The focus of this talk is the comparison of hydroclimate variability from the middle Holocene to present in paleoclimate data to that from state-of-the-art climate models. We highlight three different model-(paleo)data comparisons, each with a distinct motivation and approach, in part to deal with the idiosyncrasies of the associated paleoclimate data. In the first, we use high-resolution spatial data for drought over the last millennium to validate the simulation of centennial timescale hydroclimate variability across the Northern Hemisphere. We find that while climate models can simulate realistic variability on these timescales, there are inter-model and -regional differences that must be considered. In the second, we use a single low-resolution record from a sinkhole lake in the Caribbean that is interpreted as reflecting rainfall. Leveraging many simulations of the last millennium from a single climate model we find centennial timescale hydroclimate variability in the Caribbean that is consistent with that in the sinkhole lake record, and other contemporaneous regional records. In the third, we address the question of what coral records can tell us about climate variability. To do so, we drive a coral sensor proxy system model with output from an isotope enabled climate model. We find a surprising amount of spatial heterogeneity in the relationship between these records and climate variability. Nevertheless, some records are likely to provide robust information on large-scale reorganizations of the tropical atmosphere and ocean. Developing paleoclimate data requires substantial and expensive field and laboratory efforts. We view the integration of insights from climate models and paleoclimate data as imperative to best capitalize on these efforts.

*Refreshments:
10:30 AM - 11:00 AM (Atrium)