Tilling the earth; modelling global N2O emissions caused by tillage

Abstract

Agriculture is the largest contributor of non-CO2 anthropogenic greenhouse gas emissions (GHG). Agricultural based mitigation strategies (e.g. no-tillage) are identified to reduce emissions from agricultural soils through improved agricultural management. Global ecosystem models that are usually used for finding the potential of agricultural based mitigation strategies are limited, because processes related to agricultural management are currently underrepresented in global ecosystem models. The aim of this thesis is to contribute to the representation of agricultural management in global ecosystem models, so that the potential of agricultural based mitigation practices can be better quantified. Therefore, this thesis first addressed how processes related to agricultural management can be described in global ecosystem models, with a focus on processes related to tillage and N2O emissions. This analysis resulted in a standardized framework that can be followed to incorporate other agricultural management practices in global ecosystems as well. After indicating how processes related to tillage can be described, they were implemented into the global ecosystem model LPJmL. Subsequently, the extended LPJmL model was evaluated on its performance on various fluxes (including N2O and CO2) at the global scale and for a number of experimental sites. Finally, the uncertainty caused by the upscaling of soil input data when assessing tillage effects on N2O emissions were addressed. LPJmL was not capable of accurately simulating tillage effects on N2O emissions. Hence, the potential of mitigating N2O emissions through tillage management cannot be well assessed. However, the implementation of the more detailed tillage-related mechanism into the global ecosystem model LPJmL improved the ability to understand agricultural management options for agricultural mitigation of CO2 emissions, climate change adaptation and reducing environmental impacts. The work in this thesis concludes that as processes related to agricultural management can be incorporated into global ecosystem models by following the standardized framework, and data-scarcity on agricultural management does not necessarily limit the evaluation of the extended model, there is no general barrier to extend global ecosystem models by modules for the representation of agricultural management.