Characterizing Agricultural N2O Emissions in the U.S. Midwest Using a Novel Top-Down Approach Based on Airborne In Situ Measurements

Kurzfassung

Nitrous oxide (N2O) is, after carbon dioxide and methane, the third most important long-lived anthropogenic greenhouse gas and nowadays the dominant ozone-depleting substance in the stratosphere. Anthropogenic emissions, mainly released due to fertilization practices in agricultural regions, have increased atmospheric concentrations by more than 20% since the start of the industrialization to about 334 ppb. Despite its important role, N2O is almost ignored in emission reduction plans submitted to the Paris Agreement. One of the reasons for this is the insufficient characterization of regional N2O sources due to the lack of measurements and methodologies required for thorough analyses of complex N2O area sources. This thesis investigates the hypothesis that regional-scale airborne in situ measurements of N2O are well-suited to characterize N2O emissions from intensively cultivated agricultural regions and to evaluate state-of-the-art bottom-up emission inventories. To this end, an exceptional in situ N2O dataset is used, which has been collected in the course of the Atmospheric Carbon and Transport-America (ACT-America) project (2016-2019) during five aircraft campaigns covering all four seasons over the eastern part of the U.S. It consists of high-precision flask measurements and unique continuous measurements with an absorption spectrometer (Quantum Cascade Laser Spectrometer (QCLS)), which, in the course of this work, was optimized for N2O and successfully deployed during two of the five aircraft campaigns. In combination with WRF (Weather Research and Forecasting model) simulations and available atmospheric dispersion calculations, N2O emissions in the bottom-up inventory EDGAR (Emissions Database for Global Atmospheric Research) are scaled to quantify emissions from the U.S. Midwest - a region with one of the most intensive agriculture in the world. Using a combination of QCLS measurements and WRF simulations, N2O emissions in the Midwest in October 2017 (0.42+-0.28 nmol m-2 s-1) and June/July 2019 (1.06+-0.57 nmol m-2 s-1) have been quantified. Flask measurements, available for all five ACT-America deployments, were further used to study the seasonality of emissions. Primarily due to fertilization, emissions in spring were found to be 75% higher than in summer, while in fall, they were observed to be 13% higher than in summer. In winter, estimated emissions even exceeded the summer estimates by 230%, most likely due to freezing/thawing processes of the soils. The results of this study are consistent with other ground-based top-down studies. However, further studies are needed to be able to fully capture the complexity of N2O emissions. Comparisons with the bottom-up inventory EDGAR show that EDGAR underestimates Midwest N2O emissions significantly (factors between two and ten), for exceptional cases even by factors up to 20. Monthly Midwest emission estimates for 2011-2015 calculated with the process-based model DayCent (daily time-step version of the CENTURY model) are significantly closer to the results of this thesis than EDGAR (factors between two and five), since DayCent considers regional characteristics like soil conditions and weather. A sensitivity analysis using the flask measurements and dispersion calculations indicates that the heterogeneity of N2O soil emissions in the Midwest mainly correlates with soil temperature in summer and soil moisture in spring and fall. In winter, soil emissions are dominated by freezing/thawing processes. For a thorough quantitative analysis, additional simulations with a process-based model are required. This work shows that airborne in situ N2O measurements are suitable for characterizing regional N2O emissions. This is a valuable contribution to the effort to establish a national N2O emission monitoring system, the basis for emission reduction strategies, which are urgently needed to meet the targets of the Paris Agreement.