Hu, Lu2014-08-192014-08-192014-06https://hdl.handle.net/11299/164972University of Minnesota Ph.D. dissertation. June 2014. Major: Land and Atmospheric Science. Advisor: Dylan B. Millet. 1 computer file (PDF); xii, 178 pages, appendix 1-4.This dissertation presents the first-ever in-situ tall tower measurements of volatile organic compound (VOC) concentrations. The data span August 2009 through July 2012, and provide new constraints on seasonal and long-term controls on VOC sources and their atmospheric effects. The 200 m sampling height provides a large-scale footprint, while the tall tower location, near the intersection of the main North American ecosystems and at times downwind of the Twin Cities, affords information on natural emissions from some of the most important US landscapes as well as on anthropogenic sources.I interpret the dataset using an atmospheric chemical transport model (GEOS-Chem CTM), with a focus on several key atmospheric VOCs. This dissertation finds that current models underestimate methanol emission rates for younger versus older leaves. This biased seasonality means that the photochemical role for methanol early in the growing season is presently underestimated. A Bayesian inverse analysis of the tall tower observations reveals that the prior estimate of North American anthropogenic acetone sources (based on the US EPA's NEI05 inventory) is accurate to within 20%. However, biogenic acetone emissions from broadleaf trees, shrubs, and herbaceous plants are presently underestimated (~37%), while emissions from needleleaf trees plus secondary production from biogenic precursors are overestimated by a similar amount (~40%). Model-measurement comparisons imply that isoprene emissions in the immediate vicinity of the tall tower are accurately captured by the MEGANv2.1 biogenic inventory, but that larger-scale regional emissions are underestimated, reflecting the heterogeneous land cover in this transitional landscape. Isoprene emissions play a key role in seasonal shifts between VOC-limited chemistry in the spring and fall and NOx-limited or transitional chemistry in the summer.A Bayesian inverse analysis based on the tall tower measurements suggests that: i) the RETRO global emission inventory significantly overestimates (> two-fold) US C6-C8 aromatic emissions; ii) the US EPA's NEI08 inventory likewise overestimates the toluene flux by a factor of 3, partly reflecting a bias in the estimated non-road emissions; and iii) total annual emissions of benzene and C8 aromatics in the EPA's NEI08 are accurate to within the analysis uncertainty, but with some seasonal biases for on-road emissions.en-USGEOS-ChemProton transfer reaction mass spectrometer (PTR-MS)Sources and emissionsTall tower measurementsTop-down and bottom-up estimatesVolatile organic compounds (VOCs)Thesis or Dissertation