Superluminous Spiral Galaxies
We report the discovery of spiral galaxies that are as optically luminous as elliptical brightest cluster galaxies, with -band monochromatic luminosity ( erg s). These super spiral galaxies are also giant and massive, with diameter kpc and stellar mass . We find 53 super spirals out of a complete sample of 1616 SDSS galaxies with redshift and . The closest example is found at . We use existing photometry to estimate their stellar masses and star formation rates (SFRs). The SDSS and WISE colors are consistent with normal star-forming spirals on the blue sequence. However, the extreme masses and rapid SFRs of yr place super spirals in a sparsely populated region of parameter space, above the star-forming main sequence of disk galaxies. Super spirals occupy a diverse range of environments, from isolation to cluster centers. We find four super spiral galaxy systems that are late-stage major mergers–a possible clue to their formation. We suggest that super spirals are a remnant population of unquenched, massive disk galaxies. They may eventually become massive lenticular galaxies after they are cut off from their gas supply and their disks fade.
The most massive galaxies in the universe are thought to form from the largest density peaks in the primordial matter distribution. Galaxy mergers change the initial galaxy mass function, forming more massive galaxies by combining less massive ones. The result is the galaxy mass distribution we see in the local universe, empirically described by the Schechter (1976) luminosity function, together with a morphology-dependent mass-to-light ratio. The luminosity function also depends upon the star formation history of galaxies, regulated by gas content, gas accretion, stellar feedback, and active galactic nucleus (AGN) feedback. Galaxy mergers play an important role here too, since tidal torques in merging systems force gas into the galaxy centers, leading to starburst activity that grows the stellar bulge and AGN activity that grows the supermassive black hole (Toomre & Toomre, 1972; Barnes & Hernquist, 1991; Hopkins et al., 2009).
Galaxies segregate into two major classes based on color and morphology (Strateva et al., 2001; Lintott et al., 2008). Blue, star-forming disks (late-type galaxies, LTGs) lie in one region of color-space called the blue-sequence. Red-and-dead spheroids (early-type galaxies, ETGs) lie in a different region of color space called the red sequence. LTGs demonstrate a correlation between star formation rate (SFR) and stellar mass () called the star-forming main sequence (SFMS: Brinchmann et al., 2004; Elbaz et al., 2007; Wuyts et al., 2011). The SFMS may be a consequence of an equilibrium between inflowing gas and star-formation driven outflows, with the specific star formation rate (SSFR) regulated by the halo mass growth rate (Lilly et al., 2013). Most ETGs on the other hand have much lower SFRs because they lack the cold gas needed to sustain star formation.
It appears that there is a limit to the mass of star-forming disk galaxies of roughly , with the most massive disk galaxies transitioning away from the main sequence as their SSFR declines. This decline appears to be a gradual process, occurring over a period longer than 1 Gyr after the gas supply to the galaxy disk has been interrupted (Schawinski et al., 2014). Rapid quenching does not appear to occur for most galaxies that remain disk galaxies, contrary to early attempts to explain the apparently bimodal distribution of galaxy colors.
A number of mechanisms have been suggested to explain why the gas supply is interrupted for the most massive disk galaxies. Major galaxy mergers may disrupt merging disk galaxies and transform them rapidly into elliptical galaxies (Baldry et al., 2004), though this does not explain the transformation of galaxies that remain disks. The accretion of cold gas onto a galaxy may be stopped when the galaxy halo becomes massive enough that accretion shocks develop, interrupting the cold streams of gas needed to replenish the disk (Dekel & Birnboim, 2006). Increasing AGN feedback from a growing supermassive black hole may shock or eject gas from the galaxy disk, reducing its capacity to form stars (Hopkins et al., 2006; Ogle, Lanz, & Appleton, 2014). Ram-pressure stripping of the interstellar medium (ISM) by the intercluster medium (ICM) of a galaxy cluster can also remove cold gas (Sivanandam, Rieke, & Rieke, 2014).
Studying the most massive spiral galaxies can give us clues as to which of the above evolutionary processes are primarily responsible for converting star-forming disk galaxies into red-and-dead lenticulars or ellipticals. The existence of rapidly star-forming, massive spirals with indicates that disk galaxies can postpone this fate under special circumstances. We present here the most optically luminous and biggest spiral galaxies at redshift , found by mining the NASA/IPAC Extragalactic Database (NED). We assume a cosmology with , , and for computing all linear sizes and luminosities.
This project is an offshoot of our work to determine the completeness of NED and explore its potential for systematic studies of galaxy populations (Ogle, P. et al., in preparation). NED provides a unique fusion of multi-wavelength photometry from Galaxy Evolution Explorer (GALEX), Sloan Digital Sky Survey (SDSS), and the 2-Micron All-Sky Survey (2MASS), among others, which we augment by Wide-field Infrared Survey Explorer (WISE) photometry, that allows us to estimate stellar masses and SFRs. We compared the redshift distribution of galaxies in NED at to a model redshift distribution for the universe derived using a redshift-independent luminosity function, in order to estimate the spectroscopic completeness of NED. We used the Schechter (1976) luminosity function fits of Blanton et al. (2003), which are based on 150,000 SDSS galaxies as our benchmark. We adopt their characteristic absolute magnitude value of ( erg s at 6200) for the SDSS -band luminosity function. The redshift limit was made large enough to capture the rarest, most luminous galaxies, but not so large as to require consideration of redshift evolution in the luminosity function.
