Since 1998, Illinois fabricated and non-fabricated adjusted temperatures have been diverging at a rate of 23 degrees per century.
Zeke Hausfather says that all of this disappears when you use anomalies and gridding, and that I am really stupid.
Gridding over an area the size of Illinois would have minimal effect, because the state isn’t much larger than a grid cell. The only way this could happen is :
- The algorithm being used to generate the manufactured temperatures is incorrect. Or
- There has been massive selective loss of warmer temperature records, requiring filling in of much warmer temperatures than those that are measured.
My methodology assumes a Monte Carlo distribution of error, i.e. that the missing data is randomly distributed geographically and by month of year. This is the normal way which scientists work with large data sets. Climate scientists mess with the data and introduce nonsensical biases into their results.