Modeling changes in the global carbon cycle-climate system

Steinacher, Marco (2011). Modeling changes in the global carbon cycle-climate system (Unpublished). (Dissertation, Universität Bern, Philosophisch–naturwissenschaftliche Fakultät, Physikalisches Institut, Abteilung für Klima– und Umweltphysik)

Text steinacher11phd.pdf - Other Restricted to registered users only Available under License BORIS Standard License. Download (26MB) | Request a copy

There is strong evidence that human activities are altering the Earth’s climate and biogeochemical cycles. Anthropogenic carbon emissions have increased the concentration of carbon dioxide (CO2) in the atmosphere by about 40% since preindustrial times. The elevated CO2 levels are the main drivers of changes in the atmosphere’s radiative balance as well as in the ocean’s chemistry. This thesis investigates consequences of the anthropogenic interference with the coupled carbon cycle-climate system. A main focus is on the acidification of the oceans caused by the uptake of anthropogenic CO2. Further, interactions between climate and the carbon cycle are addressed. Numerical simulations with coupled carbon cycle-climate models are used to quantify future changes under several scenarios of CO2 emissions, land-use change, and other forcings. Such projections are important for assessing potential impacts on ecological and socio-economic systems. Moreover, they provide a basis for developing mitigation and adaptation strategies and contribute to the understanding of the complex carbon cycle-climate system in general.
The introduction in chapter 1 provides an overview of the global carbon cycle with an emphasis on ocean biogeochemistry. Further, important concepts such as radiative forcing, feedbacks, and climate sensitivity are introduced. The influence of human activities, predominantly the emissions of CO2, is summarized and discussed with respect to ocean acidification and the perturbation of the radiative balance, followed by an overview of the associated changes in climate and feedbacks within the coupled carbon cycle-climate system.
Chapter 2 consists of three publications presenting projections of ocean acidification with a special focus on the Arctic Ocean. The global coupled carbon cycle-climate model NCAR CSM1.4-carbon is applied to simulate ocean acidification for the industrial period and for the two IPCC SRES emission scenarios A2 and B1 (2000–2100). The results support the important finding of earlier studies that the ocean’s chemistry is undergoing large and rapid changes in response to anthropogenic carbon emissions with potentially severe impacts on marine ecosystems. Ocean pH and the aragonite saturation state, a key variable governing impacts on calcifying organisms, decrease rapidly on a global scale. In the A2 scenario, water saturated by more than 300%, considered suitable for coral growth, vanishes by 2070 (CO2 ≈ 630 ppm), and the ocean volume fraction occupied by saturated water decreases globally from 42% to 25% over this century. By extending the simulations from 2100 to 2500 under the assumption that carbon emissions are (unrealistically) reduced to zero in 2100, it is shown that the projected changes in the 21st century are largely irreversible for at least several hundred years.
The Arctic Ocean is identified to be particularly vulnerable with respect to ocean acidification. The largest projected pH changes (∆pH = -0.45) worldwide occur in Arctic surface waters. Aragonite undersaturation in the Arctic Ocean is imminent and expected to begin within the next decade for both scenarios. By the time atmospheric CO2 exceeds 490 ppm (2040 in A2, 2050 in B1), more than half of the Arctic is projected to be undersaturated at the surface. By the end of the twenty-first century and for the A2 case, undersaturation in the Arctic Ocean also occurs with respect to calcite. The main reasons for the vulnerability of the Arctic Ocean are its naturally low saturation state and that Arctic climate change amplifies the acidification, in contrast to other regions like the Southern Ocean, where no significant interaction of climate change and ocean acidification is found in our simulations. Freshening of surface waters and increased carbon uptake in response to sea ice retreat is projected to amplify the decrease in Arctic mean saturation and pH by more than 20%. The results imply that surface waters in the Arctic Ocean will become corrosive to aragonite, with potentially large implications for the marine ecosystem, if atmospheric CO2 is not kept below 450 ppm.
