Ensemble empirical mode decomposition for tree-ring climate reconstructions

A novel data adaptive method named ensemble empirical mode decomposition (EEMD) was used to reconstruct past temperature and precipitation variability in two 2,328- and 1,837-year tree-ring chronologies from the Dulan region, northeastern Qinghai-Tibetan Plateau. Our results show that EEMD can be used to extract low-frequency signals from the Dulan tree-ring data. The extracted low-frequency temperature trends in the two chronologies correlate significantly with Northern Hemisphere temperatures over the past two millennia. In addition, the newly reconstructed precipitation data have a higher standard deviation than that of data reconstructed with the conventional ordinary least squares and variance matching methods and yield the best amplitude match to the instrumental data. This study shows that EEMD is a powerful tool for extracting the full spectrum of climate information in tree-ring chronologies.