Using the Southern Oceans Baseline

The chart above uses the Neukum et al. Southern South America temperature reconstruction archived on NASA Paleo and is spliced with GISS LOTI  regional with the full common 1880 to 1995 period of the reconstruction and the instrumental data.  With this slice, the mean of the instrumental is approximately 0.1 c above the mean of the full SSA reconstruction.

The full reconstruction is shown above.  Note that the equatorial and southern extra tropics match the reconstruction pretty well and there is nothing unprecedented in those data sets.  The Northern extra tropics are above any point in the reconstruction.

To compare the Northern Hemisphere sensitivity this chart includes the Jacoby et al. Taymyr Peninsular temperature reconstruction.  Note the mean value of the Taymyr reconstruction is nearly 0.4 C below the mean of the SSA reconstruction. That appears to be consistent with the approximate 0.4C drop in circa 1400 AD.

In 1816, "the year without summer"  a volcano caused considerable cooling in the Northern Hemisphere.

Isolating the time frame, it appears the volcano erupted several years before 1816 or there was some other cause for decline in the temperatures.  Note the sharp drop in the Taymyr reconstruction and the much more gradual decline in the SSA reconstruction.  While the data is too coarse, there is some indication of a recovery with over shoot and settling indicating a dampened response to the perturbation.  That seems to support the idea that the instrumental record started ~0.3 C below the normal for the past 1100 years.

The Tambora timing does add up to the data though, from the Smithsonian Institution Volcanic Project, there were a number of small Northern hemisphere volcanoes that could have aided in the decline.

There are still plenty of questions, but the Southern Oceans baseline seems to have a number of advantages.

This plot adds the Kobashi Greenland reconstruction which starts in 2000 BC.  The mean value lines for all of the reconstructions and data series are included to show the range of "average" conditions is dependent on choice of baseline.  For this collection of climate reconstructions, the mean of the Northern hemisphere is ~0.25C lower than the mean of the Southern hemisphere.  One interesting feature is the 1600 to 1900 period where the Greenland recostruction is out of phase with both the SSA and Taymyr reconstructions.  That may be due to a major imbalance of the North Atlantic and the rest of the oceans.  The Sargasso Sea temperature reconstruction of Keigwin also indicates a similar event with a range of Sargasso Sea temperatures of ~22C to ~25C degrees or +/-~1.5C range of temperatures similar to ranges in the reconstructions used above.