Well, we need more information on how you got to that state. If I execute the cell in that notebook the output uses Celsius, i.e. no conversion. I don’t know how the original output was generated, but I cannot reproduce any implicit conversions with what you’ve provided so far.
I tried using Ctrl+= as well as Quantity. I get no implicit conversions. And again, if I evaluate your cells in place, I do not get the outputs you had in there originally.
I also notice that your outputs are actually wrong in the sense that if a conversion had occurred, it’s faulty (which seems even worse than implicit conversion itself). Specifically, 80 degrees celcius is closer to 350 kelvins than to the over 1100 kelvins that appears in your notebook.
So, my first suspicion is that there were some previous calculations that you haven’t shown us that set up some variables or functions that are being applied in these expressions. (You could test this by killing and restarting the kernel.)
If that’s not it, then maybe there is some setting/preference that is set differently in your environment than in mine. Maybe there are options for automatically converting certain units. Seems unlikely, but I’m just brainstorming here.
It makes no sense to multiply a Celsius temperature by a coefficient. Version 13.2 has introduced temperature differences, and the behaviour in such situations has changed. The unexpected output happens on MacOS 13.1 with just this input:
Good point. I was so focused on not being able to reproduce the behavior that I didn’t even think about the semantics. I’ve even run into this before and wondered why sometimes it seems like Quantity units are treated like arbitrary tags. It’s almost like you could do Quantity[4, abc] and have subsequent operations just treat abc like a formal variable or something. But anyway, kind of nice that MMA does “the right thing” in the update.