incorrect FITS delivered through Jupyter notebook

Question

I have a jupyter notebook file that I use to download a lot of FITS images from the legacy survey dr7, but I noticed for a small fraction of the FITS that I download it does not actually have the galaxy in it. For example, I'm looking at an object around ra,dec = 16.632, -1.114 but I'm actually getting a file where the wcs says that it is centered around ra,dec = 16.6320,-0.8744. Checking some other files it always seems that the declination is always about 0.24 degrees higher than the object I desire, while the ra is about correct.

The actual download command I use is a modified version of what shows up in one of the tutorial jupyter notebooks that's mainly an svc.search at a specific ra,deg, and given FOV and then some conditions to select the right exposure.

Let me know if you have any idea what's going on here. Thanks

Answer 1

Hi, thanks for reaching out.

Re your first issue:

Could you please post here the exact SIA query you are trying, i.e. at least these lines:

DEF_ACCESS_URL = "https://datalab.noao.edu/sia/coadd_ls_dr7" # Legacy Survey DR7 stacks
svc = sia.SIAService(DEF_ACCESS_URL)
fov = 0.1 # in degrees
ra, dec = 16.632, -1.114
imgTable = svc.search((ra,dec), (fov/np.cos(dec*np.pi/180), fov), verbosity=2).to_table()

where you replace with your values the DEF_ACCESS_URL, fov, ra, dec.
I tried your coordinates (ra, dec = 16.632, -1.114) on our web version of the cutout service, with fov=0.1, and I do get rectangular (i.e. not square) images. This is usually the case when the FOV is larger than the distance of (ra,dec) from an image edge. We currently do not force the resulting image to be square (and for instance zero-padded). Please also see

  https://datalab.noao.edu/help/index.php?qa=630&qa_1=cutout-issues-with-ls_dr7-cont-d

 on this issue.

Re the 0.24 deg offset issue: can you please provide a few ra/dec pairs, plus the FOV, that give you this offset?

Many thanks!

Robert for the DL team

Comments

Sure, the lines of interests are:

DEF_ACCESS_URL = "http://datalab.noao.edu/sia/ls_dr7" # DES SIA service URL
fov = 0.05
ra,dec = 16.632, -1.114
imgTable = svc.search((ra,dec), (fov,fov), verbosity=2).to_table()
*I was curious why the ra fov looked like that, is it some small order correction? I did the math for the example there and effectively got 0.05

Re the 0.24 deg offset issue: Sure, here's a few examples (where the one in my original post also applies). I checked it by pulling it up in ds9 and switching the wcs to degrees

ra,dec
16.84763,0.89277
19.82096,-0.113676
9.668481,1.140725

Thanks! (also if you could just respond to this comment to let me know someone received this that would be much appreciated)

---

Hey Robert, just wanted to confirm that you got my last message. I think one time I commented on a reply and the other person never saw it

---

Hi Derrick, apologies for the delay. We are looking into the issue, and will report back here.
BTW, if you can see a message here, then everyone else can as well.
We'll be back, thank you.

---

Hello, just wondering if there has been any update on this since.