Methane emissions and isotopes of northern peatlands in a global vegetation model

Zürcher, Sibylle (2013). Methane emissions and isotopes of northern peatlands in a global vegetation model (Unpublished). (Dissertation, Universität Bern, Philosophisch–naturwissenschaftliche Fakultät, Physikalisches Institut, Abteilung für Klima– und Umweltphysik)

Text zuercher13phd.pdf - Other Restricted to registered users only Available under License BORIS Standard License. Download (41MB) | Request a copy

Methane is, after water vapor and carbon dioxide, the third most potent greenhouse gas in the atmosphere at present. Its concentration has strongly increased in the 20th century and reached a concentration exceeding any value ever measured over the last 800’000 years. For the effort to mitigate climate change, it is important to understand the dynamics leading to the atmospheric methane concentration. The atmospheric methane level is a balance between sources and sinks. While the present global total source and sink are relatively well known, the partition between the different contributors is not well understood. To be able to predict future CH4 atmospheric levels it is important to understand the size of the different CH4 sources and sinks and their reaction to climate change. As current observations are limited or only cover a small time window of climate and environmental change, past ice core observations in CH4 and δ13CH4 are of great value to validate models. Global wetlands are the largest natural emitters and very climate sensitive. Their feedback to a future warming is believed to be positive. Therefore, it is among many other things important to improve the process-based modelling of northern peatlands that are an important part of wetlands and a source of methane with a considerable uncertainty, variability and sensibility to climate change. Taking the isotopic composition of a source into account provides a further constraint in this difficult task. This thesis contributes to the effort of understanding the methane cycle by improving and enhancing a methane model embedded in a dynamic global vegetation model and by newly implementing and calibrating a methane isotope routine into this model framework.
This thesis is structured along 5 major chapters, while an outlook at the end of this thesis highlights research challenges and further topics to be addressed in future work.
Chapter 1 provides an introduction giving some general information about the global methane budget in the past and present, an overview on the important sources and sinks, and details on the role of wetlands and in particular peatlands. The processes leading to methane emissions in peatlands are discussed, as well as the importance of methane isotopes. The required notations are introduced as well.
In Chapter 2 an introduction to climate modelling is provided. Further, the model used in this thesis is presented in detail with a special focus on its methane routine and the implemented isotopes. At the end of the Chapter, site simulations are discussed in detail to illustrate the implemented mechanisms and explain the produced output like the methane emissions, isotopic signature or the soil profile. Note that for the simulations done during this thesis, two different versions of the model were used which are discussed separately. A first development step was to implement, improve and calibrate the methane module WHyMe (developed by R. Wania in Bristol) into Bern-LPJ. Bern-LPJ is a dynamic global vegetation model that simulates plant distribution and carbon stocks and flows. Methane isotopes were missing at that point. The simulations in the peer reviewed studies presented in Chapter 3, 4.1 and 4.2 were conducted with this model version (LPJ-Bern). A second version (called LPX) with general model improvements outside the methane routine as well as in the methane routine itself including the implementation of methane isotopes was used for the simulations discussed at the end of Chapter 2 and in Chapter 5.
Chapter 3 presents a model application in a study about the impact of an abrupt cooling event on methane emissions in northern peatlands. Rapid changes in temperature and precipitation are paralleled with substantial variations in atmospheric methane concentrations as documented in ice cores. Most studies attribute a change in methane concentrations to emission changes from boreal and sub-tropical wetlands and boreal peatlands are believed to be the dominant and most directly responding ecosystem. The calibration of the model with modern site data is described (this was redone for LPX and summarized in the Appendix). The results of the methane emission simulations in reaction to an abrupt climate change are presented including sensitivity test to model parameters and climate input data. The main result is that if the simulated changes in climate are taken as an analogy to the 8.2 kyr event, boreal peatland emissions alone can only explain about 23% of the 80 ppb decline in atmospheric methane concentrations.
Chapter 4.1 shows the results of a model inter-comparison project (WETCHIMP) where the present ability to simulate wetland extend, characteristics and corresponding methane emissions is investigated. LPJ-Bern was one of ten models taking part in the comparison. A common experimental protocol was used to drive all models with the same climate forcings.
Chapter 4.2 is the follow-up paper to the WETCHIMP paper and provides technical details for the six experiments that were conducted and about the models that took part in the study. The major conclusion is that the models demonstrate a great disagreement in their simulations of the wetland areal extend and methane emissions in space and time and it is stressed that we presently do not have sufficient wetland methane observations to evaluate model fluxes.
Chapter 5 presents applications of the implemented isotope routine. NH peatland runs for present day were performed and discussed. The sensitivity of the isotopic signature of methane emissions to the different plant types simulated in the peatland area are investigated as well as to input parameters like temperature or precipitation. Further, the correlation between the emitted signature and model parameters are investigated. The simulated emission signatures for present day are in the range of peatland methane emission measurements. Finally, a simulation of NH peatland methane emissions from the LGM till present was performed to assess the variability of δ13CH4 and the possible contribution to the atmospheric signal during the Holocene.

Interest & Impact

Downloads

0 since deposited on 07 Mar 2024
0 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