2.1. SDSS r-band Selection of the most Optically Luminous Galaxies
SDSS is the largest source of spectroscopic redshifts, with a spectroscopic selection limit of (Strauss et al. 2002). We find that NED is complete over the SDSS footprint out to for galaxies with , the most optically luminous and massive galaxies in the low-redshift universe. Our sample is chosen from the 797,729 galaxies (typeG) in NED with spectroscopic redshifts , in the SDSS footprint, and detected in SDSS band. We apply Galactic extinction corrections (tabulated by NED) and K-corrections to the -band magnitudes prior to making our sample selection. We find 1616 galaxies with redshift and luminosity , which constitute our Ogle et al. Galaxy Catalog (OGC). The most luminous galaxy in the OGC is a elliptical brightest cluster galaxy (BCG).
2.2. UV Selection Method for Super Spiral Galaxies
We make a further selection for UV emission because we are interested in finding the most massive, actively star-forming disk galaxies. We recently matched and integrated the GALEX All-Sky Survey Catalog (GASC) and GALEX Medium Sky Survey Catalog (GMSC) with NED, using an automated, statistical algorithm (Ogle et al., 2015). We inspected the SDSS images of all 196 galaxies from the OGC that are detected in the GALEX NUV band (the OGC-UV subsample). Of these, we find 46 NUV-detected, galaxies with spiral morphology (Table 1). The remaining NUV sources include 118 ellipticals, 11 galaxies with E+A spectra, 2 quasi-stellar objects (QSOs) with extended emission, and 19 galaxies with erroneous redshifts or magnitudes. The most luminous elliptical galaxy in OGC-UV is a BCG, while the most luminous spiral galaxy has .
2.3. Morphological Selection Method for Super Spiral Galaxies
We inspected the 310 brightest galaxies of the full OGC sample, those with , to see if we are missing any spirals with UV selection. Of these, we classified 11 super spirals, 253 ellipticals, 38 galaxies with erroneous redshifts or magnitudes, 6 lenticulars, 1 irregular, and 1 E+A galaxy. This inventory includes 7 additional super spirals (Table 1) that are not in the OGC-UV sample. Of these, 4 have no GALEX sources nearby, and 3 others have nearby GALEX sources that should be matched in NED, but are not, possibly because of confusion. This shows that our NUV selection, while relatively efficient (47/196) compared to morphological selection (11/310), leads to an incomplete sample, with only 4/11 super spirals recovered in this luminosity range. Part of the incompleteness (3/11) is owing to incomplete matching of GALEX with NED, while the rest (4/11) may be attributed to the GALEX detection limit or coverage.
We conduct our investigation of super spirals primarily with photometry compiled by NED. We use SDSS DR6 , , , , photometry measured with the CModel method, which combines exponential plus deVaucouleurs model fitting. GALEX FUV and NUV photometry is taken from the GASC and GMSC, measured within a Kron elliptical aperture. We use 2MASS , , total magnitudes from the 2MASS Extended Source Catalog (2MASX). NED objects are matched to AllWISE sources using the Gator tool in the Infrared Science Archive (IRSA). We use AllWISE 4.6 and 12 m photometry within the largest available fixed-radius aperture of , which is well-matched to the largest galaxy in our sample, with semimajor axis .
4. Basic Properties of Super Spirals
4.1. Optical and Mid-IR colors
The SDSS and WISE colors of super spirals lie along the blue sequence, similarly to less luminous star-forming disk galaxies (Figure 1(a)). The SDSS comparison sample is adopted from Alatalo et al. (2014), who show that LTGs and ETGs classified by Galaxy Zoo (GZ) (Lintott et al., 2008) are well-separated in WISE [4.6] vs. SDSS color space. The WISE [4.6] color ranges from 2.0 to 4.2, typical of polycyclic aromatic hydrocarbon (PAH) and warm dust emission from gas-rich, actively star-forming galaxies. The color ranges from 1.4 to 4.4, indicating star-forming disks with a range of SSFR or dust extinction. We estimate differential K-corrections of mag in the redshift range , by convolving several spectral energy distribution (SED) models (e.g., those in the Appendix) with the SDSS filter curves. These corrections are not large enough to explain the additional scatter in the observed colors of super spirals.
There is a shift in the locus of super spiral colors compared to less-massive blue sequence galaxies. Super spirals tend to have redder and bluer [4.6] colors compared to the SDSS distribution. This could in principle indicate either lower SSFR or increased extinction. However, the high SSFR of our sample (Figure 1(b)) runs contrary to the first explanation. Six super spirals have , a value not attained by less luminous SDSS LTGs. The two reddest galaxies (SS 53 and SS 09) may be misclassified peculiar elliptical galaxies with prominent shells. CGCG 122-067 (SS 50) may be redder because of its double bulge. The other 4 are clearly spirals, and require further investigation and custom photometry to determine the cause of their unusually red colors.
4.2. Stellar Mass and Star Formation Rate
We estimate stellar mass from 2MASS luminosity together with an SDSS color-dependent mass-to-light ratio estimated using the prescription of Bell et al. (2003), giving . We apply a small correction to the stellar masses to convert to a Chabrier initial mass function (IMF). This yields stellar masses that are consistent with more sophisticated SED template fitting (Appendix). We find stellar masses in the range .
