more examples of problematic galaxy modeling

[moustakas]

Problem areas identified by Maayane Soumagnac (working with Peter Nugent).

Below are a few examples we found of problems with the Tractor modelling. This
is not a complete list, we found these case serendipitously by looking at a few 
hundred cases.

    Issues with modeling big galaxies (the issue is most visible when ticking the 
"DR8 models" box in the Legacy Survey viewer. Tractor basically models a big 
galaxy with several smaller galaxies):
        277.682741 +42.362312
        266.472504 +31.710562
        206.727482 +32.442861
        284.989742 +72.267328
        357.492463 +7.073495
        273.976453 +44.696486
        328.652038 +18.115517
    Issues with star-galaxy classification (galaxies appearing in DR8 as "PSF" 
objects whereas they are galaxies)
        344.845665 +8.751283
        339.337937 +22.379571

[moustakas]

From Peter:

Source RA,Dec = 339.3381, 22.3799
Source type: P
Mags: g=18.06, r=17.66, z=17.41
Brick: 3393p225, Objid: 116
Number of exposures: g=3, r=3, z=4
---------
Source RA,Dec = 277.6812, 42.3647
Source type: E
Mags: g=20.90, r=19.87, z=19.42
Brick: 2777p422, Objid: 6392
Number of exposures: g=3, r=3, z=3

or

Source RA,Dec = 277.6808, 42.3647
Source type: E
Mags: g=17.55, r=17.04, z=16.64
Brick: 2777p422, Objid: 6386
Number of exposures: g=3, r=3, z=3

---------

Source RA,Dec = 266.4726, 31.7109
Source type: P
Mags: g=18.44, r=17.23, z=16.30
Brick: 2663p317, Objid: 1808
Number of exposures: g=3, r=3, z=3

or

Source RA,Dec = 266.4719, 31.7104
Source type: E
Mags: g=20.31, r=19.14, z=18.30
Brick: 2663p317, Objid: 1768
Number of exposures: g=3, r=3, z=3

---------
Source RA,Dec = 206.7238, 32.4438
Source type: D
Mags: g=18.24, r=17.52, z=16.93
Brick: 2067p325, Objid: 838
Number of exposures: g=3, r=3, z=3

or

Source RA,Dec = 206.7259, 32.4424
Source type: E
Mags: g=20.95, r=20.43, z=20.23
Brick: 2067p325, Objid: 842
Number of exposures: g=3, r=3, z=3

or

Source RA,Dec = 206.7285, 32.4435
Source type: E
Mags: g=21.32, r=20.77, z=20.37
Brick: 2067p325, Objid: 841
Number of exposures: g=3, r=3, z=3
---------

Source RA,Dec = 284.9792, 72.2657
Source type: E
Mags: g=16.73, r=16.27, z=15.97
Brick: 2846p722, Objid: 2098
Number of exposures: g=3, r=2, z=4

and 8 others around it.
---------
Source RA,Dec = 357.4923, 7.0722
Source type: E
Mags: g=18.94, r=18.50, z=18.23
Brick: 3576p070, Objid: 3617
Number of exposures: g=4, r=3, z=3

2 others around it

---------
Source RA,Dec = 273.9745, 44.6951
Source type: E
Mags: g=16.25, r=15.64, z=15.21
Brick: 2738p447, Objid: 1538
Number of exposures: g=5, r=5, z=2
---------
Source RA,Dec = 328.6498, 18.1147
Source type: E
Mags: g=20.53, r=19.51, z=18.25
Brick: 3286p180, Objid: 5881
Number of exposures: g=4, r=4, z=5

and 4 others around it
---------
Source RA,Dec = 344.8456, 8.7513
Source type: P
Mags: g=17.53, r=16.97, z=16.91
Brick: 3449p087, Objid: 2442
Number of exposures: g=3, r=3, z=3

[moustakas]

I spot-checked several of these and I don't think there are any specific action items. Galaxies that are large enough to be in the LSLGA (see #536 and #513) will get pre-burned and will have non-parametric ellipse-fitting quantities measured.

Galaxies too small to be in the LSLGA but large enough to have their internal / spiral structure will get shredded, but that's a known issue that comes along with using a small number of smooth, axisymmetric parametric galaxy models in Tractor.

More complex galaxy modeling will need to be done outside of the scope of the main data releases.