What was done
The authors used three well-dated Sabina Przewalskii ring-width chronologies derived from a total of 77 trees growing in three locations near Dulan, China, on the northeastern Tibetan Plateau (36.0-36.3°N, 98.2-98.6°E), to reconstruct annual precipitation variations in that region over the period AD 850-2002, after which they compared the results with instrumental temperature data for the same region over the period of temporal overlap.

What was learned
Working with 10-year moving averages, the thirteen scientists found that precipitation and temperature were "significantly correlated with r = 0.85 (p<0.0001), after the precipitation lagged temperature for 2 years." Hence, they went on to produce a 40-year moving average curve that was "significantly correlated with seven temperature curves of the Northern Hemisphere," which led them to conclude that their 40-year smoothed reconstruction "could be regarded as the millenary temperature curve for the northeastern Tibetan Plateau."

What it means
In viewing Liu et al.'s final "millenary temperature curve," it can readily be seen that the 40-year-averaged temperature proxies in the vicinity of AD 915 are definitely greater than those at the end of the 20th century, which comprise the next highest peak of the record. Hence, this study represents another specific instance where peak temperatures of the Medieval Warm Period likely were greater than the peak temperatures of the 20th century. And as we continue to accumulate ever more examples of such circumstances in the Medieval Warm Period Record of the Week section of our website, it is becoming ever more clear there is no compelling reason to believe that 20th-century global warming was in any way dependent upon the 20th-century increase in atmospheric CO2 concentration, which was far, far less a thousand or more years ago, when it was even warmer than it is today.