We estimate the SFR from the WISE 12 m luminosity using the prescription of Chang et al. (2015), which was established by SED-fitting more than 630,000 SDSS galaxies with magphys (da Cunha, Charlot, & Elbaz, 2008). While accurate for star-forming galaxies, this method may overestimate the star-formation rate for early type galaxies where dust may be heated by other sources not directly related to star formation, or in the presence of a luminous AGN. We further validate our WISE single-band SFRs against magphys SED-fitting for two representative super spirals (Appendix). The WISE 12 m monochromatic luminosities of super spirals range from erg s (), corresponding to SFRs of yr.
We compare our sample to the SDSS-WISE sample of Chang et al. (2015), who estimated SFR and with magphys. We find that most super spirals lie well above an extrapolation of the star-forming main sequence to higher mass (Figure 1(b)). This is a region of the SFR vs. mass diagram that is very sparsely populated. The vast majority of SDSS disk galaxies in this mass range have significantly lower SFR and SSFR.
Our -band luminosity plus NUV detection criteria tend to select galaxies with high global star formation rates. However, the SDSS spectra reveal a relatively old bulge stellar population for most super spirals. We do find an indication of starburst activity in the SDSS bulge spectra of 3 super spirals (SS 05–see Appendix, SS 15, and SS 44) with strong young stellar population contributions and high-equivalent width H emission. These three galaxies also have relatively blue SDSS colors and red WISE [4.6] colors, both indicative of a high global SSFR.
4.3. Active Galactic Nuclei
The super spiral galaxies in our sample contain 3 Seyfert 1 nuclei and 2 QSOs with broad Balmer lines and strong [O iii] in their SDSS spectra (Table 1). There is also 1 Seyfert 2 nucleus with strong [O iii] but narrow Balmer lines. There is likely a dominant contribution from the QSO to the IR luminosity of 2MASX J15430777+1937522 (SS 10), which has the greatest WISE 12 m luminosity of our sample ( erg s or ). The two QSOs are also detected at X-ray wavelengths by ROSAT. One additional galaxy (2MASX J10095635+2611324 SS 45) is detected in X-rays, but has no obvious signature of an AGN in its SDSS nuclear spectrum. There is so far no indication of any extended X-ray emission associated with super spirals, though none have been specifically targeted for this. It will be important to make deep X-ray observations of super spirals to quantify any X-ray halo emission in comparison to giant elliptical galaxies. Only two super spirals are detected by the NVSS radio survey (2MASX J14472834+5908314 SS 47 and CGCG 122-067 SS 50), but the resolution is insufficient to distinguish between radio emission from star formation activity or from a radio jet. The presence of luminous AGNs in 11% of super spirals indicates that they are continuing to grow their supermassive black holes. It is imperative to measure the distribution of bulge and supermassive black hole masses in our super spiral sample to see if they follow the same relation as lower-mass spiral bulges.
4.4. Size, SFR Surface Density, and Morphology
The sizes of super spirals range from 57 to 134 kpc, with a median size of 72 kpc, using the SDSS DR6 -band isophotal diameter at 25.0 mag arcsec (Table 1 and Figure 2). Their deprojected SFR surface densities range from to yr kpc. A plot of SFR vs. diameter shows considerable scatter (Figure 2). However, the five most rapidly star-forming galaxies, with , all have diameters kpc. The most MIR-luminous super spiral (SDSS J094700.08+254045.7, see Appendix), also has the largest deprojected SFR surface density. The largest super spiral, 2MASX J16394598+4609058 (SS 03, Figure 3), has a diameter of 134 kpc and a relatively low SFR surface density of yr kpc.
Super spirals display a range of morphologies, from flocculent to grand-design spiral patterns (Figure 4). At least 9 super spirals have prominent stellar bars visible in the SDSS images (Table 1: Notes). There are morphological peculiarities in several cases, including one-arm spirals, multi-arm spirals, rings, and asymmetric spiral structure (Figure 4 and Table 1). These types of features may indicate past or ongoing galaxy mergers or collisions.
4.5. Bulge-disk Decomposition
We make use of the bulge-disk decompositions of Simard et al. (2011) to quantify the relative contributions of the bulge and disk to the luminosity of super spirals (Table 2). The galaxy and band SDSS images are jointly fit by a de Vaucouleurs profile for the bulge (Sersic index ), plus an exponential disk. We compare super spirals to a representative subsample of 4686 spiral galaxies with classified by Galaxy Zoo (GZ, Figure 5), with bulge-disk decompositions also by Simard et al. (2011). We find a much narrower distribution of -band bulge to total luminosity () for super spirals, with a median value of , and a deficit of values . A Kolmogorov-Smirnoff (K-S) test shows that the distributions differ significantly, with a probability of only 0.0027 that super spirals are drawn from the same population as GZ spirals. The lack of super spirals with may be consistent with a past history of significant merger activity. The bulge ellipticity distribution of super spirals is not significantly different from that of GZ spirals (Figure 5(c)). We note that since the profile fits do not include a bar component, the ellipticities may be augmented by the presence of a bar or double bulge.
The disk inclination distribution of super spirals also differs significantly from that of GZ spirals, and from the expected dependence (Figure 5(b)). A K-S test gives a probability of that super spirals and GZ spirals are drawn from the same inclination distribution. Only 5 (9%) of super spirals have inclinations of , compared to the expectation of 50% for randomly oriented disks. This indicates that we are missing roughly 45% of the super spirals in our luminosity range, possibly because of internal extinction at the NUV selection wavelength. The GZ spiral inclination distribution also differs from the expectation for randomly oriented disks, with an excess at inclinations that may reveal a bias for GZ to classify edge on disks as spirals or to misclassify edge-on lenticulars as spirals.
