When the CRU at East Anglia disclosed that it had lost some of the raw temperature data, leaving only the “homogenized” data, some honest commentators expressed the hope that the homogenizing was competently done.
Anyone who has been following Climate Audit for the last few years knows that at least some of the adjustments to the raw data done by the major data depositories appear to have been incompetently done at best. The statistical techniques used in the scientific backwater of historical climatology are often ad hoc, bearing little relation to the techniques that are standard in other fields. In particular, their techniques for handling missing data are particularly unscientific.
Perhaps the most accessible blog post demonstrating the effects of homogenization adjustments on a set of temperature records is by Willis Eschenbach at Watts Up With That.
Here's the raw data:
Here's the data after the "scientists" pushing global warming get done with it:
If I did this with my portfolio performance reports, I would be in jail. This is Bernie Madoff stuff in the science arena.