Modelling the Terrestrial Biosphere: Uncertainties and Constraints

Lienert, Sebastian (2018). Modelling the Terrestrial Biosphere: Uncertainties and Constraints (Unpublished). (Dissertation, Universität Bern, Philosophisch–naturwissenschaftliche Fakultät, Physikalisches Institut, Abteilung für Klima– und Umweltphysik)

Text lienert18phd.pdf - Other Restricted to registered users only Available under License BORIS Standard License. Download (37MB)

Anthropogenic activities such as the burning of fossil fuels and changes in land use have increased the presence of greenhouse gases in the atmosphere, altering the global climate system. The greenhouse gas concentration in the atmosphere is attenuated by the global biogeochemical cycles. To enhance our understanding of these complex cycles, and in effect the climate system, the expertise of a wide range of disciplines has to be brought together. State-of-the-art contemporary observations (e.g. remote sensing), long time archives (e.g. ice cores) and computer models all have a vital role to play.
This thesis focuses on the modelling of the terrestrial biosphere and how inherent model uncertainties can be reconciled with observational constraints. The first emphasis of the thesis is on the assimilation of observations in a Bayesian framework. The second focus is the simulation of the stable isotope of carbon (13C) and its use as a potential observational constraint.
In Chapter 1, an introduction to the thesis is given. The mechanisms behind the response of the climate system to changes in the radiation budget of the Earth are presented and set into context by discussing the contemporary anthropogenic perturbance of the carbon cycle. In a second part, more specific topics are discussed. The Dynamic Global Vegetation Model (DGVM) LPX-Bern and its development history are introduced. The chapter concludes with techniques for generating and evaluating perturbed model parameter ensembles, an introduction to land-use change and a section on the stable isotope of carbon, an important tracer in the climate system.
Chapter 2 contains a study published in Biogeosciences applying a probabilistic approach to the LPX-Bern model. By using observational constraints a new best-guess version of model parameters is formulated, termed LPX-Bern v1.4. The parameter ensemble is also used to quantify magnitude and uncertainty in historical land-use emissions of CO2. The spatial structure of the emissions are reported and compared to other independent estimates. Furthermore, mechanistic differences in the representation of land-use processes are investigated.
In Chapter 3 the implementation of carbon isotope discrimination of LPX-Bern and the land biosphere module (CLM4.5) of the Community Earth System Model (CESM) is compared to measurements in a Biogeosciences study. The models both show consistent results with a global compilation of C3 δ13C leaf measurements. Less agreement is found when comparing model results to tree-ring records. Intrinsic water use efficiency is related to carbon isotope discrimination and shows an increasing trend over the 20th century in the tree-ring measurements. This trend is captured by LPX-Bern but overestimated in CLM4.5, possibly explained by a too strong nitrogen limitation in the model. The study highlights the potential of 13C to evaluate global land carbon cycle models.
The importance of 13C is further explored in Chapter 4, where first steps toward fully coupled 13C simulations are taken. Two alternative formulations of terrestrial plant discrimination are presented and tested. The LPX-Bern model is modified to be able to prescribe C3 and C4 crop distribution, which has a strong impact on the global terrestrial 13C signature. Finally, a simulation of LPX-Bern is used in conjunction with a Bern3D ocean simulation to obtain global atmosphere-to-surface 13C fluxes. With the help of a model of atmospheric transport (TM2) those fluxes are translated to local concentration anomalies, allowing to compare the simulated seasonal cycle of 13C with measurements. The results are promising, indicating that it is in principle possible to accurately simulate the seasonal cycle of 13C in the atmosphere, but some further steps are needed for a successful coupled simulation of 13C.
Chapter 5 is a compilation of supplementary results of studies where simulations of LPX-Bern were featured. The Global Carbon Budget (GCB) is an annual publication aiming to improve our knowledge of the contemporary carbon cycle, featuring amongst other an ensemble of DGVMs. Different versions of LPX-Bern over the different iterations of the GCB are compared with each other and to the other DGVMs and GCB estimates. A similar project, focusing on the emissions of the important greenhouse gas N2O, is the N2O Model Intercomparison Project (NMIP). It is the first DGVM model intercomparison project focusing on the modelling of the terrestrial nitrogen cycle and associated emissions of N2O. The LPX-Bern simulation is set into the context of the ensemble and an additional factorial simulation on the effect of the timing of nitrogen fertilizer application is presented. Lastly, complementary results of a study investigating the sensitivity of European forests to climate variability are provided. The study features a novel tree-ring network allowing to quantify the interannual variations in biomass increment. Results from factorial simulations to investigate climate sensitivities in the interannual variability of net primary productivity in LPX-Bern are featured. Additionally, the relationship between mean production and interannual variability is examined in the model simulation.
The last Chapter 6 offers an outlook on potential future projects and model development. The thesis is accompanied by three appendices. In the first appendix daily in- and output capabilities of the model are discussed. Results from simulations focusing on specific sites are featured in the second appendix. The last appendix provides an overview of the implementation of the benchmarking framework featured in the probabilistic approach of chapter 2.

Interest & Impact

Downloads

6 since deposited on 22 Feb 2024
6 in the past 12 months

Citations

5 Citations in Web of Science ®
6 Citations in Scopus

Search

in Google Scholar™

Services

Actions (login required)

Edit item

Actions (login required)

Edit item