The median disk exponential scale length of super spirals is 12.2 kpc, 2.3 times as large as the 5.3 kpc median for GZ spirals, confirming the giant disk sizes of super spirals (Figure 5d). A K-S test gives a probability of that super spirals and GZ spirals are drawn from the same size distribution. The galaxy smoothness parameter (Simard et al., 2009), which quantifies the fractional residuals to the model fit inside two half-light radii, is in r band. The and parameters of bulge-disk decompositions have been used by others to quantitatively select early-type galaxies, with and as criteria (Simard et al., 2009). Several super spirals in our sample meet these criteria, but we are nevertheless confident of the detection of a significant spiral disk in most of these cases.
5. Galaxy Merger Candidates
We find four super spiral merger candidates with apparent double stellar bulges or double nuclei (Figure 6). The SDSS spectra only cover the dominant or central bulge or nucleus of each system. Spectroscopy of the secondary bulges or nuclei will be necessary to confirm or rule out these merger candidates as true physical pairs or multiples.
The merger candidate 2MASX J08542169+0449308 (SS 22) appears to be a nearly equal mass major spiral pre-merger. The arms of both spirals are wound in the same direction, and the disks appear to be overlapping in the plane of the sky. The stretched out spiral arms of both spiral galaxy components, together with an apparent tidal arm at PA (measured counterclockwise from North) suggest an ongoing tidal interaction.
The merger candidate 2MASX J16014061+2718161 (SS 37) is a BCG, surrounded by several other disk galaxy companions (Figures 6 and 7). The host cluster is identified as GMBCG J240.41924+27.30444, with a photometric redshift of 0.193 (Table 3). There are clear distortions to the spiral structures of both spiral galaxy components that appear to be involved in this merger.
The merger candidate 2MASX J09334777+2114362 (SS 46) appears to be a double AGN system. The primary, central nucleus is identified as an SDSS QSO. The secondary nucleus has similar flux and color to the primary nucleus, but it does not have an SDSS spectrum to confirm that it is a true physical double AGN. The galaxy disk has high surface brightness, suggestive of starburst activity. Faint outer spiral arms are also suggestive of a recent galaxy interaction.
The merger candidate and BCG CGCG 122-067 (SS 50) appears to be a late stage :1 major merger. The double bulge is surrounded by a common inner disk. Two giant spiral arms emerge from this central disk, one from each bulge, making a complete circuit around the disk. A large gap is seen between the arms at PA. There are three other possible merging nuclei, including a bright green point source at PA, that raise the possibility that this is a five-component multiple merger system. Such multiple mergers are reminiscent of the elliptical nest galaxies that are sometimes found at the centers of galaxy clusters.
We checked NED for known galaxy clusters and groups within of each super spiral (Table 3). Seven of the super spirals are candidate BCGs, within of a galaxy cluster. Two are candidate brightest group galaxies (BGGs), within of a compact galaxy group. Most of the clusters only have photometric redshifts and have yet to be verified spectroscopically. However, the photometric redshifts are all within of the super spiral spectroscopic redshift, which suggests a true physical association. The two associations of super spirals with compact groups are only based on their small angular separation, with no independent redshift available for the groups.
We used NED’s Environment Tool to further explore the environments of the super spiral BCG and BGG candidates. This tool performs a redshift-constrained cone search for galaxies and galaxy clusters within a sphere of comoving radius 10 Mpc. Because of the high redshifts of the super spirals, only the most luminous galaxies in their neighborhoods will tend to have measured spectroscopic redshifts in NED. We tabulate the number of galaxies (N1) with spectroscopic redshifts that are within 1 Mpc and 500 km s, and the number (N10) within 10 projected Mpc and 5000 km s. The MSPM 05544 galaxy cluster, which appears to host the super spiral CGCG 122-067 (SS 50) has the largest number of cluster members with spectroscopic redshifts in NED (302), while the SDSSCGB 59704 galaxy group has the smallest number (2). These numbers should be taken as lower limits to the cluster membership, depending primarily on the SDSS spectroscopic selection limit and redshift.
There are likely more clusters to be discovered in the vicinity of super spirals. For example, a clear overdensity of galaxies is seen to the SE of 2MASX J11535621+4923562 (SS 17, Figure 8). We verify a concentration of 69 galaxies within 10 Mpc and 5000 km s (Table 3: OGC 0586 CLUSTER), using NED’s Environment tool. We estimate the mean redshift of OGC 0586 CLUSTER to be , from 12 galaxies with spectroscopic redshifts that are within 5 projected Mpc of SS 17.
While super spirals have similar structure to less luminous spiral galaxies, they are impressive in the vastness of their scale. A sense of how truly enormous these galaxies are can be gained by comparison to other galaxies in the same cluster (Figure 8: OGC 0586 CLUSTER). The 2MASX J115356214923562 (SS 17) super spiral at , with a luminosity of and a diameter of 90 kpc, can be compared to a more common, less luminous spiral galaxy which has and a diameter of 39 kpc (), at about the same redshift ().
It is natural to ask whether any analogs to super spirals have been found at lower redshift. One well-known example of a giant spiral galaxy is Malin 1 (), initially suggested to be a proto-disk galaxy because of its massive H i disk (Bothun et al., 1987). While Malin 1 does have one of the largest stellar disks known, with an exponential scale length of 70 kpc, its global -band luminosity () is not nearly great enough to make it into the OGC catalog. Its disk has very low surface brightness and is not readily visible in SDSS images. As further points of comparison, we estimate a global stellar mass of and global SFR of yr, which are both much lower than the range spanned by super spirals.
