2010ApJ...725..534F

We investigate the relationship between spiral arms and star formation in the grand-design spirals NGC 5194 and NGC 628 and in the flocculent spiral NGC 6946. Filtered maps of near-IR (3.6 µm) emission allow us to identify "arm regions" that should correspond to regions of stellar mass density enhancements. The two grand-design spirals show a clear two-armed structure, while NGC 6946 is more complex. We examine these arm and interarm regions, looking at maps that trace recent star formation–far-ultraviolet (GALEX NGS) and 24 µm emission (Spitzer SINGS)–and cold gas-CO (HERACLES) and H I (THINGS). We find the star formation tracers and CO more concentrated in the spiral arms than the stellar 3.6 µm flux. If we define the spiral arms as the 25% highest pixels in the filtered 3.6 µm images, we find that the majority (60%) of star formation tracers occur in the interarm regions; this result persists qualitatively even when considering the potential impact of finite data resolution and diffuse interarm 24 µm emission. Even with a generous definition of the arms (45% highest pixels), interarm regions still contribute at least 30% to the integrated star formation rate (SFR) tracers. We look for evidence that spiral arms trigger star or cloud formation using the ratios of SFR (traced by a combination of FUV and 24 µm emission) to H₂(traced by CO) and H₂ to H I. Any enhancement of SFR/M(H₂) in the arm region is very small (less than 10%) and the grand-design spirals show no enhancement compared to the flocculent target. Arm regions do show a weak enhancement in H₂/H I compared to the interarm regions, but at a fixed gas surface density there is little clear enhancement in the H₂/H I ratio in the arm regions. Thus, it seems that spiral arms may only act to concentrate the gas to higher densities in the arms.