I was thinking about the Carnot Theorem, and thinking how that applies to a car's engine. T_hot would be the temperature of combustion for the fuel/air mixture, which, once the car's engine has been operating for a few minutes under load, should reach a stable maximum. I think that T_cold would end up being a function of the temperature of the outside air in the environment around the vehicle (the temperature of the engine coolant would be a function of how cold it can get as it flows through the radiator, and cannot get colder than the air flowing over the radiator; it seems reasonable [though I may be wrong] to conclude that the colder the outside air, the colder the engine coolant will get). So, T_hot seems like it would be, probably, about the same in the summer in Mexico, as in the Winter in Alaska, while T_cold could go down by a pretty decent chunk in a place in the extreme North, where temperatures might get down to -45 Celsius. I realize that for Carnot's Theorem, you use temperatures on an absolute scale (that is, Kelvin), which is still 227.6 Kelvin, so it's not a huge jump, but seems like it might be a measurable difference? Are there any practical complicating factors which might negate the expected increase in efficiency?