A Relationship Between Stellar Surface Density and Star Formation Rate

Our search for new fundamental relationships in galactic disks began with a major imaging study of 34 nearby ( km s) southern spiral galaxies in both the V and I bands (the I band being more sensitive to the older, low-mass stellar population) as well as in the H line (which highlights the HII regions ionised by newly-formed massive stars, and thus can be used as a measure of the current star formation rate; Kennicutt 1983). Surface photometry using the GASP package was then carried out to determine the disk orientation parameters, and thus the radial variation of deprojected surface brightness (assuming the disk to be optically thin throughout) in each band. We then asked the question: Is there any correlation between the azimuthally-averaged stellar surface brightness at a given radius in a galaxy (as marked in the upper panel of Figure 1), and the azimuthally-averaged star formation rate (lower panel) at that same radius? In this way, we are effectively normalising by the galaxy's scale length, without the need for complex bulge-to-disk decompositions.

 
Figure 1:  Images of the nearby spiral galaxy NGC 2997 in the V-band (top) and H line (bottom) from the MSSSO 40'' telescope. The concentric circles in each image mark annular zones of equal radii, within which the mean surface brightnesses are to be compared.

As Figure 6 of Paper I shows, there is indeed a very similar trend, amongst all the galaxies in our sample, in the way that the H surface brightness decreases with decreasing I-band surface brightness, with the main departures from perfectly uniform behaviour being that some loci lie slightly above (fainter in H at all points) or slightly below (brighter in H). In order to investigate whether there may be a ``second parameter'' at work, we have plotted the size of this offset against various global galactic parameters. A good linear correlation was found with the log of the total HI gas content per unit area of the optical disk (Figure 2), in the sense that it is the more gas-rich spirals that appear brighter in H for a given I-band surface brightness than the mean. This is perhaps to be expected, on the basis that galaxies with low gas surface densities will not be capable of forming stars at the same level as those with higher gas densities. Figure 3 shows just how well the loci agree after the offset correction is applied.

 
Figure 2:  Plot of the total HI gas content normalised by the area of the optical disk, on a logarithmic scale, against the offset (in magnitudes per square arcsecond) needing to be applied to the H surface brightnesses, to bring them into agreement with the mean locus of H vs. I-band surface brightness. The dashed line is a linear least-squares fit to the points.

 
Figure:  Individual loci of H surface brightness vs. I-band surface brightness for the 34 spiral galaxies in our sample, after the offsets from Figure 2 are applied, compared with the mean trend (dashed line).

© Copyright Astronomical Society of Australia 1997