change in period evidence

[Robin Leadbeater]

The argument for something to have happened though between 1978 and 1998 is convincing from fig 4 of the 1999 Graczyk et al paper if the error bars are to be believed