[Previous] | [Session 21] | [Next]
E. J. Barton (DAO/HIA/NRC), M. J. Geller, S. J. Kenyon (CfA)
Galaxy-galaxy interactions are a significant cause of galaxy evolution, although their aggregate effects remain unknown. Observations of nearby galaxy pairs provide a method for measuring these effects at the current epoch and for calibrating the effects at high redshift. We present a comprehensive analysis of a large, complete sample of galaxy pairs selected from the CfA2 redshift survey. Observations of subsamples of the full pair sample (786 galaxies) include nuclear spectra of 502 galaxies, images of 190 galaxies, and H\alpha rotation curves of 84 spirals.
Measures of recent star formation such as H\alpha equivalent width, EW(H\alpha), correlate significantly with the orbit parameters (projected separation on the sky, line-of-sight velocity difference), providing unambiguous evidence that the sample is rich in interactions which result in star formation. The correlations provide a method for measuring both the durations and the IMFs of the tidally-triggered bursts. The starburst ages match the dynamical timescales only for continuous (\gtrsim 108 years) star formation and a Miller-Scalo-type IMF. However, superposed old stellar populations and dust complicate these conclusions.
We use B and R images with the nuclear spectra to estimate the relative strengths of the old and new stellar populations. If, as the measurements suggest, the contributions to EW(H\alpha) from the old populations and dust are moderate, the Miller-Scalo IMF remains the best description of the star formation. In this case, the central bursts of star formation can be strong (20 -- 100% in R). However, only a few percent of the galaxies lie far from the R-band Tully-Fisher relation because the bursts usually contribute \lesssim 1 magnitude in R. We discuss the limits our measurements place on luminosity evolution in pairs. In galaxy formation models, these data are a basis for semi-analytic prescriptions for interaction-triggered star formation.