Utilising palaeobotanical data to calibrate the response of vegetation to carbon dioxide in Earth system models

Rising atmospheric carbon dioxide (CO2) concentrations alter the vegetation composition indirectly through climate change and

directly through plant physiological modifications. Both responses modulate climate through changed energy, moisture, and carbon

fluxes between land and atmosphere. This makes accurate estimates of the responses crucial for future vegetation and climate

projections. Yet, large uncertainties remain about the magnitude of the direct response because vegetation's slow response time and

complex interactions of species hinder a quantification purely from instrumental and experimental data. To overcome this limitation,

PALVEGMOD explores to what extent reconstructions of vegetation changes in the past can constrain the direct vegetation response.

Palaeobotanical data provides information on vegetation cover during Earth's history. I will develop new methods to quantify the

direct vegetation response to CO2 in Earth system models (ESMs). Then, I will calibrate the direct response in the ESM HadCM3B by

utilising global compilations of vegetation reconstructions for three periods in the last 17 million years with atmospheric CO2

concentrations between ~185ppm and ~600ppm. The calibration optimises model parameters in a Bayesian framework that accounts

for the limited data availability and uncertainties of palaeobotanical data. Finally, I will derive the first estimates, which employ

constraints from palaeobotanical data on the direct vegetation response to CO2, for the impact of the direct response on vegetation

and climate in emission scenarios until 2500 CE. The new constraints can contribute to an improved quantification of the future land

carbon sink and help assess the effects of varying CO2 concentrations on the efficiency of large-scale afforestation for removing atmospheric CO2. If successful with HadCM3B, PALVEGMOD's methods can be used to improve a range of Earth system models including state-of-the art models used in the next IPCC report.