The Lick index H delta A -- why it belongs in the data set
-
by
mlpeck
I posted a version of this graph before, but this time I've done it using measurements from the MPA-JHU pipeline. Kauffmann et al. (2003a and 2003b) found that the relationship between the Lick index H delta A and the 4000 Å break index is an important diagnostic of burst strength in starburst and post starburst galaxies. This was also explored by Balogh et al. (1999) in a paper that first introduced the "narrow" form of the D4000 index. Here is a plot of H delta A against D4000 for the quench and control samples:
The cloud of light gray points and contour lines are the control sample. The colored points are the quench sample. Color coding is the same as the BPT diagrams I posted here (scroll down to the second post for the MPA-JHU derived BPT diagram).
It's interesting to compare this graph to Figures 5 and 6 -- especially Figure 6 -- of the first of the Kauffmann et al. papers. In Figure 6 they have color coded regions where strong starbursts occurred in the recent past in model galaxies. That region lies approximately above the outermost contour line of the graph above and to left of D4000=1.5. Very roughly ~28% of the quench sample galaxies fall in that region, compared to ~1.5% of the control galaxies.
Also, and I may be fooling myself here, it appears that many of the AGN/LINER galaxies fall in this region, as well as many of the BPT unclassified galaxies, which will mostly be quiescent. Starforming and "transitional" galaxies seem to fall mostly in the denser part (in the control set) of the blue sequence.
I'm a little surprised that there was little or no quantitative discussion of Balmer line absorption strengths in the recent literature on post-starburst or recently quenched galaxies. There was a histogram of H delta A index values in Schawinski et al.'s (2009) paper on blue (mostly star forming) early type galaxies, but that's all I've seen in the recent scientific literature linked here.
Posted
-
by
mlpeck
Today I'm going to focus a little more on the quench sample in the H-delta - D4000n plane. Here is a plot of the quench sample alone, color coded as above by the BPT classification:
The curved line here is approximately the lower boundary of galaxies that had experienced a strong starburst (involving 5% or more of the total mass) in the recent past in Kauffmann et al.'s (2003) models. This is taken from their figure 6.
A simpler and more conventional criterion for recently quenched post starburst galaxies is H-deltaA > 5 Å (eg. Goto et al. 2003, Goto 2007, Schawinski et al. 2009).
35% of the quench sample (1048 out of 3000) lie above the Kauffmann et al. strong burst threshold. Over 49% (1481/3000) satisfy the H-deltaA > 5 Å criterion.
Here's a breakdown of BPT class by H-delta strength:
> table(ea.quench,kclass.quench) kclass.quench ea.quench 0 1 2 3 4 FALSE 285 744 347 79 64 TRUE 557 271 252 146 255This confirms my eyeball analysis above. The strong H-delta region is heavily populated by both AGN/LINER systems and (probably) quiescent ones. Star forming galaxies are preferentially below the H-delta strong limit.
Here's a quick cut and past look at H-delta strength against a couple of morphological classes, namely "smooth" and "merging." The one noteworthy thing I see here is that the strong H-delta group has a larger number of objects that show evidence of disturbance (everything but "Neither" in the merging classification).
> table(ea.quench,class.quench$smooth) ea.quench Features or disk Smooth Star or artifact FALSE 466 1047 4 TRUE 286 1190 3 > table(ea.quench,class.quench$merging) ea.quench Both Merging Neither Tidal debris FALSE 0 14 84 1305 110 TRUE 0 27 88 1118 243Posted