A Multidecadal Climate Signal Propagating Across the Northern Hemisphere through Indices of a Synchronized Network

摘要

Proxy and instrumental records reflect a quasi-cyclic 50-to-80-year climate signal across the Northern Hemisphere. Three studies, the collection of which is presented in this thesis, document evidence, or lack thereof, of this proposed climate signal.In the first study1, chapter two, an eight-member collection of geographically and dynamically diverse twentieth-century climate indices was analyzed with multivariate statistical techniques to assess collective behavior of the network. Emergent from the results was a picture of a climate signal propagating through a sequence of synchronized atmospheric and lagged oceanic circulations across the Northern Hemisphere. Tempo of the signal’s multidecadal variability appears related to that of the low-frequency oscillatory pattern of sea-surface-temperature distribution across the North Atlantic basin, the Atlantic Multidecadal Oscillation (AMO).The third chapter features the second study, the goals of which were two-fold: to gain insights into mechanism of the propagating signal identified in the first study and to probe the signal’s history. Data sets included twentieth-century data and proxy data spanning the interval 1700 to 2000. Findings suggest i) the observed 20th century signal-propagation has existed in somewhat similar fashion for the 300-year length of this study; ii) Eurasian-Arctic Shelf sea-ice plays a strong role in the propagation of the hemispheric climate signal; and iii) dynamics fundamental to generation of the multidecadal component of the Northern Hemisphere’s surface temperature are encoded onto the records of key proxy indices, the combined signatures of which trace the hemispheric circumnavigation of the secularly varying, sequentially propagating climate signal.In the final study in this collection, detailed in chapter four, a network of simulated climate indices, reconstructed from a data set generated by models of the third Coupled Intercomparison Project (CMIP3), were analyzed. Of sixty analyses performed on these networks, none succeeded in reproducing a propagating multidecadal quasi-oscillatory signal. This result, standing in stark contrast to those of the first two studies, may imply that physical mechanisms relevant to signal propagation may be missing from this suite of general circulation models.