Spencer and Christy’s new UAHv6.0 gl TLT dataset is impressively and conspicuously corroborated by:
# HadCRUt3 gl with Jan’98 –0.064K adjustment.
In fact, we can extend this more correct surface dataset by calibrating it with the HadCRUt4 series, constructing what could be dubbed the “Real HadCRUt” series.
“Real HadCRUt” in turn agrees very well with:
# The 3rd generation “Reanalysis Mean” (NASA MERRA+ERA Interim+JRA55), gl T_2m.
“Real HadCRUt”, JMA, “Reanalysis Mean” and UAHv6 tlt:
The agreement between UAHv6 and RSSv3.3 tlt is also pretty striking:
Not to say the least, UAHv6.0 tlt (and tmt) is corroborated by CERES OLR data: