Life and times in the early Paleogene hot-house: Using global models to reconstruct a lost paradise

This dissertation focuses on the general warm climate of the early Paleogene (50--60Ma), and more specifically on the rapid warming that marks the geologic record at the Paleocene-Eocene (P/E) boundary (∼55Ma). The primary aim of this research is to use global climate models (GCMs) in connection with a dynamic global vegetation model (DGVM) to examine the sensitivity of the early Paleogene climate and ecosystem to changes in atmospheric greenhouse gas concentrations.; The first part of this study focuses on reconciling the large range in estimates for paleo-pCO2 from proxy data for the early Eocene. Comparisons between proxy climate estimates from Eocene fossil flora and model output from a CO2 sensitivity study with an atmosphere and land GCM coupled to a sea ice and a slab ocean model suggest that CO2 must have been high (∼1000ppm) in order to support the warm climate flora found at high latitudes during this time. While this sensitivity study does not specifically consider the warming at the P/E boundary, it does have implications for this event. The relative increases in global mean temperature with increasing CO2 suggest that if the warming across the P/E boundary were due to CO2 alone, CO2 must have increased by 500--1500ppm.; The second part of this study uses a DGVM to explore how the flora of the early Paleogene may have responded to a warming such as that at P/E boundary. Neglecting the physiological effects of CO2 on vegetation, model results suggest that warming due to an increase in CO2 of 560ppm has the greatest effect on high latitude vegetation. Results also suggest that the prescribed vegetation used in previous modeling studies of this time period is not necessarily appropriate for the modeled climates. Given the relative infancy of DGVMs compared with GCMs, this dissertation also explores the utility of DGVMs in paleoclimate research through a series of short experiments. Model output from these experiments indicates that further modeling studies of paleo-vegetation should carefully consider the physiological effects of CO2 on vegetation, as well as the influence of model input, specifically soil composition, on the modeled global distribution of vegetation.