search

ACES-CMZ / reduction_ACES

Reduction scripts and tools for ACES
https://worldwidetelescope.org/webclient/?wtml=https://data.rc.ufl.edu/pub/adamginsburg/ACES/mosaics/mosaics.wtml
15 stars 12 forks source link

Execution Block ID uid://A001/X15b4/X43 Sgr_A_st_m_03_7M_updated #110

Open keflavich opened 2 years ago

keflavich commented 2 years ago

Sgr_A_st_m_03_7M_updated uid://A001/X15b4/X43

Product Links:

Reprocessed Product Links:

piposona commented 2 years ago

This updated dataset was deliverd as dynamic range limited with achieved/requested rms 1.31mJy/0.62 mJy, however, the not updated SB Sgr_A_st_m_03_7M, was delivered with the right achieved/requested rms 0.68mJy/0.62 mJy. Just in case we deem it needed to request further observations.

ashleythomasbarnes commented 2 years ago

@piposona - sorry if I'm being slow, where are these numbers coming from (lines/continuum)? And how do you know this is dynamic range limited, is there a warning for this?

astroG76 commented 2 years ago

(1) Wind speed is a little bit high and shadowed percentage is relatively high, which may be one of the reasons that lots of data were flagged.

X2639_flagged X2639_shadowed

uid_A002_Xf73ead_X2639_target_highwindspeed

(2) Step 38. Tclean/MakeImages got very low scores in Spw 20, 22, 24. For Spw 20 and 24, there is a huge amount of emission in line-free mom0 and mom8 maps, especially Spw 24. For Spw 20, it seems no line is counted in the spectra, but the line was not settled in the center. For Spw 24, it may be affected by the line wing.

Sgr_A_star_SpW20_mom0_8

uid___A001_X15b4_X43 s38_0 Sgr_A_star_sci spw20 cube I iter1 image spectrum

Sgr_A_star_SpW24_mom0_8

uid___A001_X15b4_X43 s38_0 Sgr_A_star_sci spw24 cube I iter1 image spectrum

For 22, which got 0, it warned clean reached niter limit and the final model is bad.

uid___A001_X15b4_X43 s38_0 Sgr_A_star_sci spw22 cube I iter1 model sky

keflavich commented 2 years ago

spw24 has a real issue. https://user-images.githubusercontent.com/98576159/180650277-8a4d6415-f630-4c77-94cb-3442f9a138cd.png didn't show up in the post, but it's clear that the continuum is being fit to the wrong channels.

ashleythomasbarnes commented 2 years ago

Find cont for spw24 from previous observations

ashleythomasbarnes commented 1 year ago

Update cont.dat file and run pipeline - upload HPG and tell Adam - also be aware of changes in the continuum

xinglunju commented 11 months ago

Started to run the pipeline with updated cont.dat, from:

SpectralWindow: 24
98.0026453577~98.0172920687GHz LSRK
99.4116589559~99.4136118507GHz LSRK
99.4346054698~99.4360701409GHz LSRK
99.4433934964~99.4477875097GHz LSRK
99.4741515895~99.4824513924GHz LSRK

to

SpectralWindow: 24
97.8500000000~97.9000000000GHz LSRK
99.4116589559~99.4136118507GHz LSRK
99.4346054698~99.4360701409GHz LSRK
99.4433934964~99.4477875097GHz LSRK
99.4741515895~99.4824513924GHz LSRK
xinglunju commented 11 months ago

New images look good.

Uploaded them to Globus: /upload/Sgr_A_st_m_03_7M_updated_reclean/, including the updated cont.dat, the SPW24 cubes, and the continuum image. @keflavich

SPW24 spectrum and selected 'line-free' channels: uid___A001_X15b4_X43 s12_0 Sgr_A_star_sci spw24 cube I iter1 image spectrum

Continuum: uid___A001_X15b4_X43 s10_0 Sgr_A_star_sci spw16_18_20_22_24_26 cont I iter1 image tt0 I mean sky

djeff1887 commented 10 months ago

Wanted to note my "patch" to the confetti problem in spw22: I set noisethreshold=3, which as I understood it, makes it so that automultithresh interprets lower amplitude signal as real instead of as noise (which seems to mitigate the problem of "borderline" signal making the mask flicker on and off). I also set cyclefactor=3 to remove divergence that came along after changing noisethreshold. There may still be a few cases of pixels with channels that discontinuously lose or gain signal, but the issue seems significantly reduced from browsing over the cube.

