Date: Fri, 1 Aug 2008 12:10:37 -0500 (CDT) From: apjse@sheffield.ac.uk To: timw@astro.as.utexas.edu Cc: ApJ-MS75208@mss.uchicago.edu, apjse@sheffield.ac.uk Subject: Your ApJ Submission MS# 75208 Dr. Tim Weinzirl The University of Texas Department of Astronomy 1 University Station, C1400 Austin, Texas 78712-0259 USA Dear Dr. Weinzirl: Enclosed please find the referee's report on your submission to the ApJ entitled "Bulge n and B/T in High Mass Galaxies: Constraints on the Origin of Bulges in Hierarchical Models" by Tim Weinzirl, Shardha Jogee, Sadegh Khochfar, Andreas Burkert, and John Kormendy ( MS# 75208). When you resubmit the manuscript, please include a detailed cover letter containing the (mandatory) listing of the changes you've made to the text and your responses to the report. Processing of your revised manuscript will be expedited if you make your revisions to the manuscript latex file available for downloading from the ApJ Web Peer Review System (http://mss.uchicago.edu/ApJ/). This version includes your previous submission plus any modifications to latex commands necessary for smooth processing by the ApJ electronic system. The associated PDF file is the version seen by the referee. The Astrophysical Journal has adopted a new policy that manuscript files become inactive, and are considered to have been withdrawn, six months after the most recent referee's report goes to the authors, provided a revised version has not been received by that time. If you have any questions, feel free to contact me. Best regards, Richard de Grijs, Scientific Editor The Astrophysical Journal Phone: +44 114 222 4524 Fax: +44 114 222 3555 apjse@sheffield.ac.uk ************************************************* This paper presents a morphological decomposition of H-band images of 146 nearby galaxies of moderate to high mass. The driving motivation is the problem of bulgeless galaxies: through their study, the authors expect to contribute statistics on B/T ratios along the Hubble sequence, with the aim of determining how serious the low-B/T problem is for current galaxy formation models. The decomposition improves on several previous studies in that a bar component is allowed in the fit, in addition to the standard Sersic bulge and exponential disk. The decomposition is used to highlight a number of results, including (1) a prevalence of low-B/D galaxies across Hubble types S0/a to Sc, (2) a prevalence of low-n bulges, (3) that bulges contain only 20% of the galaxy mass in those galaxies, and (4) that the statistics of low-B/D bulges cannot be explained in the context of a particular SAM model of disk-bulge galaxy formation, in which bulges dominantly grow through major mergers. Minor mergers may be able to do the job. The paper is very well written and clear. Most issues related to 2D morphological decomposition, a tricky business indeed, are dealt with in detail. Indeed, I find the morphological decomposition the most valuable part. However, two aspects would need more attention in my view. (a) The issue of parameter errors, and error coupling, are mentioned, but too little is done about them. The authors 'will explore these issues in a future paper'. This is probably acceptable in a Letter, but not in a long paper. A few artificially simulated images are fitted to test the code, which is a sound way to proceed. However, the number of fitted models, six bulge-bar-disk and four bulge-disk, is far too small. Too little is said about these models to allow the reader to assess the relevance of the exercise. What parameters were explored in these limited experiments? Decoupling bulge from disk is much easier in low-n bulges than high-n bulges; identifying and fitting a bar is easier for some position angles and quite difficult in others; the ability to recover bulge parameters depends on the seeing. Etc. While I accept that publication of the paper does not need to wait for a full Monte-Carlo solution to the error determination, which would be time consuming given the multi-dimensional parameter space of the fits, more needs to be done than what's currently in Sect 4.2. The authors might at least identify those subsets of parameter space most relevant to the science of the paper, and devise tests to assess parameter coupling in those areas. Areas of interest might be coupling of bulge and bar parameters, and bulge vs disk parameters, in the cases of low-n and high-n. When this section is rewritten, I would encourage the authors to give more details on the models fitted and on the results. If the section becomes too long it may be pl! aced as an appendix. (b) The authors rightly note (Sect 3.2.) that central point sources lead to overestimating the Sersic index, a problem highlighted by Balcells et al 2003, and others. The authors then proceed to correcting for point sources in objects with n>5. Why only n>5? Fitting an n=1.5 bulge which harbors a central cluster will yield a wrong n of perhaps 3 or 3.5. Central positive deviations from the Sersic model are easily identified from the shape of the residual profile from a Sersic fit: a central peak surrounded by a negative valley (at any Sersic index). With the procedure employed in the paper, many of their Sersic indices are biased toward high values. Published statistics of central sources are in the range from 50% to 90% depending on the sample and the resolution (Ravindranath et al 2001; Boker et al 2002; for a sample with very similar selection to the present sample, Balcells et al 2007 report a frequency of 60%, or 90% when including central sources resolved by HST). T! he authors have identified point sources, and corrected for them, in 33% of the sample; it is plausible that an additional 30% to 60% of the sample need to be corrected as well. A simple way to address this issue will be to (i) take the best fit from Sect 3.3, for the subset of the S1 sample without nuclear sources in their current fit, (ii) repeat that fit adding a central point source, and (iii) generate a criterion for choosing among the solutions with and without central source. After these two points are addressed, I feel the morphological decomposition will be very valuable research tool, and I would encourage the authors to consider providing the fitted parameters in tabular form. The second part of the paper (Sect 5.8) is a comparison with a LCDM-based semi-analytical model of galaxy formation. This is essentially the model first proposed by Kauffmann et al 2006. In my own view the model is outdated. It is clear today that the model is too simplistic: it makes low-B/T bulges the oldest ones, while plenty of empirical evidence exists that low-B/T galaxies show younger populations (bulge and disk; e.g., Ganda et al 2007) --while more massive spheroids, which live in high-B/T galaxies, tend to be older. But the truth is, this model is still prominently cited in reviews, talks at conferences, and introductory sections of papers on galaxy bulges. It is therefore useful that the model is put to test in this paper, and shown to be wrong. In summary, I recommend publication of the paper, once points (a, b) above, and a number of other points listed below are satisfactorily addressed by the authors in a revised version. The recommendation is made on the basis that (1) the problem addressed is important and timely, (2) the data analysis is generally well done, (3) the results are a useful contribution to the field, and (4) the paper is clear and well written. Detailed notes -------------- 1- Abstract, and throughout the paper: 'moderately inclined' is misleading as the sample include all inclinations except near-edge ons. The authors should use a different expression. 