A Modest Cloud Cover Study (part 9)

Today I resort to standard scores (Z Scores) in order to get an overview of discrepancies between ICOADS, CRU and CLARA datasets

Contradictions, contradictions and contradictions! I get rather frustrated when different datasets tell different stories… these are supposed to be the same clouds over the same region of the UK after all, innit? When this kind of thing happens I always fancy going for the big picture but the big picture is thwarted by the fact we’ve got the CRU data in units of cloud fraction and the ICOADS data in those weird old okta values. A trick statisticians often use at this point is to resort to deriving Z-scores…

Now that’s a tune!

Let me get a coffee and roll my sleeves up…

The standard score, a.k.a. Z Score is a cunning method of turning disparate measurements into a set of unified scale score values. Instead of okta we end up with a score value; instead of cloud fraction we end up with a score value. We can then plot both on the very same graph and start cogitating on similarities and differences.

Now, the trick here is to start out with a common time period for all datasets otherwise the scores are going to get bent out of shape. If we are to compare the ICOADS series spanning 1855 – 2024 with the CRU series spanning 1901 – 2023, then the common time period we must use is 1901 – 2023. So let me press the button, take a slurp of Grumpy Mule and see what this fiddling gives: