The SIRTF Nearby Galaxy Survey is a comprehensive infrared imaging and spectroscopic survey of 75 nearby galaxies. Its primary goal is to characterize the infrared emission of galaxies and their principal infrared-emitting components, across a broad range of galaxy properties and star formation environments. SINGS will provide new insights into the physical processes connecting star formation to the interstellar medium properties of galaxies, and provide a vital foundation for understanding infrared observations of the distant universe and ultraluminous and active galaxies. The galaxy sample and observing strategy have been designed to maximize the scientific and archival value of the data set for the SIRTF user community at large. The SIRTF images and spectra will be supplemented by a comprehensive multiwavelength library of ancillary and complementary observations, including radio continuum, HI, CO, submillimeter, BV RIJHK, Hα, Paschen-α, ultraviolet, and X-ray data. This paper describes the main astrophysical issues to be addressed by SINGS, the galaxy sample and the observing strategy, and the SIRTF and other ancillary data products.

We present intermediate-resolution optical spectrophotometry of 65 galaxies obtained in support of the Spitzer Infrared Nearby Galaxies Survey (SINGS). For each galaxy we obtain a nuclear, circumnuclear, and semi-integrated optical spectrum designed to coincide spatially with mid-and far-infrared spectroscopy from the Spitzer Space Telescope. We make the reduced, spectrophotometrically calibrated one-dimensional spectra, as well as measurements of the fluxes and equivalent widths of the strong nebular emission lines, publically available. We use optical emission-line ratios measured on all three spatial scales to classify the sample into star-forming, active galactic nuclei (AGN), and galaxies with a mixture of star formation and nuclear activity. We find that the relative fraction of the sample classified as star-forming versus AGN is a strong function of the integrated light enclosed by the spectroscopic aperture. We supplement our observations with a large database of nebular emission-line measurements of individual H ii regions in the SINGS galaxies culled from the literature. We use these ancillary data to conduct a detailed analysis of the radial abundance gradients and average H iiregion abundances of a large fraction of the sample. We combine these results with our new integrated spectra to estimate the central and characteristic (globally-averaged) gas-phase oxygen abundances of all 75 SINGS galaxies. We conclude with an in-depth discussion of the absolute uncertainty in the nebular oxygen abundance scale.

We present accurate resolved WISE photometry of galaxies in the combined SINGS and KINGFISH sample. The luminosities in the W3 12µm and W4 23µm bands are calibrated to star formation rates (SFRs) derived using the total infrared luminosity, avoiding UV/optical uncertainties due to dust extinction corrections. The W3 relation has a 1-σ scatter of 0.15 dex over nearly 5 orders of magnitude in SFR and 12µm luminosity, and a range in host stellar mass from dwarf (10 7 M ) to ∼ 3×M (10 11.5 M ) galaxies. In the absence of deep silicate absorption features and powerful active galactic nuclei, we expect this to be a reliable SFR indicator chiefly due to the broad nature of the W3 band. By contrast the W4 SFR relation shows more scatter (1-σ = 0.18 dex). Both relations show reasonable agreement with radio continuum-derived SFRs and excellent accordance with so-called "hybrid" Hα + 24µm and FUV+24µm indicators. Moreover, the WISE SFR relations appear to be insensitive to the metallicity range in the sample. We also compare our results with IRAS-selected luminous infrared galaxies, showing that the WISE relations maintain concordance, but systematically deviate for the most extreme galaxies. Given the all-sky coverage of WISE and the performance of the W3 band as a SFR indicator, the L 12µm SFR relation could be of great use to studies of nearby galaxies and forthcoming large area surveys at optical and radio wavelengths.

Using the new capabilities of Spitzer and extensive multiwavelength data from SINGS, it is now possible to study the infrared properties of star formation in nearby galaxies down to scales equivalent to large H ii regions. We are therefore able to determine what fraction of large, infrared-selected star-forming regions in normal galaxies are highly obscured and address how much of the star formation we miss by relying solely on the optical portion of the spectrum. Employing a new empirical method for deriving attenuations of infrared-selected star-forming regions, we investigate the statistics of obscured star formation on 500 pc scales in a sample of 38 nearby galaxies. We find that the median attenuation is 1.4 mag in H and that there is no evidence for a substantial subpopulation of uniformly highly obscured star-forming regions. The regions in the highly obscured tail of the attenuation distribution (A H k 3) make up only $4% of the sample of nearly 1800 regions, although very embedded infrared sources on the much smaller scales and lower luminosities of compact and ultracompact H ii regions are almost certainly present in greater numbers. The highly obscured cases in our sample are generally the bright, central regions of galaxies with high overall attenuation but are not otherwise remarkable. We also find that a majority of the galaxies show decreasing radial trends in H attenuation. The small fraction of highly obscured regions seen in this sample of normal, star-forming galaxies suggests that on 500 pc scales the timescale for significant dispersal or breakup of nearby, optically thick dust clouds is short relative to the lifetime of a typical star-forming region.