Documenting and understanding the past behavior of the climate system provides a critical reference for comparison with future potential climate change. Since the record for instrumental measurements covers only the most recent 100-150 years, indirect ("proxy") measurements must be considered. Quantitative analysis of these proxy data is challenged by the fact that the proxies apply to specific locations, and their relationship to the main feature in the climate system are not well understood.
With this project, researchers from the University of Washington build on recent progress in using state-of-the-art tools from the weather prediction community to reconstruct past climate states by a method referred to as paleoclimate data assimilation (PDA). The research involves optimally blending information from climate models and proxy data to reconstruct Earth's climate over a wide range of time scales. PDA provides a transformative synthesis of irregularly spaced pointwise data into global gridded data, similar to reanalysis products derived from weather observations. Reconstructed climate states will be used for hypothesis testing using numerical models to evaluate climate sensitivity and predictability on decadal and longer timescales with robust sample sizes over a wide range of climate states.
This research leverages expertise from both the paleoclimate and climate dynamics communities. Further interdisciplinary collaboration will take place through interactions with the PAGES (PAst Global changES) community. Funding supports education and training of a graduate student, as well as outreach activities to schools, community organizations and state agencies.