These scales are both arbitrary. Celsius is super useful if you are concerned with phase transitions of water in the presence of a gas at 101,325 Pa of pressur

These scales are both arbitrary. Celsius is super useful if you are concerned with phase transitions of water in the presence of a gas at 101,325 Pa of pressure - it's hard to do much better, really. But what if we wanted to design a scale for measuring the typical range of air temperatures experienced outside in the temperate climates on Earth? You said you are from northwest Germany, which is a fairly temperate climate - to pick an arbitrary city there, the record low in Bremen is -23.6 C and the record high is 37.6 C. So perhaps for our hypothetical scale for Bremen (let's label its units B) a good range would be to set 0 B to -23.6 C and 100 B to 37.6 C to capture this range. In this scale 0 C would end up being 38.6 B and 100 C would end up being 202 B... values that are curiously close to an existing temperature scale. The B scale would be useful for talking about the outdoor air temperature in Bremen. Would the B temperature scale be useful for talking about atmospheric temperature on Mars? Not any more or less so than C... there's no real inherit advantage to either. The choice you make would likely be based on what your primary audience is more familiar with. If you wanted to force your readers to read a unit that would somehow promote scientific literacy you'd probably go with the Kelvin scale since 0 K is not arbitrary (although the unit magnitude still is).