In chapter 3, the response of the marine ecosystem productivity to climatic changes under the SRES A2 emission scenario is investigated with four comprehensive global coupled carbon cycle-climate models. The marine biological cycle is an important element of the carbon cycle and climate system, which influences the abundance of atmospheric CO2. Yet the simulation of primary productivity in general circulation models is a relatively new field and remains challenging. The four models applied here, which include representations of marine ecosystems of different complexity and structure, all show a decrease between 2 and 13% in global mean net primary productivity by 2100. Despite differences in magnitude, a decrease in productivity over the 21st century is a robust result. Reduced nutrient supply to the productive zone in response to a shallower mixed layer depth and slowed circulation is identified as the dominant mechanism in the low- and mid-latitude ocean as well as in the North Atlantic. In parts of the Southern Ocean, on the other hand, an alleviation of light and/or temperature limitation leads to a productivity increase. The results are compared to recent projections relying on an empirical model approach, which suggest a productivity increase, and differences are discussed. Further, a method based on regional model skill metrics is developed to generate weighted multi-model means of projected changes in productivity. The multi-model estimate of the decrease in net primary production by 2100 is -2.9 GtC yr−1 (-8%).
In chapter 4, the Bern3D-LPX Earth System Model of Intermediate Complexity is applied to perform a large set of coupled climate-carbon cycle simulations over the last millennium and into the future under different scenarios. At first, the new model setup is presented and the newly implemented land surface albedo model component is described in detail and validated by comparing results with satellite observations. Further, the impact of land-use changes on the surface albedo and radiative forcing is quantified and compared to internal feedbacks. For the future projections, the model is forced as specified by the four representative concentration pathways (RCP), which have been selected for simulations with regard to the next Assessment Report of the IPCC, and their extensions up to 2300. Projections of global warming, steric sea level rise, and ocean acidification are provided and committed changes under different assumptions are quantified. Projected changes in the global mean temperature by 2100 range from 1.8 ◦C in the low mitigation scenario (RCP3-PD) to 4.6 ◦C in the high baseline scenario (RCP8.5). Further, changes in the carbon cycle are analyzed and simulations with prescribed CO2 concentrations (diagnosed emissions) are compared to emission-driven simulations. When prescribing CO2 emissions, the model simulates significantly higher CO2 concentrations than specified for the RCP scenarios. This is a consequence of a relatively weak oceanic carbon sink, a considerable effect of land-use induced changes in the terrestrial sink capacity, and losses of carbon from soils. Finally, allowable emissions are quantified in four simulations where global mean temperatures are stabilized at 1.5 to 4 ◦C.
The impact of climate change mitigation on ocean acidification projections is discussed in chapter 5. A large set of baseline (no climate policy) and mitigation emission scenarios is explored in simulations with the Bern2.5CC and NCAR CSM1.4-carbon models. Emission scenarios provide an indication of the potential effect of mitigation policies but are often idealized and assume that new technologies and climate policies can be introduced over a relatively short period of time, especially in the lowest mitigation scenarios. The physical impacts in terms of ocean acidification are lower in mitigation then baseline scenarios. Early decisions are required to meet specific mitigation targets. The low scenarios, which result in similar global mean surface carbonate saturation states by 2100 than observed today, depart from the corresponding baseline emissions around 2015–2020. They require socio-political and technical conditions that are very different from those now existing.
Chapter 6 presents a study that investigates the response of the carbon cycle to large changes in ocean circulation. Ensemble simulations with the NCAR CSM1.4-carbon model are forced with freshwater perturbations in the North Atlantic and in the Southern Ocean that lead to reductions of the Atlantic Meridional Overturning Circulation. Changes in the physical climate fields, in turn, affect the land and ocean biogeochemical cycles and cause a variations in the atmospheric CO2 concentration by up to 20 ppm. The response depends on the location where the freshwater perturbation is applied. In the case of a North Atlantic perturbation, the land biosphere reacts with a strong reduction in carbon stocks and atmospheric CO2 levels are increased. This response is strongest in the tropical regions due to a shift of the Intertropical Convergence Zone, and can be found most clearly in South America. The results are mainly discussed with respect to abrupt climatic changes in past and they are compared to proxy records. However, the results also have some relevance with respect to anthropogenic climate change. Many models project a reduction in the Atlantic Meridional Overturning Circulation under future scenarios and the simulated responses in the carbon cycle presented here illustrate potential feedback mechanisms.
Finally, an outlook in chapter 7 outlines how the work presented in this thesis will be continued in future studies. In the appendices, the coupling of the Bern3D and LPX models is documented along with several other technical improvements that have been implemented during this thesis.

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