In an automatic weather station, a temperature-measuring electronic sensor (which is typically 1,5 m in the air above a grassy ground and covered from direct sunlight) is in a certain way coupled to the computer inside the station and enables to show the temperature on a computer screen to one decimal place (such as 21,7 C). I understand that „the refresh rate” of the temperature value shown is each 10 seconds (on a Vaisala sensor) or more. The computer memorizes all these values and could in principle extract out of them the maximum and minimum air temperature value over a timespan of 12 hrs, which is an important climatological parameter.
Question to PSE community:
I know about the so-called thermoelectric effects, Joule, Peltier and Seebeck, but not in detail, of course. They were not in my HS or university curriculum. My question is how does the heat transfer air - metallic sensor gets translated into a micro electric current (?) and which formula is used to determine the temperature in C/K/F from the current in A or mA or some other measured quantity.
P.S. There are temperature sensors everywhere, in a smartphone, on a PC motherboard, on the bottom of lateral mirrors of cars, on the wings of an airplane, etc. I assume all use the same physical principle (which is that exactly and what mathematical formulas does it use?).
P.P.S (Just for the fun of reading the introduction above). Automatic weather stations accross US (administrated by NWS which is part of NOOA) also determine the temperature in F to one decimal place, but for some unknown reason NWS reports and stores only the integer values in the database. As of such, a certain accuracy compared to measuring temperature in C is lost. For example, the hottest temperature ever in Phoenix AZ would be 118 F which is rounded to 47.8 C. A new temperature record in Phoenix according to NWS would be 119 F, that is 48.3 C, while if the temperature were measured the European way, a new temperature record in Phoenix could also be merely 0,1 C more, that is 47.9 C.
