Hacker News Reader

Comment by gcbw2 | Hacker News Reader

The hypothesis is they should see CMB distortion between galaxies, which is hidden by the distortion caused by the galaxies themselves. The solution was to assume they were removing the distortion caused by the galaxy halo, and what was left was the original distortion they were looking for.

sounds like the work to rule out false-positives would be huge. This is putting a lot of weight on a technique that is not fully described in the paper (i might have missed, just glanced at them for now, and i am an amateur that just like to fiddle with similar data).

View on HN

You're right about the first two papers (which used the same approach). Ruling out a FP result would be difficult, which is why it wasn't fully embraced by the scientific community.

The third team took a different approach [1], with results that are both more accurate, less prone to FP, and generally agree with the other findings.

> Here we report observations of two absorbers of highly ionized oxygen (O VII) in the high-signal-to-noise-ratio X-ray spectrum of a quasar at a redshift higher than 0.4. These absorbers show no variability over a two-year timescale and have no associated cold absorption, making the assumption that they originate from the quasar’s intrinsic outflow or the host galaxy’s interstellar medium implausible.

[1] https://www.nature.com/articles/s41586-018-0204-1

I'm not really sure of what you mean by false positives.

After my understanding of the stacking technique as used in other contexts in astrophysics and without going into too much details :

The problem they were facing is that the signal to noise of the images of the filaments was too low to say that they had detected anything in any individual images. However by stacking (adding) images they were able to detect it because the signal grows roughly with N (N being the number of images) and the noise grows with sqrt(N). So by stacking enough images you'll get the signal to noise necessary to say you've detected sth.

Think like that: there are two bright lights in front of you. There will always be a halo between those two bright lights.

Now, my hypothesis is that "between all two bright lights, there is third, dimmer one, hidding". And then i prove it by filtering X from the two bright light halo, and prove that Y is left proving that the third light is there.

Now, how can i be sure Y is really a third dimmer light? and not just noise on the function i used to try to clean up the halo of the two bright light?

You are right that this could be the case, and it's mentioned in at least one of the papers [1] where they say that a better understanding of the physical state of this gas is needed to estimate its contributions to the baryonic mass.

[1] https://arxiv.org/abs/1805.04555

Now we can get to discuss the third paper ;)

I was wondering why stacking N different galaxies was any different than just taking one image and multiplying it by N, but your comment just provided the answer: noise.

There's more confidence in the result because two teams using very different methods have arrived at the same conclusion.

The three papers used the same 'filtering' methodology, to keep the simplification analogy.

Reply on news.ycombinator.com