The image below is this one, found in this group of Hubble space telescope images of the interstellar comet 2I/Borisov (C/2019 Q4).

Can anyone recognize this pair of artifacts in the upper right corner of this HST diagnostic image?

enter image description here


2 Answers 2


I'm going to guess that these are reflection artifacts of one or two nearby bright stars. The fact that they have about the same length as the stellar streaks, suggests they are smeared by the telescope's tracking of the comet. The fact that they are oriented with a slightly different angle than the star streaks may be due to the reflection shifting slightly as the telescope is moved.

In fact, Figure 6.9 on this page from the WFC3 Handbook (the instrument used for the comet images) shows a clear examples of "figure-eight shaped ghosts" due to bright stars in the lower-right quadrant, described thus:

When a point source is positioned in the lower right quadrant of the UVIS detector, out-of-focus reflections between the CCD and windows appear along a diagonal from the source towards the upper left, well removed from the source. These figure-eight shaped ghosts gradually move outside the field of view as the target moves out of the lower right corner. They contain a few percent of the flux of the target. The ghosts of several bright stars are visible in the exposure of 47 Tuc shown in Figure 6.9.

Figure 6.9 from HST WFC3 Instrument Handbook Figure 6.9 -- note the clear example paired elliptical rings ("figure-of-eight shaped ghosts") on the left part of the upper half of the image.

(Note that the comet images were taken with a sub-array of the WFC3 UVIS detector, which is probably why you don't see the star that produced the ghosts.)

You see ring-like ghost images in ground-based images, too; I've been examining an image taken by some collaborators with a telescope in the Bavarian Alps, and it's got several examples (fortunately not on top of the galaxy we're interested in!).

  • $\begingroup$ Oh! Since this image tracked stars rather than the comet, the artifacts have not moved as the ones shown in the question did, and so their donut-ness (annular aperture out-of-focus) is more obvious, and now I can notice the central obstruction in the question's image's artifacts as well. Thanks! $\endgroup$
    – uhoh
    Oct 17, 2019 at 22:17

I suspect they are brightish extended galaxies being smeared by the motion as HST tracks on the comet - you can see the star streaks e.g. in the bottom center of the image. Given the time of the image, it should be possible to identify which galaxies they were. Pretty much everything else in the image (apart from the comet) are cosmic rays - being above the atmosphere hugely increases the cosmic ray flux in the CCD detectors compared to ground-based observations. The cosmic rays section of the WFC3 Instrument handbook says:

between 5 and 9% of WFC3 CCD pixels will be affected in a 1800s exposure in SAA-free orbits

(SAA is the South Atlantic Anomaly which normally prevents observations)

Observations with HST are normally split into pairs (e.g. Exposure Time section of the WFC3 Instrument Handbook) to help deal with by allowing e.g. median combining of the images to remove the cosmic rays.

  • $\begingroup$ If you look at the star streaks, they point in a slightly different direction; you can see a star streak just to the left of the two features. So that's not it. $\endgroup$ Oct 17, 2019 at 19:02
  • $\begingroup$ The star streaks could well be in slightly different directions due to distortion in the WFC3 optics $\endgroup$ Oct 17, 2019 at 20:25
  • $\begingroup$ There is a star streak in the upper right of the image, right next to -- partly overlapping with -- the leftmost of the two features uhoh is asking about. That's nothing to do with WFC3 distortions (which change gradually across the field of view). And you can see the star streak at the bottom image has the same orientation as the brighter one. $\endgroup$ Oct 17, 2019 at 20:45
  • 2
    $\begingroup$ Also, there's the fact this is an image made by undoing the distortions. (That's why it has that funny parallelogram shape.) $\endgroup$ Oct 17, 2019 at 20:54

You must log in to answer this question.

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