Tidally triggered star formation in close pairs of galaxies

We analyze optical spectra of a sample of 502 galaxies in close pairs and N -tuples, separated by less than or equal to 50 h(-1) kpc. We extracted the sample objectively from the CfA2 redshift survey, without regard to the sur roundings of the tight systems; we remeasure the spectra with longer exposu res, to explore the spectral characteristics of the galaxies. We use the ne w spectra to probe the relationship between star formation and the dynamics of the systems of galaxies. The equivalent widths of H alpha [EW(H alpha)] and other emission lines ant icorrelate strongly with pair spatial separation (Delta D) and velocity sep aration; the anticorrelations do not result from any large-scale environmen tal effects that we detect. We use the measured EW(Hcl) and the starburst m odels of Leitherer et al. to estimate the time since the most recent burst of star formation began for galaxies in our sample. In the absence of a lar ge contribution from an old stellar population to the continuum around H al pha that correlates with the orbit parameters, the observed Delta D-EW(H al pha) correlation signifies that starbursts with larger separations on the s ky are, on average, older. We also find a population of galaxies with small to moderate amounts of Balmer absorption. These galaxies support our concl usion that the sample includes many aging bursts of star formation; they ha ve a narrower distribution of velocity separations, consistent with a popul ation of orbiting galaxies near apogalacticon. By matching the dynamical timescale to the burst timescale, we show that th e data support a simple picture in which a close pass initiates a starburst ; EW(H alpha) decreases with time as the pair separation increases, account ing for the anticorrelation. Recent N-body/smoothed particle hydrodynamics simulations of interacting pairs suggest a physical basis for the correlati on-for galaxies with shallow central potentials, they predict gas infall be fore the final merger. This picture leads to a method for measuring the dur ation and the initial mass function of interaction-induced starbursts: our data are compatible with the starburst models and orbit models in many resp ects, as long as the starburst lasts longer than similar to 10(8) yr and th e delay between the close pass and the initiation of the starburst is less than a few times 10(7) yr. If there is no large contribution from an old st ellar population to the continuum around H alpha, the Miller-Scale and cuto ff (M less than or equal to 30 M.) Salpeter initial mass functions (IMFs) f it the data much better than a standard Salpeter IMF.