|Background and Motivation:
Multicolor imaging of large angular size galaxies potentially provides unique information on the nature of the stellar population as function of position. In turn, this data can be used to investigate stellar population mean ages or metallicity as a function of position, trace out regions of current and past star formation, detect radial gradients, and build an extinction map, if the color baseline is large enough. In addition, using various image processing techniques, small scale surface brightness fluctuations can be used to plausibly detect stellar cluster candidates and the color distribution of those cluster candidates can be directly compared to the underlying stellar populations. The technique of grid photometry, perhaps first utilized by Bothun 1986, is a simple but relatively powerful way to analyze multicolor pixel data. In this methodology all different filter images are registered to the same coordinate system and then a sliding square aperture is applied to each filter. Within each box, median filtering or sigma clipping effectively rejects contaminating stars, cosmic rays or other large fluctuations against the background (including stellar clusters in this first pass). The result is a mean surface brightness and mean color of each box. A balance needs to be struck with respect to the size of the box. Too large of boxes smooth over too much of the intrinsic variation in the stellar population. Too small of boxes would have too large of random errors on the mean color for the data to be useful. The most ideal situation occurs where a very large number of pixels contain the whole galaxy image. This would clearly be the case for the ACS M51 data.
To better illustrate the kinds of maps that can be produced via this method, we turn to various published examples. In all cases, what is shown in the figures at the end of this document is the XY pixel location for each square aperture that falls within a given color range. Shown are published data for NGC 4449 (Bothun 1986), The LMC (Bothun and Thompson 1988), and IC 342 (White and Bothun 2003).
For the case of NGC 4449 we use a color baseline of B-R and present a blue color cut and a red color cut. A clear difference in morphology is seen and, not surprisingly, the blue cut corresponds to the location of current and recent star formation. For this data the scale is 1 pixel = 12 pc (for a nominal distance to NGC 4449) and each square aperture was 5x5 pixels. Thus our stellar population resolution scale is 60 pc. The image immediately below the continuum B-R color cuts represent grid photometry that is performed on the residual image that occurs when the original image is smoothed and subtracted from itself. This is a simple but useful technique to detect any embedded stellar-like images that may be stellar clusters. As can be seen, the blue cut clusters correspond to the star forming regions and while the redder clusters are more spherically distributed and some are globular cluster candidates.
A more innovate example of this approach occurs in the case of the LMC where a Parking Lot CCD Camera was used to image the entire LMC in a single CCD frame. Again, the procedure of grid photometry clearly reveals a strong dependence of galaxy morphology on color (which may actually be important in interpreting the images of high redshift galaxies). Again the color cut is B-R as measured in 5x5 pixel boxes. The blue cut shows an extended distribution with the highest density of points occurring in the Constellation X region while the red cut clearly reveals the LMC bar. This data was used by Bothun and Thompson (1988) to asses the mean age of the stellar bar and to identify significant differences in mean age of the stellar population as function of position. That analysis suggests that current star formation is radiating outward from the main LMC body.
As a final example, we turn to an object most similar to M51, namely IC 342. IC 342 is a challenging galaxy due to its large angular extent and low galactic latitude. In the two color cuts shown, this time using the color baseline of B-I, the bluest cut shows the true spiral arm structure while the redder one shows the inter arm structure. White and Bothun (2003) were able to use this data to make an extinction map of IC 342 and to test how much variable reddening there could be across the face of this galaxy.
Proposed Application to M51:
The new ACS M51 Mosaic will be the first of its kind produced with Space Based Instrumentation. Because M51 is a spiral galaxy it will contain many different kinds of stellar populations which will reveal themselves as differences in mean color. The exquisite spatial scale of M51 will allow for the method of grid photometry to explore mean colors on a spatial scale of approximately 10-20 pc. Furthermore, M51 is clearly an interacting galaxy and thus we will be studying the stellar populations in a galaxy that is being slightly disrupted by a satellite companion. This potentially provides a very powerful and constraining test of current dynamical models. What we measure is the current distribution of grid colors as a function of position (e.g. the stellar populations) after the encounter has occurred. In principle, one can put in model distributions of stellar populations and dust in the pre-encounter, presumably quiescent, late type spiral. Those model distributions are then perturbed (by the N-body code representing an encounter with a satellite galaxy) and the spatial position of each stellar population and/or dust distribution can be followed. Those distributions can then be directly compared to the data on spatial scales unprecedented in this level of modeling. Suddenly we have the potential to go well beyond modeling the gross properties of interacting galaxies (e.g. bridges and tails) to modeling the response of a galactic disk to an encounter on very local scales. This is an exciting prospect that may take our understanding of satellite encounters well beyond the velocity field modeling done by Mihos and Bothun (1997).
