SPATIAL AND TEMPORAL MODES OF VARIABILITY IN ARCTIC SUMMER TEMPERATURE DURING THE PAST 500 YEARS

Spatial arrays of high-resolution (annual-decadal) paleoclimate records from throughout the Arctic can be used to distinguish different modes of variability and trace their behavior back in time. Previous compilations of primarily annual-resolution records from varved lake sediments, tree rings, ice cores, and marine sediments provided a view of average Arctic summer temperature documenting dramatic 20th-century warming that ended the Little Ice Age in the Arctic and caused dramatic retreats of glaciers, melting of permafrost and sea-ice, and alteration of terrestrial ecosystems. Some evidence suggests that these changes may be linked to a rising trend in the Arctic Oscillation (AO), and that the positive trend in the AO itself may theoretically be due to greenhouse warming. Unfortunately, combining records into a single Arctic average does not exploit the valuable spatial information in the Arctic-wide array, and can not shed light on past AO behavior. A new international collaboration has created a high-resolution spatial array of Arctic paleotemperature records for the past ~500 years. Annually resolved archives were used wherever possible (e.g., tree rings, varved lake sediments, and annual ice layers), but sub-decadal resolution records from ice cores and high deposition-rate lake sediments were included as well.

Empirical Orthogonal Function (EOF) analysis was used to characterize the spatial and temporal modes of variability contained in the proxy array. The leading modes of proxy variability all have highly significant correlations to leading modes identified in NCEP-NCAR reanalysis data, and thus are likely associated with dynamically significant processes, including: 1) a circum-Arctic temperature trend with rapid 20th-century warming; 2) the Arctic Oscillation; and 3) a Urals Trough wave number three circulation pattern. Our analyses demonstrate the ability to identify the major modern observed modes of Arctic SAT variability within an array of proxy data, and indicate the feasibility of of reconstructing these modes back in time. Analysis of this compilation of high-resolution Arctic proxy data will provide insight into the long-term natural background variability of the AO, as well as other dynamic systems, and place observed recent positve trends into a pre-anthropogenic context.