Physical Sources of Scatter in the Tully-Fisher Relation

We analyze residuals from the Tully-Fisher relation for the emission-line galaxies in the Nearby Field Galaxy Survey, a broadly representative survey designed to fairly sample the variety of galaxy morphologies and environments in the local universe for luminosities from MB = -15 to -23. For a subsample consisting of the spiral galaxies brighter than M = -18, we find strong correlations between Tully-Fisher residuals and both B-R color and EW(Hα). The extremes of the correlations are populated by Sa galaxies, which show consistently red colors, and spiral galaxies with morphological peculiarities, which are often blue. If we apply an EW(Hα)-dependent or B-R color-dependent correction term to the Tully-Fisher relation, the scatter in the relation no longer increases from R to B to U but instead drops to a nearly constant level in all bands, close to the scatter we expect from measurement errors. We argue that these results probably reflect correlated offsets in luminosity and color as a function of star formation history. Broadening the sample in morphology and luminosity, we find that most nonspiral galaxies brighter than M = -18 follow the same correlations between Tully-Fisher residuals and B-R color and EW(Hα) as do spirals, albeit with greater scatter. However, the color and EW(Hα) correlations do not apply to galaxies fainter than M = -18 or to emission-line S0 galaxies with anomalous gas kinematics. For the dwarf galaxy population, the parameters controlling Tully-Fisher residuals are instead related to the degree of recent evolutionary disturbance: overluminous dwarfs have higher rotation curve asymmetries, brighter U-band effective surface brightnesses, and shorter gas consumption timescales than their underluminous counterparts. As a result, sample selection strongly affects the measured faint-end slope of the Tully-Fisher relation, and a sample limited to include only passively evolving, rotationally supported galaxies displays a break toward steeper slope at low luminosities.