Other giant spiral galaxies are found in the local universe, though they also have considerably lower luminosities than the super spirals in our sample. Romanishin (1983) find 107 spiral galaxies in the Uppsala General Catalog of Galaxies (UGC) at , with cosmology-corrected -band isophotal diameters (at mag arcsec) of kpc, similar to super spirals. The 39 giant UGC spirals with SDSS photometry in NED have -band luminosities of , stellar masses of to yr, and SFRs of yr. Because of their considerably lower stellar masses, they cannot be faded super spirals, but could be useful analogs for understanding giant disks. One of the largest giant spiral galaxies, UGC 2885 has a rotational velocity of 280 km s at a radius of 60 kpc, and has undergone fewer than 10 rotations at its outer edge in the age of the universe (Rubin, Ford, & Thonnard, 1980).
Super spirals may also be related to the cold sub-mm galaxies (SMGs) discovered at redshift (Chapman et al., 2002). The relatively cold ( K) dust temperatures of these SMGs may indicate starburst activity in a disk rather than a spheroid. In comparison, the FIR SED of super spiral 2MASX J13275756+3345291 (SS 05, see Appendix) is fit by the sum of a cold dust component with K, likely from the disk, and a warmer dust component with temperature K, likely from the starbursting bulge. GN20, one of the most luminous sub-mm detected star bursting galaxies, shows molecular gas and star formation distributed in a 10 kpc scale disk at (Carilli, C. L. et al., 2010). Deep near to mid-IR imaging of SMGs at intermediate redshifts will be necessary to measure their sizes and stellar masses and better determine their relationship to super spirals.
7.2. Formation and Survival
We estimate an average super spiral number density of Gpc at , correcting for incompleteness at high disk inclination (§4.5). The space density of super spirals is therefore only of the space density of elliptical galaxies in the same -band luminosity range. Even the largest galaxy evolution simulations to date, such as the Illustris simulation (Vogelsberger et al., 2014; Snyder et al., 2015), covering Gpc, are not big enough to manufacture a significant number of super spirals. Therefore, no adequate prediction exists for the expected number of super spirals at , nor are there simulations showing how these giant disk galaxies might form.
Super spirals could be formed by gas-rich major spiral-spiral mergers. Simulations that collide two gas-rich disk galaxies are able to produce post-merger spiral galaxies, albeit at smaller scale (Barnes, 2002; Springel & Hernquist, 2005; Robertson et al, 2006; Hopkins et al., 2009). While merging stellar disks are typically destroyed, the gas in the outer disks may combine to reform an even larger gas and stellar disk. Orbital geometry may also be important, with misaligned or retrograde orbits leading to more gas-rich final merger products. If the dynamical timescales are longer and the merger-induced torques are even weaker in the outer disks of super spiral mergers, this may also be conducive to the preservation of gas disks and reformation of stellar disks. Alternatively, super spirals might be formed more gradually, from the inside out by accretion of cold gas. This may require a relatively low halo mass in order to avoid accretion shocks, which might prevent the gas from settling onto the outer disk Dekel & Birnboim (2006). It will be important to study the spatial distribution of both neutral gas and star formation in super spirals to gain further insight into how their disks are formed.
It appears that the super spirals in our sample have so far avoided the fate of the vast majority of the most massive galaxies and continue to form stars in spite of their extreme mass, bucking the trend of cosmic downsizing. There are several possible reasons for this success. First, super spirals may be robust to mergers because of their massive, dissipative gaseous disks. It appears that several super spirals in our sample have survived recent major mergers with their star-forming disks intact. Second, the supermassive black holes in super spiral bulges may not be large enough to provide enough feedback to drive away the gas in the giant galaxy disk. Third, the halo mass may not be large enough to cut off cold accretion onto the disk via accretion shocks. Finally, a large enough gas reservoir may have already settled into the disk to fuel star formation for a long time into the future. Observations across the electromagnetic spectrum are called for to distinguish among these possibilities.
7.3. Connection to Quenched Disk Galaxies
Super spirals occupy a relatively empty corner of the SFR vs. stellar mass diagram (Figure 1b). They lie above an extrapolation of the star-forming main sequence, at the most extreme mass and SFR. We find that most super spirals have SSFR Gyr. They are forming stars at a rate that would allow them to build up their mass in less than the age of the universe. This is unlike similarly massive, yet much more common disk galaxies (early type spirals and lenticulars) that fall below the star-forming main sequence, in what we shall call the disk quenching sequence (DQS: the disk galaxy subset of the green valley population). The disk-quenching sequence is discussed in the context of SSFR and UV color evolution by Schawinski et al. (2014), and in the context of IR color evolution by Alatalo et al. (2014). Quenching disk galaxies are likely greatly reduced in their ability to form stars because their supply of cold gas has been cut off (e.g. Dekel & Birnboim, 2006).
The most densely populated ridge of the DQS is close to the median stellar mass of our super spiral sample (). We suggest that the majority of disk galaxies along this ridge were once super spirals. At a minimum, galaxies of this mass must have attained an average SFR yr in order to have formed in less than the age of the universe. This would put them squarely in the SFR and SSFR range of super spirals. A further implication is that their kpc diameter stellar disks must have faded dramatically. If fossil giant disks are detected around massive lenticular galaxies with deep imaging, it will provide strong evidence for this hypothesis. In addition, deep H i and CO observations may reveal if their cold gas reservoir has been entirely depleted or reduced to a level that is not conducive to star formation.
