We explore the properties of the submillijansky radio population at 20 cm by applying a newly developed optical color-based method to separate star forming (SF) from AGN galaxies at intermediate redshifts (z<1.3). Although optical rest-frame colors are used, our separation method is shown to be efficient, and not biased against dusty starburst galaxies. This classification method has been calibrated and tested on a local radio selected optical sample. Given accurate multi-band photometry and redshifts, it carries the potential to be generally applicable to any galaxy sample where SF and AGN galaxies are the two dominant populations. In order to quantify the properties of the submillijansky radio population, we have analyzed ~2,400 radio sources, detected at 20 cm in the VLA-COSMOS survey. 90% of these have submillijansky flux densities. We classify the objects into 1) star candidates, 2) quasi stellar objects, 3) AGN, 4) SF, and 5) high redshift (z>1.3) galaxies. We find, for the composition of the submillijansky radio population, that SF galaxies are not the dominant population at submillijansky flux levels, as previously often assumed, but that they make up an approximately constant fraction of 30-40% in the flux density range of ~50 microJy to 0.7 mJy. In summary, based on the entire VLA-COSMOS radio population at 20 cm, we find that the radio population at these flux densities is a mixture of roughly 30-40% of SF and 50-60% of AGN galaxies, with a minor contribution (~10%) of QSOs.

We discuss the panchromatic properties of 99,088 galaxies selected from the SDSS Data Release 1 spectroscopic sample (a flux-limited sample for 1360 deg^2). These galaxies are positionally matched to sources detected by ROSAT, GALEX, 2MASS, IRAS, GB6, FIRST, NVSS and WENSS. We find strong correlations between the detection fraction at other wavelengths and optical properties such as flux, colors, and emission-line strengths. Using GALEX, SDSS, and 2MASS data, we construct the UV-IR broad-band spectral energy distributions for various types of galaxies, and find that they form a nearly one-parameter family. For example, based on SDSS u- and r-band data, supplemented with redshift, the K-band 2MASS magnitudes can be "predicted" with an rms scatter of only 0.2 mag. When a dust content estimate determined from SDSS data by Kauffmann et al. (2003) is also utilized, this scatter decreases to 0.1 mag. We demonstrate that this dust content is indeed higher for galaxies detected by IRAS and that it can be used to "predict" measured IRAS 60 micron flux density within a factor of two using only SDSS data. We also show that the position of a galaxy in the emission-line-based Baldwin-Phillips-Terlevich diagram is correlated with the optical light concentration index and u-r color determined from the SDSS broad-band imaging data, and discuss changes in the morphology of this diagram induced by requiring detections at other wavelengths. We study the IR-radio correlation and find evidence that its slope may be different for AGN and star-forming galaxies and related to the H_alpha/H_beta line strength ratio.

We synthesize the rest-frame Stroemgren colors using SDSS spectra for 99,088 galaxies selected from Data Release 1. This narrow-band ~200 AA photometric system (uz, vz, bz, yz), first designed for the determination of effective temperature, metallicity and gravity of stars, measures the continuum spectral slope of galaxies in the rest-frame 3200-5800 AA wavelength range. Galaxies form a remarkably narrow locus (~0.03 mag) in the resulting color-color diagram. The Bruzual & Charlot population synthesis models suggest that the position of a galaxy along this locus is controlled by a degenerate combination of metallicity and age of the dominant stellar population. Galaxy distribution along the locus is bimodal, with the local minimum corresponding to an ~1 Gyr old single stellar population. The position perpendicular to the locus is independent of metallicity and age, and reflects the galaxy's dust content, as implied by both the models and the statistics of IRAS detections. A comparison of this locus with the galaxy locus in the H_delta-D_n(4000) diagram, utilized by Kauffmann et al. (2003) to estimate stellar masses, reveals a tight correlation, although the two analyzed spectral ranges barely overlap. Overall, the galaxy spectral energy distribution in the entire UV to near-IR range can be described as a single-parameter family with an accuracy of 0.1 mag, or better. This nearly one-dimensional distribution of galaxies in the multi-dimensional space of measured parameters strongly supports the conclusion of Yip et al. (2004), based on a principal component analysis, that SDSS galaxy spectra can be described by a small number of eigenspectra. Apparently, the contributions of stellar populations that dominate the optical emission from galaxies are combined in a simple and well-defined way.

We discuss the UV, optical, and IR properties of the SDSS sources detected by GALEX as part of its All-sky Imaging Survey Early Release Observations. Virtually all of the GALEX sources in the overlap region are detected by SDSS. GALEX sources represent ~2.5% of all SDSS sources within these fields and about half are optically unresolved. Most unresolved GALEX/SDSS sources are bright blue turn-off thick disk stars and are typically detected only in the GALEX near-UV band. The remaining unresolved sources include low-redshift quasars, white dwarfs, and white dwarf/M dwarf pairs, and these dominate the optically unresolved sources detected in both GALEX bands. Almost all the resolved SDSS sources detected by GALEX are fainter than the SDSS 'main' spectroscopic limit. These sources have colors consistent with those of blue (spiral) galaxies (u-r<2.2), and most are detected in both GALEX bands. Measurements of their UV colors allow much more accurate and robust estimates of star-formation history than are possible using only SDSS data. Indeed, galaxies with the most recent (<20 Myr) star formation can be robustly selected from the GALEX data by requiring that they be brighter in the far-UV than in the near-UV band. However, older starburst galaxies have UV colors similar to AGN, and thus cannot be selected unambiguously on the basis of GALEX fluxes alone. With the aid of 2MASS data, we construct and discuss median 10 band UV-optical-IR spectral energy distributions for turn-off stars, hot white dwarfs, low-redshift quasars, and spiral and elliptical galaxies. We point out the high degree of correlation between the UV color and the contribution of the UV flux to the UV-optical-IR flux of galaxies detected by GALEX.

We summarize the detection rates at wavelengths other than optical for \~99,000 galaxies from the Sloan Digital Sky Survey (SDSS) Data Release 1 ``main'' spectroscopic sample. The analysis is based on positional cross-correlation with source catalogs from ROSAT, 2MASS, IRAS, GB6, FIRST, NVSS and WENSS surveys. We find that the rest-frame UV-IR broad-band galaxy SEDs form a remarkably uniform, nearly one parameter, family. As an example, the SDSS u and r band data, supplemented with redshift, can be used to predict K band magnitudes measured by 2MASS with an rms scatter of only 0.2 mag; when measurement uncertainties are taken into account, the astrophysical scatter appears not larger than ~0.1 mag.