There is in fact no photograph of treetops glowing. There is a digital UV-wavelength video of the corona, and a visible-wavelength video of the trees. The paper [1] contains a sole picture with tiny circles indicating where the UV-video detected corona events, overlaid over a frame of the visible-wavelength video. The paper does also contain a video [2] which overlays a somewhat processed version of the UV video over…
Sorry, in what way is this not a photograph? Are you saying that a video is not a sequence of photographs, that UV photons captured by a sensor don’t count because human retina sensitivity is low in that range, or some hopefully-less-semantic argument?
At least personally I scanned the article for it and only found the picture at the top, which I was then frustrated to learn that's just a lab photo, and I came here wondering where the actual image is of it in the field so I found OPs comment helpful to indicate that the suggestion there would be a beautiful picture of glowing canopy somewhere is basically a result of editorializing.