Unusual sky spectra

[ivan-katkov]

While comparing spectra from the SkyE and SkyW telescopes, which observed nearly the same point on the sky, I found several very strange cases. This issue records the expnums of these exposures.

I was working with DRP 1.1.1, using SFrames files with full spectra (FLUX+SKY).

Expnums are grouped by the type of problem:

3683 3684 3805 3806 3807 3808 3810 3811 7355 7529 7530 8673 8674 21691 28955

8289 8290

21692 21693 21694 21695 21696 21697 21698 21699

[gblancm]

Hi Ivan, my 2 cents Do you have a sense of how frequent these cases are? I noticed they are all early science frames (tileid=11111). Do we have recent cases? Also, it would be worth checking the guider astrometry solution, as we might have NOT be pointing at the same position in the sky, even though we thought we were. Best Guillermo

…

On Mon, Apr 28, 2025 at 9:10 AM Ivan Katkov ***@***.***> wrote: *ivan-katkov* created an issue (sdss/lvmdrp#209) <#209> While comparing spectra from the SkyE and SkyW telescopes, which observed nearly the same point on the sky, I found several very strange cases. This issue records the expnums of these exposures. I was working with DRP 1.1.1, using SFrames files with full spectra (FLUX+SKY). Expnums are grouped by the type of problem: 3683 3684 3805 3806 3807 3808 3810 3811 7355 7529 7530 8673 8674 21691 28955 image.png (view on web) <https://github.com/user-attachments/assets/eeb6cb1e-77bb-4ac3-adc3-d6fa6455f33f> 8289 8290 image.png (view on web) <https://github.com/user-attachments/assets/b3f4234e-855c-400f-80ea-5ec78740a474> 21692 21693 21694 21695 21696 21697 21698 21699 image.png (view on web) <https://github.com/user-attachments/assets/aa3bfc8c-b7f7-4134-8212-311d9c004610> — Reply to this email directly, view it on GitHub <#209>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AIOKNY3UOLWT6ILR2SKFCIT23YSC5AVCNFSM6AAAAAB4AMD3SSVHI2DSMVQWIX3LMV43ASLTON2WKOZTGAZDIOBYGY3TCMY> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

--

------------------------------------------------------------------ Guillermo A. Blanc Associate Director of Strategic Initiatives Observatories of the Carnegie Institution for Science and Las Campanas Observatory (LCO) 813 Santa Barbara St, Pasadena, CA, 91101, USA ***@***.*** ***@***.***>

------------------------------------------------------------------ *I believe in the importance of personal time. I do not expect a response to this email outside of your normal working hours.* *Creo en la importancia del tiempo personal. No espero una respuesta a este correo fuera de su horario laboral normal.*

[ivan-katkov]

For this comparison, I used observations where the distance between (skye_ra, skye_dec) and (skyw_ra, skyw_dec) is less than 2 degree and excluded observations where the coordinates point to (ra=0, dec=0). There are 163 such records in DRPall file. All exposures mentioned above come from these 163.

If needed, I can relax distance criterion or think of method to identify such unusual spectra.

I used the Sky telescope coordinates from sdsswork/lvm/spectro/redux/1.1.1/drpall-1.1.1.fits. Could these coordinates be incorrect?

[kslong]

It's interesting. My guess is that this might be a general calibration problem with those frames. I know that the sky fibers are spread among the spectrographs, but I don't know there distribution. I think it would be interesting for those examples to plot some kind of average of the nearest science fibers for the two sky telescopes, possibly just add one to the fiber number, and see whether one sees a similar pattern. I would try to determinewhether this is a sky telescope issue, or a more general one.

[ivan-katkov]

Exposures 3683 3684 7355 7529 7530 8673 8674 21691 28955 8289 8290 21692 21693 21694 21695 21696 21697 21698 21699 have Spec spectra that are entirely NaNs

Exposures 3805 3806 3807 3808 3810 3811 also show strange spectra that seem to appear out of nowhere in North hemishpere
https://data.sdss5.org/lvmvis/?q=3805+%3C=+expnum+%3C=+3811&expnum=3811&fiber=278

https://data.sdss5.org/valis-lvmvis-api/lvm/plot_exposure_spectrum/?l=id:1.1.1/3805;type:skyflux;telescope:Sci&l=id:1.1.1/3805;type:skyflux;telescope:SkyE&l=id:1.1.1/3805;type:skyflux;telescope:SkyW&l=id:1.1.1/3805;type:skyflux;telescope:Spec&legend=long

Are we interested to know how many Spec spectra consisting entirely of NaNs?

[ivan-katkov]

Minor addition:

Among the exposures listed in the drpall-1.1.1.fits file, there are 202 exposures where Spec spectra ((slitmap['telescope'] == 'Spec') & (slitmap['fibstatus'] == 0)) contain only NaN values. These exposures are:

3582 3583 3584 3585 3683 3684 3821 3823 3824 3822 3826 3825 3829 3827 3830 3828 3833 3835 3834 3836 3837 3838 3863 3862 3864 3865 3867 3868 3866 3869 3874 3871 3877 3870 3875 3876 3879 3878 3937 3938 3939 3936 3940 3942 3944 3943 3941 3945 3948 3950 3951 3949 3952 3953 3955 3982 4001 4002 4003 4004 4008 4005 4006 4007 4013 4009 4015 4014 4016 4017 4019 4018 4051 4052 4050 4054 4055 4053 4056 4059 4063 4057 4058 4062 4065 4064 4069 4066 4068 4067 4135 4136 4137 4139 4140 4138 4142 4141 4144 4148 4143 4147 4150 4149 4152 4151 4154 4181 4183 4182 4184 4185 4186 4187 4188 4189 4190 4194 4193 4195 4196 4197 4198 5849 6641 7122 7355 7529 7530 8289 8290 8583 8584 8673 8674 9019 9018 9219 10750 13450 13578 13577 13778 13834 14110 14353 14357 14407 14512 14513 14671 14719 14720 15025 15026 15027 15231 15353 15457 15514 15638 15640 15639 15745 15945 16200 18619 18620 20057 21690 21692 21693 21694 21695 21696 21699 21691 21698 21697 21959 21958 22071 23815 23959 23958 24205 24208 24204 24206 24207 24209 24212 24210 24211 24213 24284 24285 27637 28955 29082 29087 29085

These histograms below show distribution of exposures based on the number of NaNs pixels in spectra, separated by telescope.