al. 2016, Vazdekis, A., S anchez-Bl azquez, P., Falc on-Barroso, J., et A&A, 594, A36. doi:10.1051/0004-6361/201628661 al. 2010, MNRAS, 404, 1639. S anchez, S. F., Kennicutt, R. C., Gil de Paz, A., et al. doi:10.1111/j.1365-2966.2010.16407.x 2012, A&A, 538, A8. doi:10.1051/0004-6361/201117353 Veale, M., Ma, C.-P., Greene, J. E., et al. 2017, MNRAS, S anchez-Bl azquez, P., Peletier, R. F., Jim enez-Vicente, J., 471, 1428. doi:10.1093/mnras/stx1639 et al. 2006, MNRAS, 371, 703. Volders, L. M. J. S. 1959, BAN, 14, 323 doi:10.1111/j.1365-2966.2006.10699.x Wake, D. A., Bundy, K., Diamond-Stanic, A. M., et al. Shen, S., Mo, H. J., White, S. D. M., et al. 2003, MNRAS, 2017, AJ, 154, 86. doi:10.3847/1538-3881/aa7ecc 343, 978. doi:10.1046/j.1365-8711.2003.06740.x Wechsler, R. H., Bullock, J. S., Primack, J. R., et al. 2002, Shepp, L. A., Vardi, Y. 1982, IEEE Transactions on ApJ, 568, 52. doi:10.1086/338765 Medical Imaging, 1, 113. doi: 10.1109/TMI.1982.4307558 Whitmore, B. C. & Forbes, D. A. 1989, Ap&SS, 156, 175. Smee, S. A., Gunn, J. E., Uomoto, A., et al. 2013, AJ, 146, doi:10.1007/BF00646362 32. doi:10.1088/0004-6256/146/2/32 Whitmore, B. C., Forbes, D. A., & Rubin, V. C. 1988, ApJ, Smith, R., Choi, H., Lee, J., et al. 2016, ApJ, 833, 109. 333, 542. doi:10.1086/166765 doi:10.3847/1538-4357/833/1/109 Wiegert, T. & English, J. 2014, NewA, 26, 40. Sofue, Y. & Rubin, V. 2001, ARA&A, 39, 137. doi:10.1016/j.newast.2013.04.006 doi:10.1146/annurev.astro.39.1.137 Yan, R., Bundy, K., Law, D. R., et al. 2016, AJ, 152, 197. Springel, V., White, S. D. M., Tormen, G., et al. 2001, doi:10.3847/0004-6256/152/6/197 MNRAS, 328, 726. doi:10.1046/j.1365-8711.2001.04912.x Yoon, Y., Im, M., & Kim, J.-W. 2017, ApJ, 834, 73. Swaters, R. A., Sancisi, R., van Albada, T. S., et al. 2009, doi:10.3847/1538-4357/834/1/73 A&A, 493, 871. doi:10.1051/0004-6361:200810516

http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png

Astrophysics arXiv (Cornell University)

http://www.deepdyve.com/lp/arxiv-cornell-university/rotation-curves-of-galaxies-and-their-dependence-on-morphology-and-8vO1uJSKcN

Rotation Curves of Galaxies and Their Dependence on Morphology and Stellar Mass

Loading next page...

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

eISSN: ARCH-3330
Publisher site: See Article on Publisher Site

Abstract

Journal

Astrophysics – arXiv (Cornell University)

Published: Oct 12, 2021

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Rotation Curves of Galaxies and Their Dependence on Morphology and Stellar Mass

Rotation Curves of Galaxies and Their Dependence on Morphology and Stellar Mass

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Rotation Curves of Galaxies and Their Dependence on Morphology and Stellar Mass

Rotation Curves of Galaxies and Their Dependence on Morphology and Stellar Mass

References

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies