Empirical Orthogonal Function Spectra of Extreme Temperature Variability Decoded From Tree Rings of the Western Himalayas

The ability to distinguish different natural frequency modes from a complex noisy temperature record is essential for a better understanding of the climate response to internal/external forcing. Here we investigate the empirical orthogonal function and spectra of a newly reconstructed tree ring temperature variability record decoded from the western Himalayas for a period spanning 1227 A.D. to 2000 A.D., allowing frequency resolution of interdecadal and interannual oscillatory modes. The spectral analysis of first principal component (PC1) with similar to 61.46% variance reveals the dominance of significant solar cycles notably peaking around 81, 32, 22, and 8-14 years. Although longer solar cycles are dominant and statistically significant at more than 95% confidence level, the average 11 year solar cycle peaking at a period ranging from 8 to 14 years is less significant (not >90%) and might indicate chaotic phenomena. Similar spectral analysis of PC2 (variance 26%) and PC3 (variance 13.05%) reveals interannual oscillations peaking at a period range of 2-8 years, which are probably related to the global aspect of the El Nino-Southern Oscillation phenomena. Our present analysis in the light of the recent ocean-atmospheric model results suggests that even small variation in solar output in conjunction with the atmospheric-ocean system, and other related feedback processes could cause the observed abrupt temperature variability at the time of "criticality" through the triggering mechanism.