2- Sect 2.1. survey and data quality parameters should be described: telescope(s), pixel size, exposure times, image depths, seeing, photometric calibration. 3- Table 1: the source(s) of the galaxy distances should be given, together with statements on their assumed uncertainties. 4- Sect 2.2. The masses from Bell et al 2003 are ok, but, why not use NIR imaging. If their H-band images are not calibrated (which Eskridge et al. 2002 says they are), the authors can use 2MASS images for the calibration. K-band magnitudes from 2MASS must be quite accurate for massive, nearby galaxies. There are plenty of reliable sources of M/L_K -- including Bell et al 2003. Providing galaxy masses derived from the two methods, and comparing the results, would enhance the quality of the paper. 5- Sect 2.2. Errors in the derived masses should be provided, including the contribution from the error in the galaxy distance. 6- Sect 5.2. Still on masses: a global M/L is derived from the global B-V color. But all galaxies show color gradients. For the mass ranges studied here, color gradients are mostly negative outward, hence in most cases M/L is higher in the central parts, and only in a minority of cases M/L is lower in the central parts. As a result, there is a danger that the current derivation of mass in bulges, disks and bars penalizes bulges. The discussion of these effects in Sect 5.2, paragraphs 2 and 3, is too neutral -- by first mentioning the possibility that bulges are younger than their surrounding disks, the authors give the idea that such circumstance is more common than the opposite one. 7- Sect 5.3. The discussion around Figure 15 is weak, as there is no way for the author to assess from the images whether a bulge is 'conspicuous'. The visual assessment will depend entirely on the stretch with which the images are displayed. In the particular stretch used, the differences between the size of the black, saturated central region which, I believe, the authors call the bulge, is small when comparing, e.g., NGC 1371 and NGC 1808 with NGC 3810 and NGC 4254. This discussion might become more useful if focussed on showing/reminding that Hubble types were not originally assigned as a measure of B/D, but as a combination of three distinct criteria including dust and topology of the spiral arms. The caption to Figure 15 should focus on describing the figure: what band are these images? what depth? what gray contrast ? 8- Sect 5.5. The discussion of bar strength could be improved. While the authors correctly explain that ellipticity is only a partial measure of the bar strength, they then proceed to promoting bar mass to the status of bar strength measure; this sure is partial also. Wouldn't bar strength scale with _both_ bar mass and bar ellipticity (and, also, bar length)?. I suggest the authors rethink and rewrite this section. 9- Sect. 5.6. The unveiled statistical association between bars and n<2 bulges does invite one to propose a cause-effect relationship. But all the authors have is a statistical association. Personally, I think that this kind of statements, which turn a statistical association into a cause-effect relationship, even if just 'suggested', are rather dangerous, as they may propagate as truth what is simply an expectation, wish, or belief based on evidence other than the one presented in the paper. My suggestion would be to rephrase the statement, and certainly remove the italics. 10- Sect 5.7. The description of the physics of minor mergers contains inaccuracies, which for the sake of completeness should be corrected. (1) The discussion misses one key process through which minor mergers contribute to the growth of B/T - inflow of disk stars to the bulge region as a result of the distortions triggered by the tidal field of the merging secondary. This process is modeled in Eliche-Moral et al. 2006, and should be included in the description of relevant physical processes for bulge growth. (2) It also misses the effects of multiple minor mergers, most relevant in the context of this paper (Bournaud, Jog & Combes 2007). (3) While one often sees written that minor mergers excite bars, which in turn drive gas to the center, and papers generally cite Hernquist & Mihos 1995 in support of the statement, the truth is that, what Hernquist & Mihos 1995 claimed, was that the _tidally distorted disk_ creates the dominant torque which sends the gas inward. It is not clear to me that they ever claimed that distortion to be a bar --e.g., the authors might search for the word 'bar' in that paper. 11- Some references are odd, or missing. A list of noted problems follows: * (Sect 5.7) in support of the statement that low-luminosity ellipticals having n<4 , the authors cite Kormendy et al. (2008, in prep), but fail to cite published papers that would be appropriate, going back to Caon et al. 1993 who established this concept, and, I believe, should deserve the central credit in this context. * (Sect 5.7) Quinn et al 1993 is mentioned when talking about bar-driven gas inflow, while that paper does not include gas and does not mention bars. * (Sect 5.7) important references for the physics of minor mergers are missing, e.g., Bournaud, Jog & Combes 2007, Eliche-Moral et al. 2006. * that bulges typically have lower than n~3 sersic indices was first emphasized by Balcells, Graham, Peletier et al 2003, which deserve being cretited for it. * That in early-type disks is low had been found before (Balcells, Graham & Peletier 2007, =0.25 for S0s), * I couldn't find Mihos & Hernquist 1995 on the reference list. * The reference list is not in complete alphabetical order. 12- Figure 14 is not mentioned in the text. 13- Some figure captions contain interpretation of the figure, which does not conform to apj style.