I'd like to better understand how a global temperature anomaly is calculated from 1)measured data and 2) models and I haven't been able to find this info online so am asking here.
1) Given that there are finite number of measurements of temperature on land and sea, how is the anomaly calculated? For example, is it calculated in each grid cell and then those anomalies averaged? Does it only use measured data or are interpolated data (because there are no measuring stations in those remote locations) used?
What is the uncertainty in the temperature anomalies calculated from data? For example, how different is the anomaly if one uses or doesn't used interpolated values? I've rarely seen the uncertainty provided in the temperature anomaly so was curious how large it was.
2) What is the methodology used to calculate the temperature anomaly from models? Does it use model values at the same locations where there are measuring stations or does it calculated the anomaly per grid cell and average? How does it handle altitude differences in grid cells?
Thank you.