The Berkeley Earth Surface Temperature (BEST) project has updated and is still updating its site. A discussion of the pluses and minuses of the revised products is going on at Climate Etc. among other places and has finally brought to the forefront the issue of Land Amplification. The minions of the Great and Powerful Carbon tend to miss the subtle issues of a mainly land located in the northern hemisphere, "Global" temperature series and what drives the temperatures of the majority of the land in that land only data set, the northern oceans. So in yet another soon to fail attempt I have plotted the Reynolds Oiv2 SST for the Northern Extra Tropical oceans from 24N to 70N latitude. This time though I thought showing the BEST Tmax and Tmin instead of Tave might be better.
Since the Oiv2 data only starts in November of 1981 and has seasonal variation that needs to be removed to see the ~0.4 C shift starting in 1985 thanks to volcanic forcing plus a Pacific Decadal Oscillation (PDO) mode swing, I used the monthly averages for the 31 full years common to the data sets for the baseline and seasonal cycle removal. That should be pretty clear to anyone that halfway follows the surface temperature debate. BEST, being predominately northern hemisphere data has a strong northern hemisphere seasonal cycle. While my approach may be less than ideal, at least there are 31 years and I explained what I did.
Pretty much anyone, other than the minions of the Great and Powerful Carbon, should notice that there is a strong correlation between the satellite measured northern extra tropical oceans and the BEST "Global" land surface temperature data. Most would notice that the correlation is better with the Tmin than the Tmax. Most would notice that the northern extra tropical oceans are curving back toward the mean prior to both the BEST Tmin and Tmax.
The minions of the Great and Powerful Carbon manage to have the land data wagging the sea dog tail when the opposite would appear to be the case. As I have pointed out before, Land Amplification should be considered and that because of land amplification, the northern Atlantic region (AMO) makes a damn good proxy for "Global" land and ocean "surface" temperatures. That means that the Central England Temperature record makes a pretty good proxy for "Global" "surface" temperature and since the Indo-Pacific Warm Pool correlates well with the longer term average north Atlantic SST, the Indo-Pacific Warm Pool makes another pretty damn good proxy for "Global" "Surface" temperature anomaly. All because of land amplification.
Figure 53. This figure shows the amplification ratio between 30N and 60N (roughly from the middle of Texas to the bottom tip of Greenland) compared to the amplification ratio in the Arctic. The observed value from Berkeley Earth is shown in red and the GCMs are in blue. The GCMs tend to overestimate the amplification at the pole and underestimate the amplification in the mid-latitudes.
Remarkable isn't it?
While we are at it let's consider some baseline dependence issues.
this site. This time seasonality was not removed which may be a minor issue.
With this baseline, the BEST Tmin starts at the lowest initial condition with Tmax more closely tracking the northern extra tropical SST. The "Global" diurnal temperature trend would have decreased until circa 1980 then started increasing roughly around 1985. The majority of the warming was due to Tmin and from the start of the data until about 1960 then more uniform warming took place. Just a different view of the same old data.