Homogenization of Temperature Data: An Assessment

Kevin Cowtan


As a novice to the field of homogenization, I asked experts in the field to review the material for errors. Some of the comments have been incorporated into the final version of the report, as noted below.

Peter Thorne, ISTI

Peter's comments are in this document (PDF). The comments are PDF annotations on an earlier version of the report. To view them you will need an appropriate PDF viewer such as Adobe Acrobat Reader - look for the yellow boxes in the document and click on them. PDF viewers built into modern web browsers and operating systems do not typically have this functionality.

In response to Peter's comments I have made the following changes to the report:

  • The chapter "Testing for bias" was originally at the end of the document, reflecting when the work was done. This created an ambiguity as to whether the tests applied to the GHCN PHA algorithm or my own simpler FHA algorithm. The chapter has been moved to follow the chapter "Evaluating the GHCN adjustments", to clarify that these tests were applied to PHA.
  • Additional references were added to relevent ISTI work and other work on the use of renanalyses in homogenization.
  • Descriptions of the CRU data and GHCN versions clarified.

The remaining comments are also valuable but have not yet been incorporated.