A TEST OF STAR FORMATION LAWS IN DISK GALAXIES

We use observations of the radial profiles of the mass surface density of total, Sigma(g), and molecular, Sigma(H2), gas rotation velocity and star formation rate surface density, Sigma(sfr), of the molecular dominated regions of 12 disk galaxies from Leroy et al. to test several star formation laws: a "Kennicutt-Schmidt power law," Sigma(sfr) = A(g)Sigma(1.5)(g,2); a "constant molecular law," Sigma(sfr) = A(H2)Sigma(H2,2); the "turbulence-regulated laws" of Krumholz & McKee (KM) and Krumholz, McKee, & Tumlinson (KMT), a "gas-Omega law," Sigma(sfr) = B(Omega)Sigma(g)Omega; and a shear-driven "giant molecular cloud (GMC) collisions law," E(sfr) = B(CC)Sigma(g)Omega(1 - 0.7 beta), where beta equivalent to d ln nu(circ)/d ln r. We find the constant molecular law, KMT turbulence law, and GMC collision law are the most accurate, with an rms error of a factor of 1.5 if the normalization constants are allowed to vary between galaxies. Of these three laws, the GMC collision law does not require a change in physics to account for the full range of star formation activity seen from normal galaxies to circumnuclear starbursts. A single global GMC collision law with B(CC) = 8.0 x 10(-3), i.e., a gas consumption time of 20 orbital times for beta = 0, yields an rms error of a factor of 1.8.