plough

I think it is the other way around: SD is calculated from 1982-2020, while all measurement readings in the plot are 1982-2026. I believe this is meant to not introduce an unwanted shift but compare to sort of a 'stable process'. However, that should have been described and argued somewhere.