Star formation rate estimates

The star formation rates that you will find here are based on the technique discussed in Brinchmann et al (2004, B04) and you should cite that paper if you use these SFRs. In particular the fits to star forming galaxies are carried out using the Charlot & Longhetti (2001, CL01) model and the method to obtain SFR estimates for other classes of galaxies (AGN, Composite, low S/N SF, low S/N LINER and Unclassifiable) is near identical to that described in B04.

However there are some modifications in the methodology:

Estimates of SFR for non-SF galaxies

For those galaxies where we cannot use the emission lines to estimate the SFR in the fibre directly, such as AGN and composite galaxies, we have made a slight modification to the procedure.

To recall, in B04, we constructed the likelihood distribution the specific SFR, SFR/M*, as a function of D4000 using the star-forming sample. This use of D4000 to estimate specific SFR was found to work fairly well. Likewise for the low-S/N SF galaxies we constructed the average conversion factor from observed Hα luminosity to SFR and used this to estimate SFRs for those galaxies. However it is important to realise that the resulting PDFs would be appropriate only in average and that it implicitly applies a dust correction similar to that of the average in the SDSS SF sample, about 1 magnitude at V.

In B04 we used SDSS DR2 and the sample size did not allow us to do this properly for different dust attenuations. This has changed now and we have therefore modified our method somewhat in that when the galaxy has Hα and Hβ with S/N>3 we construct the PDF using galaxies with similar Hα/Hβ. This removes all trends with dust attenuation and removes a bias. It does mean that the SFRs estimated for AGN and Composite galaxies are typically somewhat higher than with the previous technique.

Aperture corrections of SFRs

In B04 the aperture corrections were done in an empirical manner: The distribution of SFR/M* at a given (g-r, r-i) colour was constructed and then applied to the photometry outside the fibre.

This was found to overestimate the SFR of galaxies with low levels of star formation by Salim et al (2007). The problem here is again that the average distributions are inappropriate in some regions of parameter space. In particular the average SFR/M* estimated within fibre has a larger contribution from dusty high-metallicity star-bursts than is seen in the outer regions of galaxies.

To remedy this we build on the Salim et al work - they showed that by fitting photometry to stochastic models it was possible to remove most of that bias. Thus our procedure here is to calculate the light outside the fibre for each galaxy, and then fit stochastic models, similar to those used by Salim et al to this photometry. This works well and should be a much better aperture correction.


With those preliminaries out of the way, here are the files for the (specific) star formation rates both measured within the fibre and outside (note: you can also download them directly from the Data directory).

The AVG, MEDIAN, MODE give the mean, median and mode of the PDF of the SFR or SFR/M*, for each object, while P2P5, P16, P84, P97P5 are the 2.5th, 16th, 84th and 97.5th percentiles respectively. The ENTROPY key provides the entropy of the PDF defined as the sum over P ln(P) in the PDF. The FLAG keyword indicates the status of the SFR estimation. If the FLAG is set to 0 then all is well and for statistical studies in particular, it is recommendable to focus on these objects as in all other cases the detailed method to estimate SFR or SFR/M* will be (slightly) different and can introduce subtle biases.