It is worth noting that method for producing the data 1900-2010 data and the 0-1900 data in the top graph is wildly different.
First and foremost: The black line is an actual temperature measurement, while all the other lines are estimates based on proxies. Except of course: Secondly all the proxies appears to be correlated to the black line, because they have been chosen to do so. That is, the correlation to the black line is manufactured, and not a finding.
The rest of the graph is made in a completely different manner, and the proxies are allowed to carry the signal for themselves. I think it is very interesting that you can see a small medieval warming period from the proxy data, but nevertheless one shouldn't be too eager to compare the right part and the left part of the graph, as the methods used are actually not the same.
One can also show that the noise in the proxies will tend to pull the data towards the average and away from the extremes, which is exactly the lower rate of change and lower peaks you see. So that graph is really not very good, at all...
First and foremost: The black line is an actual temperature measurement, while all the other lines are estimates based on proxies. Except of course: Secondly all the proxies appears to be correlated to the black line, because they have been chosen to do so. That is, the correlation to the black line is manufactured, and not a finding.
The rest of the graph is made in a completely different manner, and the proxies are allowed to carry the signal for themselves. I think it is very interesting that you can see a small medieval warming period from the proxy data, but nevertheless one shouldn't be too eager to compare the right part and the left part of the graph, as the methods used are actually not the same.
One can also show that the noise in the proxies will tend to pull the data towards the average and away from the extremes, which is exactly the lower rate of change and lower peaks you see. So that graph is really not very good, at all...