2
$\begingroup$

Dr. Becky's Feb. 11, 2022 video Move over JWST! 5 new telescopes to get excited for makes a quick mention of this image produced by the Meerkat array and published in Heywood et al. 2022 The 1.28 GHz MeerKAT Galactic Center Mosaic (also arXiv).

The caption of the mosaic begins:

Figure 1. The full MeerKAT total intensity mosaic, covering 6.5 square degrees of the Galactic center region at an angular resolution of 4". This is a linear mosaic formed from the 20 pointings described in Section 2, the centers of which are showning the figure as “+” markers. The image has dual color schemes, with a linear greyscale covering the faint end, and the heat map covering the bright end with a square-root stretch function. Pixel scales for the two colormaps are inset...

I noticed that the "linear greyscale covering the faint end" is centered on zero, and ranges from -0.4 to +0.4 mJy/beam.

Question: How do we get -0.4 mJy/beam? What does negative power mean in this context?

related:


Screenshot from the arXiv preprint:

Figure 1. The full MeerKAT total intensity mosaic, covering 6.5 square degrees of the Galactic center region at an angular resolution of 4". This is a linear mosaic formed from the 20 pointings described in Section 2, the centers of which are showning the figure as “+” markers. The image has dual color schemes, with a linear greyscale covering the faint end, and the heat map covering the bright end with a square-root stretch function. Pixel scales for the two colormaps are inset. Heywood et al. 2022 https://arxiv.org/abs/2201.10541

$\endgroup$
2
  • 1
    $\begingroup$ In single-dish observations, this typically results from emission in the OFF position, subtracted from the ON. These are interferometric data, so I think along similar lines that negative flux results from imperfect deconvolution and absolute flux calibration. A very minor imperfection given that the deviation is 1e-4 from the peak flux. $\endgroup$ Sep 7, 2023 at 8:44
  • $\begingroup$ @MarkusRoellig sure that certainly sounds plausible - if you feel it rises to the level of an answer please feel free to post as such, thanks! It just feels that they've chosen the mixed intensity scales to highlight the -0.4 to +0.4 range (symmetric about zero) and make the swings back and forth across zero particularly striking. (small nit, the +/-0.4 deviations are 1e-3 of the peak flux of 350 if I'm understanding it correctly) $\endgroup$
    – uhoh
    Sep 7, 2023 at 10:08

1 Answer 1

1
+100
$\begingroup$

Astronomical radio emission is typically much weaker compared to ambient (terrestrial) emission, e.g., emission by the atmosphere, the telescope, the surroundings, and the receiver, so one needs to clean the measured signal from these contributions. (Fun fact: a typical radio dish even collects some emission from behind the telescope.)

In single-dish observations, this is done by observing a position as close to the target source as possible and subtracting this OFF emission from the ON emission. Assuming that both lines of sight are similarly affected by all local error contributions, this leaves the source's emission. This requires the OFF to be emission-free, which is surprisingly hard depending on the observed frequency. Imagine you want to observe molecular emission from a giant molecular cloud. Molecules form once they are sufficiently shielded from harmful UV radiation, so pointing away from the cloud should do the trick unless you accidentally hit a different cloud that happens to be in the foreground or background. (There are many interstellar clouds in the Milky Way. If you choose an OFF that is too far away, you may end up with a line-of-sight that passes different atmospheric (terrestrial) volume elements with slightly different emissions. This will lead to a worse ON-OFF correction.

The MeerKAT data is from an interferometric observation which do not rely on ON-OFF but gets rid of all contaminations by deconvolution followed by an absolute flux calibration using a particular calibration target PKS B1934-638 (Details: Heywood et al. 2020 ).

All these data reduction procedures are imperfect, given that any signal sufferers from systematic (and stochastic) noise. Therefore, it is not uncommon that the reduced and calibrated date shows some (low) negative flux values. In this case, we are talking about 1e-3 relative to the peak flux. In other words, -0.4 mJy/beam is still zero emission.

For complex gain calibration at the NRAO see here: https://science.nrao.edu/facilities/vla/docs/manuals/obsguide/calibration and this PDF. If you are interested in single-dish fulx calibration there is a manual from the IRAM 30-m. There is a nice description of all steps from the Lab Course M at I. Physics Institute of the Universität zu Köln: Manual

$\endgroup$
2
  • 1
    $\begingroup$ Thanks for a speedy answer! I wonder if "stochastic noise" is the culprit, or if "systematic errors" might be a better label for such smooth, larger areas of (slightly) negative power. For example maybe the precipitable water vapor wasn't corrected for sufficiently, or gains were drifting, etc. Here's another, possibly of interest to you, and maybe you can add something there also Why did it take five years to "figure out" how to use astrometric calibration sources to deblur LOFAR images? $\endgroup$
    – uhoh
    Sep 7, 2023 at 12:19
  • 2
    $\begingroup$ @uhoh Good point. If we closely look at the image we note, that the regions with negative flux are in close vicinity to the brightest parts of the map which hints at a systematic bias. Instrument and receiver drift is another difficulty of course. $\endgroup$ Sep 7, 2023 at 13:35

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .