IMAGe Brown Bag Seminar - Combining Proxy Data and Climate Simulations for Paleoclimate Reconstructions

06/28/2017 - 12:00pm to 1:00pm
ML - Chapman Room

Combining Proxy Data and Climate Simulations for Paleoclimate Reconstructions

Nils Weitzel

Meteorological Institute of the University of Bonn

Spatio-temporal reconstructions of past climate are important for the understanding of the long term behavior of the climate system and the sensitivity to forcing changes. Unfortunately, they are subject to large uncertainties, have to deal with a complex proxy-climate structure, and a physically reasonable interpolation between the sparse proxy observations is difficult. As Bayesian models can include multiple sources of information (e.g. different proxy types, physical models for interpolation, output from climate simulations) and quantify uncertainties in a statistically rigorous way, they are in theory a good method to approach these problems.

We present a Bayesian framework that combines a network of pollen records with a spatial prior distribution estimated from an ensemble of climate simulations. The use of climate simulation output aims at a physically reasonable spatial interpolation on a regional scale. To transfer the pollen data into (local) climate information, we use a forward version of the probabilistic indicator taxa model. In addition, we can include Gaussian distributed data from preprocessed proxy records. The Bayesian inference is performed using MCMC methods following a Metropolis-within-Gibbs strategy.
As applications of our framework, we show reconstructions of European temperature for the Mid Holocene (~6,000 years BP) and the Last Glacial Maximum (~21,000 years BP) using an ensemble of climate simulations from the PMIP3 project and pollen syntheses from Bartlein et al. (2011) and Simonis et al. (2012).

Future extensions of our method are the inclusion of other proxy types, the extension to spatio-temporal reconstructions of the last deglaciation, and the incorporation of additional types of uncertainty.


Wednesday, June 28th, 2017

12:00pm – 1:00pm

Mesa Lab, Chapman Room

(Bring your lunch)