tclean(vis=['../uid___A002_Xf73ead_X2326.ms.contsub',` '../uid___A002_Xf73ead_X2639.ms.contsub', '../uid___A002_Xf7519f_X159c.ms.contsub', '../uid___A002_Xf75b8f_X1a8f.ms.contsub'], field='Sgr_A_star', spw=['0', '0', '0', '0'], antenna=['0,1,2,3,4,5,6,7,8,9&', '0,1,2,3,4,5,6,7,8,9&', '0,1,2,3,4,5,6,7,8,9&', '0,1,2,3,4,5,6,7,8,9&'], scan=['7,10,13,16,19', '6,9,12,15,18', '7,10,13,16,19', '7,10,13,16,19'], intent='OBSERVE_TARGET#ON_SOURCE', datacolumn='corrected', imagename='uvcontsub_noisethresh3_cf3', imsize=[270, 270], cell=['2.1arcsec'], phasecenter='ICRS 17:47:12.7761 -028.20.06.419', stokes='I', specmode='cube', nchan=2042, start='87.8853036790GHz', width='0.0305213MHz', outframe='LSRK', perchanweightdensity=True, gridder='mosaic', mosweight=True, usepointing=False, pblimit=0.2, deconvolver='hogbom', restoration=True, restoringbeam='common', pbcor=True, weighting='briggsbwtaper', robust=0.5, npixels=0, niter=999999, threshold='0.525Jy', nsigma=0.0, cyclefactor=3.0, interactive=0, usemask='auto-multithresh', sidelobethreshold=1.25, noisethreshold=3.0, lownoisethreshold=2.0, negativethreshold=0.0, minbeamfrac=0.1, growiterations=75, dogrowprune=True, minpercentchange=1.0, fastnoise=False, restart=True, savemodel='none', calcres=False, calcpsf=False, parallel=True)

This gif shows the impact on the spectra: Presentation1

djeff1887 commented 10 months ago

Went back and looked at the images and residuals for spw22. In general, the new images have lower fluxes than the pipeline versions, and I think that's reflected in the residual too. The residual is typically lower or more negative than the pipeline versions. Although, it's hard to say if there's a strict pattern. The clean looks to be better in some places and worse in others (which maybe isn't too surprising). Orange is new residual, blue is pipeline residual:

(a confetti area, several parsecs northeast of the cores) image

(over Sgr B2M) image

(another confetti area, just north of N) image

(another confetti area, several parsecs to the west of the cores) image

ashleythomasbarnes commented 10 months ago

I'm a little confuse, how come the residual have negative fluxes here? Could you also check the model images for where you previously had a spike and now you don't? It would be interesting to see what the model looks like for both...

keflavich commented 10 months ago

It's normal for the residual to contain negative fluxes - dirty images always do. Ash, why do you want to see the model images? I'm not sure what that tells you - it's obvious from the smaller residuals that there are now (orange) model components where there previously none, right? I would like to see the images now, though - did the speckling / "confetti" go away?

djeff1887 commented 9 months ago

Just following up: here's a quick comparison of the new images vs the old ones. (Old) pipelineconfetti

(New) clean_noconfetti

The speckling is gone, and I have confirmed that the mask and model in several confetti areas are now present where they used the used to drop out: (model) model-northern (mask) mask-northern

ashleythomasbarnes commented 9 months ago

Interesting... Then is isn't super obvious why the flux towards the confetti was higher previously? Is this this region (https://github.com/ACES-CMZ/reduction_ACES/issues/110#issuecomment-1811864545)?

djeff1887 commented 9 months ago

No, that was from a single pixel in a different part of the field, in the southwest. But checking that pixel and the region around it, I see the same behavior. It's easier to compare to my last message by looking at the region: (image) ashregion_image (model) ashregion_model (mask) ashregion_mask (residual) ashregion_residual