Tractor Catalog Photometry

[Jiachuan Xu]

Hi all,

    I'm new to the tractor catalog and I'm trying to reproduce the BGS sample selection as a sanity test. I want to reproduce the star-galaxy separation (Figure 2 in https://arxiv.org/abs/2208.08512, screenshot attached below) using one of the sweep catalogs. However, mine figure (see the 1st attached figure) has a systematic shift in r-z color around 1, which can not be explained by any factor like galactic extinction, etc. The r and z magnitude are calculated as 

    r = 22.5 - 2.5*np.log10(FLUX_R)

    z = 22.5 - 2.5*np.log10(FLUX_Z)

    where FLUX_R or FLUX_Z are the model flux in nanomaggy

    Am I doing anything significantly wrong such that my color is wrong by ~ 1?

Best,

Jiachuan

 

[John Moustakas]

Thanks for your email, Jiachuan.

Unfortunately, this looks like an error in the axis label (or the figure, depending on the interpretation). Here is the code (written by ChangHoon Hahn) which generated those figures, and if I'm interpreting everything correctly, he's actually plotting the g-z (not r-z) color along the bottom axis (scroll to the bottom of the notebook)--

  https://github.com/desi-bgs/bgs-cmxsv/blob/main/doc/sv_paper/sv_overview_bgs_selection.ipynb

Hopefully this helps.

If you have any further questions---especially DESI-related ones---feel free to use the DESI help desk, too--

  https://help.desi.lbl.gov/

Cheers,

-John

[Jiachuan Xu]

Some other questions I have about BGS target selection paper https://arxiv.org/abs/2208.08512

- For the BGS Faint sample, can I take it as (criteria in Sect. 3.1 Selection Cuts) & (criteria in Sect. 3.3 BGS Faint Sample)? Then Sect. 3.3 requires 19.5<r<20.175, however Sect. 3.1 already required r<20 (equation 2) thus the cut at 20.175 is not necessary. The resulting number density I get is much lower than the BGS Faint density in the paper (360 v.s. 530 targets per deg2). Is BGS Faint relaxing the criteria of equation 2 on the faint end?

- In figure 18 which shows the redshift distribution of BGS Bright and Faint from SV3 data, I get a ~50% lower dndz when plotting the SV3 catalog. If I just read off from the figure 18, the total number of BGS Bright is approximately 0.5*(0.5/0.02)*20000 = 250,000 = 1786 targets per deg2 x 140 deg2, which is much higher than the 864/deg2 density. 

Best,

Jiachuan