Amidst a number of verifications of already well - established results on the fidelity of the
surface station trends, they also claim to have discovered something new.
Interestingly, if you pre-combine the Byrd data like Eric does and perform the reconstruction exactly like S09, the resulting infilled
ground station trend at Byrd is 0.13 Deg C / Decade (fairly close to our gridded result).
The
Byrd station trend used to produce this result, however, clocks in at a rather respectable 0.75 Deg C / decade.
The S09 gridded result, however, is 0.25 Deg C / Decade — or almost double the ground
station trend from their own RegEM infilling, and closer to our gridded result than to theirs.
By the way, Eric also fails to note that the kgnd = 5 and 6 Peninsula trends, when compared to the
corresponding station trends, are outside the 95 % CIs for the stations.
I also note that the TTLS reconstructions do a poorer job of capturing the Byrd
ground station trend than the iRidge reconstructions, which is the opposite behavior suggested by Eric (and this was noted during the review process as well).
If
the station trends are plotted on a map (Figure 3), two things stand out.
The NNR yields a relatively uniform and statistically significant trend in this region, which is statistically similar to two of the four
station trends.
The facts are that the surface
station trends are used by GISS GCM as an input.