We report the discovery of a large sample of the most optically luminous (), biggest, and most massive spiral galaxies in the universe, which we call super spirals. These galaxies are very rare ( Gpc ), but are easily observed out to because of their high luminosities and gigantic sizes. Super spirals are forming stars at yr, a rate greater than their mean SFR over the age of the universe. Bulge-disk decompositions confirm the presence of giant stellar disks, with a median exponential scale length of 12.2 kpc, 2.3 times the median scale length of less luminous spirals at the same redshift. The bulge-to-total optical luminosity distribution is also significantly different for super spirals, showing a deficit of galaxies with , and a concentration of galaxies with . Roughly 11% of super spirals have Seyfert or QSO nuclei, suggesting that they are still actively adding mass to their supermassive black holes. We find evidence that several super spirals are undergoing major mergers, but manage to keep their star-forming disks intact, and avoid being transformed in to red-and-dead elliptical galaxies. Some super spirals are brightest cluster galaxies, while others appear to be isolated in the field. We suggest that super spirals may be the progenitors of red and dead lenticular galaxies of similar mass.
- Alatalo et al. (2014) Alatalo, K., Cales, S. L., Appleton, P. N. et al. 2014, ApJ, 794, L13
- Baldry et al. (2004) Baldry, I. K., Glazebrook, K., Brinkmann, J., et al. 2004, ApJ, 600, 681
- Barnes (2002) Barnes, J. 2002, MNRAS, 333, 481
- Barnes & Hernquist (1991) Barnes, J. E. & Hernquist, L. E. 1991, ApJ, 370, L65
- Bell et al. (2003) Bell, E. F., McIntosh, D. H., Katz, N., & Weinberg, M. D. 2003, ApJS, 149, 289
- Blanton et al. (2003) Blanton, M. R., Hogg, D. W., Bahcall, N. A, et al. 2003, ApJ, 592, 819
- Bothun et al. (1987) Bothun, G. D., Impey, C. D., Malin, D. F., & Mould, J. R. 1987, AJ, 94, 23
- Brinchmann et al. (2004) Brinchmann, J., Charlot, S., White, S. D. M., et al. 2004, MNRAS, 351, 1151
- Carilli, C. L. et al. (2010) Carilli, C. L., Daddi, E., Riechers, D., et al. 2010, ApJ, 714, 1407
- Chang et al. (2015) Chang, Y.-Y., van der Wel, A., da Cunha, E., & Rix, H.-W. 2015, ApJS, 219, 8
- Chapman et al. (2002) Chapman, S. C., Smail, I., Ivison, R. J., Helou, G., Dale, D. A., & Lagache, G. 2002, ApJ, 573, 66
- da Cunha, Charlot, & Elbaz (2008) da Cunha, E., Charlot, S., & Elbaz, D. 2008, MNRAS, 339, 1595
- Dekel & Birnboim (2006) Dekel, K. & Birnboim, Y. 2006, MNRAS, 368, 2
- Elbaz et al. (2007) Elbaz, D., Daddi, E., Le Borgne, D. et al. 2007, A&A, 468, 33
- Hopkins et al. (2009) Hopkins, P. F., Cox, T. J., Younger, J. D., & Hernquist, L. 2009, ApJ, 691, 1168
- Hopkins et al. (2006) Hopkins, P. F., Hernquist, L., Cox, T. J., Di Matteo, T., Robertson, B., & Springel, V. 2006, ApJS, 163, 1
- Joye & Mandel (2003) Joye, W. A., & Mandel, E. 2003, ASPCS, 295, 489, eds. Payne, H. E., Jedrezejewsk, R. I. & Hook, R. N.
- Lilly et al. (2013) Lilly, S. J., Carollo, C. M., Pipino, A., Renzini, A., & Peng, Y. 2013, ApJ, 772, 119
- Lintott et al. (2008) Lintott, C. J., Schawinski, K., Slosar, A., et al. 2008, MNRAS, 389, 1179
- Ogle et al. (2015) Ogle, P. M., Mazzarella, J., Ebert, R., Fadda, D., Lo, T., Terek, S., Schmitz, M., & The NED Team. 2015, ASPCS, 495, 25, eds. Taylor, A. R., & Rosolowsky, E.
