edit: the article goes into a little more detail and talks about how its due to the different speeds of light in water and vaccuum, and I don't know enough about semantics or physics to say what this means about my original question :)
For the work in Nuclear Sessions I’ve used a 2017 dataset of radon potential, published by the Spanish Nuclear Safety Council.
This dataset was selected purely on the basis of availability: it was the easiest (and first) geographic dataset regarding radiation to be found.
They could have gone for the full Australian 256 channel raw (or post processed full spectrum or "artifical colour" U-K-Th 3-channel) dataset covering the bulk of the continent.https://www.ga.gov.au/scientific-topics/disciplines/geophysi...
The usual practice is to accumulate one second sample windows of the gamma spectrum while travelling at 70 m/s, normalise the data to remove various wobble factors, and extract three significant channels to form an RGB image - you can always add fiddle layers to indicate radon or out of band features (such as uncommon trace elements from atomic tests).
That said when reading the title I though this might be about the new trend in radiometric sampling - using a scintallation source surround by layered spinning masks - when you get a gamma spark you can better guess to some degree the direction of the source, when many gamma interactions accumulate you can build up a pretty decent 3-D image of gamma sources surrounding your instrument.
One of these was recently trialed in the HN infamous Western Australia Mining Company loses Radiactive Source! stories from earlier this year.
If you want to identify stuff, you'll be looking for an energy spectrum.
If you're after an actual spectrum and you're thinking DIY then a starting point is a thalium doped sodium iodine cyrstal and electronics
https://alphaspectra.com/scintillation-detector-manufacturin...
https://www.alibaba.com/product-detail/2-inches-NaI-Scintill...
Alibaba is listing a single 50mm round x 50mm thick crystal plus electronics at 3K USD each (less than 10) which seems steep to me given years back we used 42 litre crystal packs (with CSIRO grown crystals and in house PhD electronics engineers, etc - so YMMV).
Now your problem is sampling the output levels several thousand times a second, binning the counts, and having a calibration source and compensating for tempreture drift.
You can buy a box that'll do that for you or you can <cough> DIY that with reference to radiometric survey field guides (there's one from AGSO - probably Bob Minty has his name on it).
A good all in one handheld is probably something like https://www.radiationsolutions.ca/wp-content/uploads/2020/03...
which is one of those "contact us and get a quote" jobs from a Canadian company of good repute with Jens Hovgaard, a Finn, as the Tech CEO | President.
He knows his stuff and has an industry algorithm named after him .. although others did similar work elsewhere about the globe.
A relevant spec sheet to compare against what was outlined above is:
https://www.gammadata.se/assets/Uploads/CsITl-and-Na-data-sh...
Almost but not quite. The speed of light is always the speed of light. There's nothing wrong with gamma rays.
It's charged particles such as beta and alpha rays that generate the blue Cherenkov radiation. The writer's link to Wikipedia already reflects that.
Gamma rays always follow the speed of light in the medium that they travel through, because they are light. They cannot exceed the local speed of light, thus they cannot generate Cherenkov radiation. It must be particles with charge and non-zero mass that are capable of generating Cherenkov radiation.
I saw one at the German Museum of Technology (Deutsches Technikmuseum) in Berlin: https://technikmuseum.berlin/en/spectrum/world-of-experiment...
Don't forget Dr Manhattan from Watchmen!