DMI averages weirdly by latitude, Zach averages properly by area. I guess the differences are not that large when the temperatures are nearly uniform between the Pole and the 80 line, but I'm sure there should be a difference somewhere. Or I have misunderstood the weighting methods.