|
Abstract:
|
One of the most-widely available climate proxy data are tree pollen collected in sediments. Pollen grains in sediments are counted and the relative abundance of different tree species is a function of the underlying climate state. Thus, reconstructing patio-temporally correlated climate from pollen involves estimating a complex, non-linear relationship from multinomial data making traditional Markov Chain Monte Carlo methods difficult. In this work, I apply a Polya-gamma data augmentation scheme to enable conjugate parameter updates and reduce computational costs, allowing for Bayesian paleoclimate reconstructions from pollen to be performed at regional-to-continental scales.
|