As shown in the White and Bothun (2003) analysis of IC 342, if you have a sufficiently big color baseline (e.g. U-I), you can use that information in comparison with color indices over shorter color baselines to approximately determine the locations of high and low extinctions. Thus we also have the opportunity to produce a high spatial resolution extinction map for NGC 5194/5. Using the ACS mosaic data on M51 would allow for an unprecedented comparison between observations and interaction models on spatial scales perhaps 2 orders of magnitude smaller than has been possible previously. While there is clearly likely to be degeneracy in this comparison, we are interested in mapping out the possible pre-encounter 2D density distributions of the ISM (e.g. dust) and stellar populations that when convolved with interacting galaxy models are consistent with the present day observations.
The overall plan of attack would be to apply the method of grid photometry to the HST M51 data in the same way it has been successfully applied to targets on the ground. Such an analysis should lead to the following:
A determination of the mean color of M51’s stellar populations on spatial scales of 10-20 parsecs. This should help to define locations of current and past star formation and to define the extent of the underlying old age disk in a similar way that such identification has been made in the LMC.
A thorough map of embedded point sources which are cluster candidates can easily be made given the high angular resolution of the data. In addition, individual clusters can be identified perhaps directly from the data by correlating color and surface brightness fluctuation. In general, a cluster represents a high surface brightness region embedded against a smoother background of a different color. Depending on the yield, it may be possible to then study the spatial location of cluster formation in response to the interaction dynamics.
Grid photometry as applied to the HCG 90 by White etal 2003 reveals that this method is also quite good for identifying and characterizing any diffuse or intracluster light population. Thus we expect to be able to properly characterize the stellar population and possible stellar population differences within the diffuse tidal bridge between NGC 5194/5.
Using a combination of short color baselines (e.g. something equivalent to U-B) and long color baselines (e.g. V-I) it is possible to being to break the degeneracy between age and metallicity (see Bothun etal 1984; Bothun and Gregg 1990). While the use of near-IR indices in this technique is better, V-I is an adequate substitute if significant differences in giant branch metallicity exists. In this case, the data may reveal metallicty differences in the older disk population within M51 and/or the extent to which interaction dynamics may mix these populations thus obscuring the signature.
The products of our analysis will be an “Atlas” of grid photometry images, like those depicted earlier in this proposal. While this may be less spectacular than a digital coffee table book showing the many colors of M51, grid photometry is the best way of visualizing the structure of a galaxy as a function of color. Given the potential quality of the data, we would expect these maps to be far more detailed and encompass more filter ratios than those done to date. From those details we hope to reconstruct the overall star formation history of M51 as well as its small scale response to interaction dynamics. We have a proven track record in these research areas and believe that the application of grid photometry to the ACS Mosaic of M51 is both straightforward and quite scientifically useful and furthermore can be done in a relatively timely manner.
Figure 1: NGC 4449 – left hand side is the blue B-R color cut and right hand side is the red cut. The top row represents is for the distributed star light while the bottom row is for the embedded images that are presumably stellar clusters.
Figure 2: Top row is B-R blue and Red Color Cuts for the LMC. The X and Y axis have a total dimension of 6 angular degrees on the sky. The bottom row represents two color cuts of IC 342 using B-I. The left hand cut (the bluest one) most effectively reveals the spiral structure while the more intermediate color cut shows the inter- arm region.
Bothun etal 1984 AJ 89 1300
Bothun 1986 AJ 91 507
Bothun and Thompson 1988 AJ 96 877
Bothun and Gregg 1990 ApJ 350 73
Mihos and Bothun 1997 ApJ 481 741
White and Bothun 2003 PASP 115 1135
White etal 2003 ApJ 585 739