Significance to Astronomy
Objects that are located in the universe at low surface brightness levels are naturally underrepresented in any flux limited survey at any wavelength. This naturally leads to a biased view in which models of galaxy evolution are constrained by what is available in the easily detected Universe. As a quick reminder of the severity of this bias, extant galaxy catalogs contain only about 1/12 of the required baryons to satisfy the results from BBN nucleosynthesis and/or the concordance model (see Bothun 2003). Hence the question Where are the Baryons? is not a rhetorical one but rather a serious astrophysical one. While some of the “missing” baryons are undoubtedly present in the warm IGM, current measured levels (e.g. the X-ray forest; the UV Background) are insufficient to account for most of them. In a similar vein, this proposal hopes to motivate the review panel that another seemingly innocuous question Do the disks of galaxies ever end? is an important astrophysical question. For sometime, conventional wisdom held that galactic disks break their exponential nature at 3-5 scale lengths (see Pohlen etal 2002; Kregel etal 2002; Erwin etal 2005) and that this “break” radius may hold astrophysical significance. Note that for a typical disk with central surface blue brightness of B = 21-22 mag/” , 5 scale lengths (1 scale length = 1.09 magnitudes drop in SB) represents a disk environment seldom probed by most observations due to surface brightness limitations. So, we don’t really know what is going on in most disk galaxies at these large radii. Strong observational confirmation that disks may not really end comes from the GALEX Atlas of Nearby Galaxies which shows that 10-20% of all imaged spirals have greatly extended UV emission (the most spectacular case being M83 which shows star formation at radii about 50 kpc – see Gil de Paz etal 2007). Thus, it would appear that the outer environments of disk galaxies have become a new, astrophysically interesting target to explore and GALEX has proven itself to be a capable tool for such exploration.
Background and Motivation
The emergence of extended UV disks in spiral galaxies (often those with known extended H I distributions) has been regarded as one of the more interesting and surprising results from the
ALEX mission to date. The above image shows the case of NGC 4569 (see Gil De Paz etal 2007) in which the (green) ellipse denotes the nominal optical extent of the disk. The excess NUV emission is quite obvious. For the purposes of this proposal, the discovery of XUV disks is taken as confirmation that LSB features are most easily detected in the NUV window due to the very low sky background as originally suggested by O’Connell (1987) and shows the ability of GALEX to detect diffuse stellar populations in the very lowest surface brightness regions of the galaxy disk environment. While the existence of young stars obviously helps in this detection (and of course can also be detected via ionized hydrogen) most any stellar population of age a few Gyr will effectively radiate at 2000 angstroms. We therefore wish to exploit the LSB detection capability of GALEX to image a small sample of gas rich, large scale length, relatively isolated, disk galaxies to confirm in these systems, what has been observed for NGC 300 – namely the detection of disk light out to 10 optical scale lengths. For the sample in question, this would result in the establishment of disks of diameter 100-200 kpc, and such large structures become a key challenge to understand in any context of galaxy formation due to the very fact that stellar orbits in the outer radius would be at most, 3 dynamical timescales old.
As emphasized in Ferguson (2007), the distant outskirts of galaxies have been largely unexplored but may hold important clues to the assembly process involved in galaxy construction. GALEX provides another good tool in which to make these relevant measurements.
This proposal is then motivated by two basic considerations. 1) The experience of a seasoned observer (e.g. the aged PI) whom has long known that, beyond the nominal optical radius of galaxies, there is still a lot of stuff (distant H II regions, diffuse H I, carbon stars, stellar clusters, supernova Ia occurrences) that is part of the galaxy. Indeed, all one has to do is observe M31, M33 and M101 to become immediately convinced that both young and old stellar populations can exist at very large radii (see review of Ferguson 2007) 2) The extremely intriguing and unprecedented detection of 10 (uninterrupted) optical scale lengths (depending on how the background is explicitly subtracted) of disk light in the galaxy NGC 300 by Bland-Hawthorne etal 2005. For reference, 10 scale lengths correspond to a surface mass density which is less than 1 solar mass per square pc. The detection of such extended disk light is a dynamical paradox and raises the following issues:
How could stars every form in such a low surface mass density environment? Even plowing a spiral density wave through that medium would not work as the Toomre Q parameter would remain high (due to low mass density).
Did the stars gravitationally scatter from the inner disk to occupy these regions? (if so, how is the coherency of the disk maintained?)
Is this extended disk the signature of accumulated accretion events due to infalling intergalactic gas into the potential of NGC 300 (if so, again, how did the stars form; why is the exponential disk so coherent?)
Why hasn’t this greatly extended disk been tidally sheared away? NGC 300 is not totally isolated and does feel an overall tidal field from the Sculptor group and other mass concentrations?
The basic goal of this program is to confirm that this phenomenon is not restricted to only NGC 300 but may be a basic feature of certain types of disk systems. Our candidate objects consist of large, relatively isolated disk galaxies with ground based U-band observations (obtained at the Pine Mountain Observatory, University of Oregon) that already yield 4-5 scale lengths of disk light. If this goal is achieved, then the existence of such extended disks (with associated long dynamical timescales) must be accounted for in any structure formation scenario. For the moment, NGC 300 stands alone as a theorists nightmare – to pressure structure formation theories, similar occurrences of extended disks require detection.
