METHANE POLLUTION IN THE SOUTHERN GREAT PLAINS: SOURCE QUANTIFICATION AND EXAMINATION OF IMPACTS OF LAND-ATMOSPHERE PROCESSES USING ROUTINE OBSERVATIONS, FIELD CAMPAIGN DATA, AND SIMULATIONS

Abstract

Methane (CH₄) is a potent greenhouse gas, and accurately characterizing its emissions and distribution in the atmosphere is critical for advancing our understanding of its role in climate change. This dissertation consists of two key components of CH₄ dynamics: land-atmosphere interactions and emission quantification. By combining advanced modeling techniques with field observations, this work seeks to identify CH₄ emission sources and enhance our understanding of how land-atmosphere processes influence CH₄ concentrations. This is particularly important for improving inverse modeling approaches, where accurately linking CH₄ concentrations to emission sources relies on understanding how planetary boundary layer (PBL) dynamics and land surface processes affect CH₄ distribution in the atmosphere. These insights are crucial for refining CH₄ emission inventories and enhancing the effectiveness of mitigation strategies aimed at reducing CH₄ emissions. The first focus of the dissertation is to examine land-atmosphere interactions specific to the Southern Great Plains (SGP) site in Oklahoma, as these dynamics play a crucial role in the transport and mixing of CH₄ within the PBL. We optimized seven key parameters of the Noah-Multiparameterization Land Surface Model (Noah-MP) to improve its representation of land-atmosphere interactions. Using Bayesian optimization (BO), we fine-tuned the model parameters using data from flux tower observations. This optimization not only enhanced the simulation of surface energy fluxes and meteorological state variables like temperature, but also provided insights into soil characteristics, suggesting that the top layer of soil at the SGP site may contain more sand than indicated by existing classifications. These improvements extended beyond the initial nine-day optimization period to the entire growing season in 2016, enhancing the accuracy of atmospheric state variables and top-layer soil conditions. Building on the improved simulation of land-atmosphere processes, the second part of the research investigates CH₄ concentration enhancements at the SGP site from 2017 to 2020. Analysis of local meteorological data revealed that stable boundary layer conditions—characterized by strong temperature inversions and shallow PBLs—trap CH₄ near the surface, particularly during nighttime, leading to significant concentration spikes. Wind direction analysis identified a concentrated animal feeding operation (AFO) located 5.8 km northwest of the SGP site as a potential source of these CH₄ emissions. In a field campaign conducted in June 2024, we measured CH₄ concentrations exceeding 6000 ppb downwind of the AFO, corresponding to an emission rate of approximately 95 kg·hr⁻¹, as calculated using the mass balance method. Additional nearby sources, including open-range cattle and a natural gas pipeline, were also identified as potential contributors to elevated CH₄ levels. The final component of this dissertation integrates the field campaign data with high-resolution atmospheric modeling using the Weather Research and Forecasting model with Greenhouse Gas (WRF-GHG) model to assess whether the AFO could be responsible for the observed CH₄ enhancements at the SGP site. Using the WRF-GHG model, we simulated CH₄ dispersion from the AFO and validated these simulations against field data. The model successfully captured the spatial dispersion of CH₄ from the AFO and aligned well with field measurements, particularly around the AFO itself. However, the simulated CH₄ enhancement at the SGP site was lower than observed, even when assuming an AFO emission rate of 100 kg·hr⁻¹. This discrepancy suggests that the actual emission rate from the AFO may be higher than previously estimated, underscoring the need for more accurate emission inventories. The field campaign also revealed that rainfall events significantly increase CH₄ emissions, particularly from manure ponds and open lots, where microbial activity is enhanced in wet conditions. These findings highlight the importance of improving CH₄ emission inventories and refining atmospheric models, particularly in agricultural regions where localized sources can have substantial impacts on regional CH₄ concentrations. This dissertation bridges gaps in our understanding of CH₄ emissions by optimizing land-surface models, quantifying local CH₄ sources, and validating atmospheric simulations. By integrating empirical field data with advanced modeling approaches, this research enhances our understanding of CH₄ dynamics at local and regional scales, with important implications for global CH₄ inventories and climate policies aimed at mitigating CH₄ emissions.