Converting mock image counts to flux / magnitudes

Andrew Cooper
  • 28 Jul '20

Dear Dylan,

According to their headers, the mock SDSS FITS images from SKIRT for TNG100 are in units of "counts per second per pixel".

Please can you tell me how I should convert those counts to flux (ergs/s/cm^2, nanomaggies, or whatever) so I can correctly compare (say) the g and r magnitudes measured in apertures on the images against the whole-galaxy magnitudes reported in SubhaloStellarPhotometrics?

I didn't see any explicit mention of zeropoints or similar in the documentation or in Rodriguez-Gomez et al.; apologies if I'm missing something obvious.

Thanks in advance,

Vicente Rodriguez-Gomez
  • 31 Jul '20

Dear Andrew,

Thanks for your interest, and sorry for the omission in the documentation.

The conversion factor here is "fluxmag0" (following HSC nomenclature), which is the instrumental flux (in electrons/s) that would correspond to a magnitude of zero (3631 Jy in the AB magnitude system). I used the following values when generating the images (these were calculated theoretically using data from the SVO Filter Profile Service):

fluxmag0(sdss_g) = 2.40238e+10 electrons/s
fluxmag0(sdss_r) = 2.3826e+10 electrons/s
fluxmag0(sdss_i) = 1.59001e+10 electrons/s
fluxmag0(sdss_z) = 2.66542e+09 electrons/s

So, in order to convert the fluxes measured in the apertures (e.g. "flux_circ") to AB magnitudes, the following should do the job:

m_AB = -2.5 * [log10(flux) - log10(fluxmag0)]

By the way, I cannot remember if the r-band measurements were included in the public data release (apart from the g & i bands), but I must have them somewhere. Please let us know if you need them.


Andrew Cooper
  • 31 Jul '20

Dear Vicente,

Excellent, that's exactly what I was looking for. Many thanks for your quick and helpful response.

I asked about these zeropoints because I want measure aperture fluxes myself from the 2D images, which are available for g,r,i,z. I didn't know about 'flux_circ', which I now understand is one of the outputs in your StatMorph tables. Indeed, it seems those tables are only available in g and i. Nevertheless it's very helpful to know that aperture measurement (and the r band equivalent) exists, as a cross-check.

Thanks again,

  • Page 1 of 1