Relevance to NASA goals
The environment we intend to probe is very much related to the issue of galaxy formation or assembly and thus is related to the overall origin of structure in the Universe. In addition, if significant populations of weakly bounds stars can be detected, then we have another laboratory for studying the motions of objects on extremely low acceleration scales. As such, these systems may ultimately, pending the construction of new facilities, prove to be useful probes of the behavior of gravity on weak acceleration scales and these environments may be useful in helping to prove/disprove MOND or other alternative theories of the behavior of gravity.
Bland-Hawthorne etal 2005 ApJ 629,239
Bothun 2003 The IGM/Galaxy Connection: The Distribution of Baryons at z=0, ASSL Conference Proceedings Vol. 281. Edited by Jessica L. Rosenberg and Mary E. Putman. ISBN: 1-4020-1289-6, Kluwer Academic Publishers, Dordrecht, 2003, p.11
Erwin etal 2005 ApJ 626 L81
Ferguson, A. 2007 astro-ph/0702224
Gil de Paz etal 2007 ApJ 661 115
Kregel etal 2002 MNRAS 343,646
O’Connell 1987 AJ 94, 876
Pohlen et al 2002 A&A 392 807
Description of the Observations
We have constructed our target list of 6 galaxies, from the ground-based U-band survey. As luck would have it, our two highest priority targets (IC 983 and UGC 2885) are deemed unsafe to observe. Many of the lower priority targets were also deemed unsafe, leaving us with a sample of 6 good candidates for observations for which U-band data exists. In general, the target galaxies exhibit strong spiral structure, are relatively isolated, have measured scale lengths between 5 and 10 kpc, and have angular sizes larger than 5 arcminutes. (meaning radial velocities less than 5000 km/s) Our primary goal is to obtain sufficiently deep GALEX observations to detect the extended stellar disk to verify the existence of stellar populations in galactic disks of size 100-200 kpc and our U-band data is suggestive that some or all of these galaxies will have extended NUV detectable disks. Thus we have high probability of extending the phenomena observed in NGC 300 (and to a lesser extent in NGC 4625) to a class of intrinsically big disk galaxies and showing that on the 50-200 kpc scale, galaxy disks exist! The establishment of such extended disks would clearly show that star formation can and does occur in very low surface density environments and this presents a profound theoretical challenge which not only has obvious implications in galaxy formation and evolution, but also helps us understand where some of the missing baryons are located. Furthermore, these structures are clearly inconsistent with the current theoretical prejudice that galaxy disks are constructed from the inside out as there is likely insufficient time to build structures this large. This potential pool of weakly bound stars may also represent the parent environment of intergalactic stars or clusters ,dwarf galaxies, and tidal streams as hierarchical clustering and associated tidal effects gradually remove these stars from this reservoir (making the only surviving extended disks likely to be located in isolated systems).
While the primary goal of the observations lies in the detection, through surface photometry, of very extended disks, we will also obtain a secondary goal of measuring the total NUV flux and NUV morphology of very large disk galaxies. Such measurements would be relevant to measuring the potential contribution of very large star forming spiral galaxies to the metagalactic UV flux. In addition, our targets (except for the giant nearby LSB spiral UGC 9024) are galaxies that are easily detected at high redshift; knowledge of their NUV morphology might therefore be very helpful in comparing to large disk galaxies selected at high redshift through the same rest frame continuum light. Their respective morphologies when convolved to the same physical resolution scale, could then be directly compared. To serve as a rough indicator of the bulk nature of our target galaxies, the galaxies in this sample have total disk areas, luminosities, or H-alpha luminosities an order of magnitude higher than M101.
Feasibility and the Need for GALEX
GALEX has proven its unique capability to detect extended LSB debris around the disks of normal galaxies. Its large field, good sensitivity and the dark NUV background all combine to facilitate this detection. Exposure time calculators are never sufficient to return expected exposure times as a function of the surface brightness of the object. Moreover, the ability to reliably detect low surface brightness features depends more on how well the data can be flat fielded and how high the background is than the actual counts per pixel.. To serve as an exposure guide, we therefore simply take the cases of M83 and NGC 4625 as template examples of the exposure depth required to reach reasonably faint surface brightness levels in order to detect extended populations. These cases show that exposure times of 1 – 3.5 ksec are sufficient to easily detect extended stellar populations. As M83 is an extreme case, we regard NGC 4625 as more typical and correspondingly adopt 5 ksec for all exposure times to reach slightly deeper isophotal levels than were obtained in the NGC 4625 observations. We note that this exposure time is likely not sufficiently deep to get all the way down to 10 scale lengths (as this will be mostly sensitive to the background behavior and not the actual counts per pixel) but the exact isophotal depth is unimportant as we mainly seek to detect a significant population of NUV emitting stars (either from current or past star formation) at locations well beyond the nominal optical radius of the galaxy.