Please explain to us all how homogenisation does not invalidate any data so homogenised, given doing so means the error confidence levels of the data will be breached in doing so.
For example, Albury temp changed by 15 degrees ( hint less than 1 degree change needed only to breach the error confidence limits ).