Northern Hemisphere Temperature Reconstruction Clearly Shows the Medieval Warm Period and Little Ice Age (Plus a Whole Lot More)

The pattern of temperature change over the past 2000 years is important for several reasons. First, if there has been no long-term pattern to it, then any warming in recent decades can more easily be attributed to human activity. On the other hand, if there have been natural cycles or centennial-scale excursions of temperature, then it is more difficult to claim that recent warming is unnatural or unprecedented. Second, impacts of climate change are often based on the assumption that "unusual" warmth is harmful to life and human society; but if it was equally warm, or warmer than today, a thousand years ago, then there is much less basis for predicting harm.

The study of Ljungqvist (2010) expands the data available for reconstructing past climate by utilizing 30 datasets. However, the author felt that data from arid zone tree species should be disqualified, due to the confounding of the effects of warm temperatures with those of drought stress. Thus, bristlecone pine, foxtail pine, and Mongolia data that show a pronounced (and likely spurious) hockey-stick shape were not used.

The CPS (Composite Plus Scale) method of analysis was employed; and the data were linearly interpolated. For each dataset, the data were normalized to zero mean and unit variance. The 30 data sets were combined by computing their mean. The composite score was shown to have a correlation of 0.95 with the Hadley instrumental temperature record for the zone 90-30°N and was scaled to match the variance of the Hadley data.

While the CPS method has the disadvantage that certain proxies which might not relate to local temperature are combined with others and correlated with the hemispheric temperature (vs. doing strictly local calibration), it has the advantage that individual proxies are given equal weights, so that one is not selectively picking out those with the "right" shape by chance and giving them too much influence.

The resulting reconstruction has values at each decade. It shows a clear repeating warm/cool pattern of Roman Warm Period, Dark Ages Cold Period, Medieval Warm Period, Little Ice Age and recent warming. This long-term pattern was stated to relate to the 1500-year Bond cycles. Looking at the reconstruction itself, the RWP and MWP were both warmer than the decades from 1961-1990; and the composite proxies do not show rapid warming after 1990. Thus, the study clearly shows that large temperature excursions on centennial scales are real features of the past 2000 years, and that recent warmth is not unprecedented.