Although color-magnitude diagrams have now been obtained for a few very close galaxies - most of them dwarf -, allowing to study directly their stellar populations (e.g. Schulte-Ladbeck & Hopp 1998; Aparicio 1998), the synthesis of the spectral energy distribution (SED) of nearby galaxies remains the most efficient and systematic method to trace their star formation history up to .
Early optical studies have shown a clear correlation between the SED - and hence the star formation history - and the galaxy Hubble type. However, the age-star formation timescale and age-metallicity degeneracies make it highly desirable to extend the wavelength range to the ultraviolet (UV) and the near-infrared. Whereas the ultraviolet is related to the young stellar populations in currently star-forming galaxies, the near-infrared is dominated by old giant stars and is a measure of the star formation rate integrated from the beginning, i.e. of the stellar mass of the galaxy. When compared to shorter wavelengths, the NIR may moreover put constraints on the stellar metallicity, owing to the high sensitivity of the effective temperature of red giants to the abundance of heavy elements. In combination with optical data, it finally provides clues on the amount and the distribution of dust because of the different extinction at both wavelengths.
Observations in the UV and the NIR are however hampered by the atmosphere. Though interesting to study the detailed star formation history of a specific galaxy, the spectra obtained from Space are too scarce to afford a statistical analysis from which the general trends as a function of type or mass, or the effect of dust, may be derived. In spite of the loss of spectral resolution, broad-band colors in the J, H and K atmospheric windows are more promising. As the spectra, they are usually obtained in small apertures. Because of the color gradients, they are not representative of the whole galaxy, especially for disk galaxies, and have to be extrapolated to be compared to the optical data. Such extrapolation is however much easier for colors than for spectra and can be applied to hundreds of objects.
In the following, we analyze a catalog of infrared aperture magnitudes (Sect. 2) and combine it with optical catalogs to derive NIR growth curves of the magnitude as a function of the aperture (Sect. 3). We then compute total and effective magnitudes and colors (Sect. 4). The NIR and optical-to-NIR colors are analyzed statistically as a function of type, luminosity and inclination in Sect. 5. We finally discuss our results in Sect. 6.
© European Southern Observatory (ESO) 1999
Online publication: November 16, 1999