Galaxy Halpha distribution versus environment
Spurred by a recent tweet, I remembered some nice movies I made to accompany a paper a few years ago. After a bit of a search I’ve found them and thrown them up here. They look a bit old-school in these web 2.0 days, but are still very relevant to studies of the galaxy population versus environment.
The grey surface in the animated figure above (click for a large version) shows the Halpha equivalent width distribution (actually an offset logarithm of EW) versus local galaxy density, as determined using a conditional (probability) density estimator using non-parametric mixture model regression (Rojas et al. 2006, Bamford et al. 2008). The coloured surfaces show the four Gaussian components comprising the distribution, determined as the optimum for fitting the data by the method.
The paper is entitled “Revealing components of the galaxy population through non-parametric techniques“. It applies a cool statistical method, developed by Alex Rojas, to study the dependence of galaxy populations on environment (local density) using the Halpha equivalent-width distribution of galaxies from SDSS. The key findings are:
- Using only the Halpha EW distribution, one can determine that the overall distribution comprises four distinct components.
- These components may be identified as star-forming, Seyfert, LINER and passive galaxy populations (by comparing with classifications from line ratio diagnostics).
- While the overall Halpha EW distribution varies greatly with environment, the distributions of the individual components barely change.
- The variation in the overall distribution with environment is simply due to the components varying in relative number.
There is also a movie showing slices through the distribution, which I think is a really effective visualisation.
These plots were created using an early version of S2PLOT (converted and animated using ImageMagick).