LS Shape Measurement

[Jiachuan Xu]

Hi all,

    Hello! I have a question about the shape measurements `SHAPE_E1` & `SHAPE_E2` in the LS photometry catalog. We want to select bright disk galaxies for some weak lensing study, and our forecast needs the ellipticity distribution (sqrt(SHAPE_E1**2 + SHAPE_E2**2)) of our galaxy sample. 

See the figure below; I compare the ellipticity of LS galaxies:

    - [EXP OR SERSIC] MORPHTYPE classified as EXP or SER

    - [Late-type Galaxy] MORPHTYPE classified as EXP or (SER and SERSIC<=2)
    - [Sample 3] satisfies our sample definition, which is  (EXP or (SER and SERSIC<=2)) and (SHAPE_R > 1.0")

with the ellipticity of CFHT galaxies (the red dash-dotted line)

    The LS galaxies are missing many low-ellipticity, round-ish galaxies, which we like for some reason. I notice that galaxies with MORPHTYPE==(PSF OR REX) are fixed to zero ellipticity. However, realistic PSF generally has non-zero ellipticity. 

    May I ask how the MORPHTYPE is classified and how the SHAPE_E1/2 is measured?

Could it be some low-ellipticity galaxies are classified as REX, and their ellipticity is artificially fixed to 0? How reliable are r_mag<20 disk galaxies' shape measurements for weak lensing purposes?

      Thanks! And wish you a happy holiday break!

Best,

Jiachuan

[Dustin Lang]

Hello,

The SHAPE_E1,E2 values are the "above-the-atmosphere" intrinsic galaxy ellipticities, and they are fit by forward-modeling galaxy models correlated with the PSF model of individual images.

Morphological TYPE is chosen through a model-selection process that chooses the lowest chi-squared model fit but penalizes the more complicated models.  PSF is the simplest model, REX has only one additional parameter (the radius), EXP and DEV have two more parameters (SHAPE_E1,E2), and SER has one more (the Sersic index).

As you say, round EXP galaxies will tend to get classified as REX, because the REX model is exactly equal to an EXP with SHAPE_E1=SHAPE_E2 =0, and EXP galaxies with small E1,E2 values are *pretty close* in appearance to REX.  So indeed, what you see is expected.

We do have all the parameter values for all the model fits in the "all-models.fits" files, eg https://portal.nersc.gov/cfs/cosmo/data/legacysurvey/dr10/south/metrics/000/all-models-0001m002.fits

cheers,

dustin

--
You received this message because you are subscribed to the Google Groups "DECam Legacy Survey" group.
To unsubscribe from this group and stop receiving emails from it, send an email to decam-legacy-su...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/decam-legacy-survey/a25a1456-7f68-4fac-96b8-e9e15118c9d6n%40googlegroups.com.