CMG Collaborative Research: Development of Bayesian Hierarchical Models to Reconstruct Climate Over the Past Millenium

点击翻译按钮获取中文摘要

The hockey stick controversy is a scientific dispute over reconstructed estimates of the mean surface temperature changes experienced in the Northern Hemisphere over the past one thousand years. This project is a collaboration between a team of experts in statistics, scientific computation, paleoclimate proxies, temperature field reconstruction, and climate modeling to better evaluate the uncertainty of past temperature reconstructions, as exemplified by the hockey stick graph (which shows an arguably flat pre-20th century followed by a strong warming trend over the past 100years). The socio- political significance of the controversy over the hockey stick graph centers on its use as part of the evidence for unambiguous, anthropogenically caused, global warming. Several aspects of the broader community dispute centered on criticism and defense of both the statistical methodologies and the validity of data sets used in creating the graph. This work will seek to incorporate several proxy climate records, which are known to have differing temporal characteristics and uncertainties, into paleotemperature reconstructions by combining their information using Bayesian Hierarchical Models (BHM). A BHM framework is favorable in isolating observational errors in proxy data (e.g. tree ring widths) from the underlying physical processes of interest (e.g. past surface temperature) thus improving climate reconstructions of the last millennium.