- Ogle, Lanz, & Appleton (2014) Ogle, P., Lanz, L., & Appleton, P. 2014, ApJ, 788, L33
- Robertson et al (2006) Robertson, B., Bullock, J. S., Cox, T. J. , et al. 2006, ApJ, 645, 986
- Romanishin (1983) Romanishin, W., 1983, MNRAS, 204, 909
- Rubin, Ford, & Thonnard (1980) Rubin, V. C., Ford, W. K., & Thonnard, N. 1980, ApJ, 238, 471
- Schawinski et al. (2014) Schawinski, K., Urry, C. M., Simmons, B. D., et al. 2014, MNRAS, 440, 889
- Schechter (1976) Schechter, P. 1976, ApJ, 203, 297
- Simard et al. (2011) Simard, L., Mendel, J. T., Patton, D. R., Ellison, S. L., & McConnachie, A. W. 2011, ApJS, 196, 11
- Simard et al. (2009) Simard, L., Clowe, D., Desai, V., et al. 2009, A&A, 508, 1141
- Sivanandam, Rieke, & Rieke (2014) Sivanandam, S., Rieke, M. J., & Rieke, G. H. 2014, ApJ, 796, 89
- Snyder et al. (2015) Snyder, G. F., Torrey, P., Lotz, J. M. et al. 2015, MNRAS, 454, 1886
- Springel & Hernquist (2005) Springel, V., & Hernquist, L. 2005, ApJ, 622, L9
- Stoughton et al. (2002) Stoughton, C., Lupton, R. H., Bernardi, M., et al. 2002, AJ, 123, 485
- Strateva et al. (2001) Strateva, I., Ivezic, Z., Knapp, G. R., et al. 2001, AJ, 122, 1861
- Strauss et al. (2002) Strauss, M. A., Weinberg, D. H., Lupton, R. H. et al., AJ, 124, 1810
- Toomre & Toomre (1972) Toomre, A. & Toomre, J. 1972, ApJ, 178, 623
- Vogelsberger et al. (2014) Vogelsberger, M., Genel, S., Springel, V. et al. 2014, MNRAS, 444, 1518
- Wuyts et al. (2011) Wuyts, S., Forster Schreiber, N. M., van der Wel, A., et al. 2011, ApJ, 742, 96
- Wyder et al. (2005) Wyder, T. K., Treyer, M. A., Milliard, B., et al. 2005, ApJ, 619, L15
|SS||OGC||NED Name||()||aaIsophotal diameter (kpc) at mag arcsec.||bb () or upper limit, based on 2MASS luminosity and color.||SFRcc SFR ( yr) or 95% confidence upper limit, based on WISE 12 m luminosity.||RedshiftddSDSS DR9 redshift.||NUV||Notes|
|01||0065||2MASX J103015760106068||13.9||81.3||11.25||1.54||0.28228||21.02eeGALEX NUV-band photometry measured in aperture. Source not in GASC or GMSC.||16.92||2.54||bar|
|04||0170||2MASX J101007073253295||11.6||87.1||11.27||1.40||0.28990||20.14||17.10||2.68||BCG, bar|
|05||0217||2MASX J132757563345291||11.2||68.8||11.05||1.81||0.24892||19.44||16.72||2.02||starburst, bar|
|07||0265||SDSS J115052.98460448.1||10.8||88.1||10.94||0.28946||21.51eeGALEX NUV-band photometry measured in aperture. Source not in GASC or GMSC.||17.19||3.25||faint spiral|
|08||0290||2MASX J123430995156295||10.6||62.4||11.13||1.71||0.29592||19.57||17.25||1.67||Sy1, asymm.|
|09||0299||2MASX J090944802226078||10.5||83.1||11.26||0.28539||21.40||17.25||3.73||BCG, shells?|
|10||0302||2MASX J15430777+1937522||10.5||65.5||11.37||2.45ffThis SFR may be overestimated by a large factor because of the QSO nucleus.||0.22941||17.07||0.40||QSO, tidal arm|
|12||0345||2MASX J092608052405242||10.3||81.2||11.27||1.38||0.22239||19.61||16.57||3.38||BCG, face-on|
|16||0543||2MASX J094700102540462||9.6||99.3||11.07||1.13||0.10904||17.74||14.83||2.57||bar, Sy1?|
|17||0586||2MASX J115356214923562||9.5||90.2||11.11||1.43||0.16673||19.92||15.90||2.64||BCG, Sy2|
|22||0789||2MASX J085421690449308||9.0||86.0||10.96||1.30||0.15679||18.68||15.83||2.49||2 spirals, bar|
|34||1250||2MASX J123215151021195||8.3||71.4||10.95||1.06||0.16588||19.69||16.04||2.76||2 bulges?|
|37||1304||2MASX J160140612718161||8.3||82.3||11.03||1.17||0.16440||17.60||16.06||1.60||BCG, 2 spirals|
|40||1352||SDSS J101603.97303747.9||8.2||68.8||10.73||ggWISE data compromised by nearby IR-bright star.||0.25191||21.16||17.13||2.94|
|46||1501||2MASX J093347772114362||8.1||63.6||11.00||1.69||0.17219||17.84||16.17||1.60||QSO, 2 nuclei|
|50||1559||CGCG 122-067||8.0||81.4||11.13||1.00||0.08902||18.27||14.56||3.13||BCG, 2 bulges|
|51||1606||SDSS J121644.34122450.5||8.0||77.9||1.13||0.25694||20.12||17.22||1.76||bar, Sy1|
|53||1611||2MASX J003807810109365||8.0||83.9||11.31||0.91||0.20828||21.33||16.65||4.38||E with shells?|
|SS||NED Name||ScaleaaScale [kpc/].||bbBulge fraction in SDSS DR7 -band image.||ccBulge eccentricity.||ddDisk exponential scale length.||eeDisk inclination ().||PAffDisk PA ().||ggSmoothness in r band.|
|SS||NED Name||Redshift||N1aaNumber of galaxies within 1 Mpc and 500 km s.||N10bbNumber of galaxies within 10 Mpc and 5000 km s.||Cluster Name||Type||Redshift||ztypeccRedshift type, from NED. EST–estimated, PHOT–photometric, and SPEC–spectroscopic.||Sep()|
|02||2MASX J10405643-0103584||0.250303||1||8||SDSS CE J160.241898-01.069106||GClstr||0.254019||EST||0.013|
|04||2MASX J10100707+3253295||0.289913||2||17||GMBCG J152.52936+32.89139||GClstr||0.319000||PHOT||0.001|
|09||2MASX J09094480+2226078||0.285386||1||9||GMBCG J137.43670+22.43538||GClstr||0.303000||PHOT||0.000|
|12||2MASX J09260805+2405242||0.222451||1||22||WHL J092608.1+240524||GClstr||0.178000||PHOT||0.000|
|13||2MASX J17340613+6029190||0.275807||1||2||SDSSCGB 59704||GGroup||0.276000ddThe association of the super spiral galaxy with the compact group is based only on proximity on the sky. The group redshift in NED for SDSSCGB 59704 appears to be based only on the redshift of the super spiral. None of the other galaxies in SDSSCGB 16827 have measured redshifts.||SPEC||0.450|
