Climate forcings and feedbacks from the terrestrial biosphere. From Greenhouse-gas emissions to anthropogenic land use change

Stocker, Benjamin (2013). Climate forcings and feedbacks from the terrestrial biosphere. From Greenhouse-gas emissions to anthropogenic land use change (Unpublished). (Dissertation, Universität Bern, Philosophisch–naturwissenschaftliche Fakultät, Physikalisches Institut, Abteilung für Klima– und Umweltphysik)

Text stocker13phd.pdf - Other Restricted to registered users only Available under License BORIS Standard License. Download (32MB)

This thesis addresses different aspects of the interaction between the terrestrial biosphere and the climate system. Land carbon storage and emissions of greenhouse-gases are not only affected by direct anthropogenic impacts, e.g., through land use and land use change, but are also sensitive to changes in climate and atmospheric CO2, which induces multiple feedbacks. Interactions of the carbon (C) and nitrogen (N) cycles may limit future C sequestration and determine emissions of N2O. Its representation in global vegetation models is thus key to understanding the land’s role in the climate system. In this thesis, I present the implementation of a model for the coupled C and N cycling in the LPX-Bern Global Dynamic Vegetation Model. This model is then applied to investigate different climate-land feedbacks and anthropogenic impacts on the terrestrial C cycle.
After a brief introduction into some of the most important processes shaping the role of the terrestrial biosphere in the Earth system, with a particular focus on C and N cycle interactions and a description of their implementation in LPX-Bern, I will turn to how feedbacks between land and climate can be quantified in a consistent conceptual and mathematical framework (Chapter 1). This sets the basis for the study presented in Chapter 2, where the coupled Bern3D-LPX Earth System Model of Intermediate Complexity is applied to quantify the strength feedbacks arising from the sensitivity of terrestrial C storage, CH4 and N2O emissions, and surface albedo to climate and atmospheric CO2. The applied model is forced by a comprehensive set of input data for anthropogenic forcings and integrates earlier model developments in our group to include effects from wetland and peatland C balance and CH4 emissions, land use change, land albedo and the coupling to the Bern3D ocean-atmosphere model. Simulated historical N2O and CH4 emissions are consistent with atmospheric budgets and prescribed non-soil emissions. We show how future emissions depend on the scenario for climate and CO2 changes (RCP2.6 and RCP8.5 are investigated), how anthropogenic inputs of reactive N into terrestrial ecosystems amplify N2O emissions, and how the pattern of climate change as suggested by different CMIP5 climate models implies large uncertainties in future C storage. Under a business-as-usual scenario, the coupled model suggests a climate and CO2-induced amplification of future atmospheric N2O and CH4 concentrations by ∼40%. This feeds back to increase projected temperatures by 0.4–0.5°C above what would be expected without accounting for these feedbacks. As the negative feedback from CO2-fertilisation of terrestrial C storage becomes weaker in a high-CO2 world, positive feedbacks from N2O, CH4, and albedo exert an increasingly strong additional warming.
Chapter 3 addresses the influence of anthropogenic land use change on the C cycle of the Holocene, and whether human impacts are responsible for the observed rise in CO2 after 7000 yr BP (7 kyr BP). A set of historical land use change scenarios is applied, covering the plausible range of per-capita land requirement and representing contrasting shares of pre-industrial versus industrial era land use change to converge at the present-day distribution of agricultural areas. For the standard scenario (HYDE data), we found a CO2 increase of ∼1 ppm by 2 kyr BP, rising to ∼3 ppm at pre-industrial. Scenarios with a higher share of early agricultural expansion imply an earlier rise but smaller maximum elevation in CO2. This is linked to the time scales of ocean C uptake. However, such scenarios appear difficult to be reconciled with C budget constraints for the historical period. Relatively large uncertainties remain, in particular with respect to the long-term effect of soil cultivation on C storage and historical land use reconstructions. However, the sharp rise in CO2 between 7 and 5 kyr BP is neither captured by bottom-up modeling of land use change relying on upper-range assumptions for per-capita land requirement and assuming large impacts on soil C stores, nor do additional constraints from atmospheric δ13C measurements and independent estimates of peatland C uptake support an anthropogenic C source of sufficient magnitude.
Chapter 4 presents results for historical and future land use change C emissions based on a further model development of how transitions between areas under different anthropogenic land use are implemented in LPX-Bern. This method accounts for parallel expansion and abandonment of croplands and pastures within a single gridcell (shifting cultivation-type agriculture) and simulates biomass removals due to wood harvest. Additionally, a method is presented to derive full land use area transitions based on minimum input data and evading methodological obstacles but yielding quantitatively very similar results as when prescribing this data. Our results are in line with other studies suggesting that effects of shifting cultivation and wood harvesting contribute an additional C source of ∼0.5 PgC/yr, leading to ∼60% higher estimates for present-day total emissions from land use and land use change (1.2 PgC/yr in early 2000s). Under future scenarios of decreasing agricultural expansion, this share is even larger and is on the same order as the indirect emissions arising from the replacement of forests acting as potential C sinks in a future high-CO2 world. This illustrates the importance of simulating the impacts of anthropogenic land use change in high process detail. Furthermore, I highlight the necessity for a consistent framework to quantify land use change emissions and its indirect component fluxes caused by feedbacks.
Chapter 5 presents the development of a model to simulate the extent of wetlands and peatlands based on high-resolution topography information, coarse-resolution soil water content, runoff, and peatland growth conditions. The implementation follows a TOPMODEL approach and enables highly cost-efficient applications in Earth system models without having to rely on prescribed present-day wetland and peatland maps. Model parameters are tuned to match observational data for the seasonal extent of inundated areas and peatland maps. First applications in combination with LPX-Bern yield promising results and will enable us to revisit the peatland C budget and the feedback from CH4 under climatic conditions significantly different from today’s.
Appendix A.1 contains a specific documentation of relevant model changes implemented in Bern-LPX, version 1.0. Appendix A.2 documents the implementation and testing of soil water drainage limitation to (potentially) improve the representation of inundated areas. Two co-authored publications are not included in this thesis. In Spahni et al. (2013), LPX-Bern is applied to simulate the buildup of present-day peatlands after the Last Glacial Maximum. Le Qu´er´e et al. (2013) present updated results for the modern C budget where land use emission estimates from LPX-Bern were used in combination with other land use emission estimates and a wealth of other data sources and methods to quantify the atmospheric CO2 growth rate, fossil fuel emissions, ocean uptake and the residual terrestrial sink.

Interest & Impact

Downloads

3 since deposited on 07 Mar 2024
3 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