Analysis and reconstruction of the relationship between a circulation anomaly feature and tree rings: Linear and nonlinear approaches

Abstract

Tree rings can be reliable recorders of past weather and climate variations. Tree rings from mountain regions can be linked to upper air atmospheric sounding observations and large-scale atmospheric circulation patterns. A "synoptic dendroclimatology" approach is used to define the relationship between tree rings and a specific upper air anomaly feature that affects climate in the western US. I have also reconstructed this anomaly feature using both regression and fuzzy logic approaches. Correlation analysis between 500 mb geopotential heights and tree rings at a site near Eagle, Colorado reveals an important anomaly centered over the western US. This center can be viewed as a circulation anomaly center index (CACI) that can quantitatively represent the relationship between atmospheric circulation and tree growth variations. To reconstruct this index from tree rings, I used both a multiple linear regression (MLR) and a fuzzy-rule-based (FRB) model. The fuzzy-rule-based model provides a simple structural approach to capture nonlinear relationships between tree rings and circulation. The reconstructing capability of both models is validated directly from an independent data set. Results show that the fuzzy-rule-based model performs better in terms of calibration and verification statistics than the multiple linear regression model. The reconstructed anomaly index can provide a long-term temporal context for evaluation of circulation variability and how it is linked to both climate and tree rings.