|17||2MASX J11535621+4923562||0.166892||3||69||OGC 0586 CLUSTER||GClstr||0.166187||SPEC||0.000|
|35||2MASX J12005393+4800076||0.278617||1||13||GMBCG J180.22479+48.00211||GClstr||0.252000||PHOT||0.001|
|37||2MASX J16014061+2718161||0.164554||3||163||GMBCG J240.41924+27.30444||GClstr||0.193000||PHOT||0.000|
|43||2MASX J13475962+3227100||0.223113||1||13||SDSSCGB 16827||GGroup||ddThe association of the super spiral galaxy with the compact group is based only on proximity on the sky. The group redshift in NED for SDSSCGB 59704 appears to be based only on the redshift of the super spiral. None of the other galaxies in SDSSCGB 16827 have measured redshifts.||0.748|
|50||CGCG 122-067||0.089008||5||302||MSPM 05544||GClstr||0.089190||SPEC||0.001|
Appendix A Custom Photometry and Validation of and SFR
In order to validate our stellar mass and SFR estimates, which are based on , , , and WISE 12 m photometry, we make a more detailed analysis of two representative examples from our super spiral sample. We remeasure their photometry in matched apertures, rather than relying on catalog photometry. Then we fit their SEDs to make full use of the available multi-band photometry to estimate more accurate and SFR. We chose SDSS J094700.08+254045.7 (SS 16) for this analysis because it is one of the brightest super spirals in our sample, with good photometry in many bands, and typical colors. The SDSS spectrum of its bulge is also typical of most super spirals, being dominated by an old stellar population (Figure 9). We also make a detailed study of 2MASX J13275756+3345291 (SS 05), which is the most luminous (non-QSO) mid-IR source in our sample and has an SDSS nuclear spectrum with strong young stellar component and high-equivalent width H emission (Figure 10), characteristic of starburst activity.
We remeasured GALEX (FUV, NUV), SDSS (, , , , ), 2MASS (, , ) and WISE band 1-4 photometry for SS 16 using the SAOImager ds9 (Joye & Mandel, 2003) on images retrieved from MAST, SDSS, and IRSA (Figure 11). Aperture and color corrections were applied as necessary and the GALEX and SDSS photometry was corrected for foreground extinction due to the Milky Way dust (Wyder et al., 2005; Stoughton et al., 2002). The Galactic extinction is a modest mag (NED). We used an elliptical aperture with semimajor and semiminor axes of and , respectively, in order to capture the full flux of the spiral disk in all bands. This corresponds to major and minor diameters of 125 kpc and 102 kpc. We also compute IRAS upper limits based on the rms uncertainty measured by IRSA’s Scan Processing and Integration tool (SCANPI) to constrain the FIR luminosity.
We present the SED of SS 16 in Figure 12. The galaxy is detected in all GALEX, SDSS, 2MASS, and WISE bands, but is undetected by IRAS. The UV through near-IR data points reveal a massive old stellar population plus a young stellar population. Mid-IR emission appears to be dominated by PAHs and warm dust from star formation. We fit the SED using magphys template fitting (da Cunha, Charlot, & Elbaz, 2008). This gives a total stellar mass of and SFR of yr. We get a consistent estimate of for the stellar mass from the color and -band luminosity, applying the Bell et al. (2003) prescription for color-dependent mass-to-light ratio (Table 1). The WISE band 3 luminosity gives a consistent SFR of yr, using the prescription of Chang et al. (2015). Lacking FIR detections, we do not have a good handle on the total dust mass, however, the SED fit formally yields a dust mass of , based on the PAH emission and FIR upper limits. This corresponds to roughly of gas, assuming a standard gas/dust ratio of 100.
We remeasured GALEX (FUV, NUV), SDSS (, , , , ), 2MASS (, , ) and WISE band 1-4 photometry for SS 05 (Figure 13), using a similar procedure. We also retrieved Spitzer IRAC and MIPS, and Herschel PACS and SPIRE images from the respective IRSA and ESA archives to measure the IR fluxes. We used a circular aperture with (156 kpc) radius for most bands. However, at SPIRE wavelengths, we used the larger point source apertures of , , and , in order to contain the broader point-spread function. The Galactic extinction is only mag (NED).
We present the SED of SS 05 in Figure 14. The galaxy is detected in all measured bands except the SPIRE 500 m band. In contrast to SS 16, there is a stronger component of emission from young stars, and much more luminous IR emission from star formation activity. We fit the SED using magphys, yielding a total stellar mass of and SFR of yr. The stellar mass is consistent with the value of that we obtain from the color and -band luminosity (Table 1). The WISE  luminosity gives a somewhat larger SFR of yr, using the conversion factor of Chang et al. (2015). The Herschel FIR measurements yield a secure estimate of total dust mass from the SED fit of , corresponding to of gas, assuming a standard gas/dust ratio of 100.