Increasing atmospheric concentrations of carbon dioxide (CO2) and its impacts on climate change have driven the implementation of regulatory policies such as emissions trading schemes. Accurate quantification of CO2 emissions from individual sources is crucial for effective policy design and mitigation strategies. Uncrewed aerial vehicles (UAVs) present promising tools for emissions monitoring, but existing approaches often rely on manual flight patterns and simplified assumptions, leading to high uncertainties. This study aims to develop an efficient UAV-based flight strategy and quantification methodology for CO2 emissions from a resource recovery facility, considering real-time wind properties.

  In this study, an anemometer was integrated onto the UAV platform to measure wind vectors simultaneously with atmospheric CO2 concentrations at the boundary of the facility's stack. Efficient flight patterns were developed to balance plume detection and detailed concentration mapping within operational constraints such as UAV battery issue. The acquired data facilitated emission rate quantification using mass balance and inverse Gaussian methods, with results compared against Environmental Protection Agency (EPA) methods and Intergovernmental Panel on Climate Change (IPCC) Guidelines.

  Through 45 flight campaigns, an efficient flight strategy was developed, combining the advantages of non-hovering and hovering patterns. Non-hovering flights facilitated rapid scanning for initial plume localization, while hovering flights enabled high-accuracy concentration mapping within the detected plume boundaries. Previous studies indirectly measured plume locations using data from ground-based weather stations and onboard anemometers, which were then applied to plume rise models to predict plume positions. Compared to ground-based weather stations, onboard anemometers were able to more accurately predict the direction and altitude of the plume. In determining the altitude for hovering flights, wind data collected during non-hovering flights were applied to the plume rise model, which informed the subsequent hovering flight operations. This sequential approach of non-hovering scanning, hovering detection, and horizontal detection was employed in repeated experiments to comprehensively quantify the CO2 emissions.

  The emission quantification results from mass balance methods and inverse Gaussian methods demonstrated improved accuracy compared to existing methods, with lower mean absolute percentage error (MAPE). A comparative analysis between the mass balance and inverse Gaussian dispersion methods revealed that as the standard deviation of wind direction increased, indicating higher wind variability, the mass balance method outperformed the inverse Gaussian approach. This highlights the advantage of incorporating real-time wind properties in the mass balance method over the steady-state assumptions in the inverse Gaussian model. Furthermore, utilizing real-time wind data within the mass balance method improved emission estimates by an average of 34.66±39.96% compared to using mean wind assumptions. However, some errors still persisted, with lower peak concentrations contributing significantly to quantification errors.

  This research demonstrates the potential of UAV-based measurements and optimized flight strategies for accurate emissions quantification from point sources, addressing limitations in existing methods. The findings contribute to advancing atmospheric monitoring techniques and supporting effective implementation of emission mitigation policies.

  Specific objectives addressed in my paper include the following:

• Before the flight campaign, the performance evaluation of various sensors was conducted, including low-cost particulate matter sensors, when mounted on UAVs. While not the primary focus, the study includes an assessment of this particulate matter sensor in environmental chamber tests to assess their accuracy under diverse particle concentration scenarios, thereby highlighting their potential and limitations in field applications.
• The first part of the research is about the development of flight strategies for Uncrewed Aerial Vehicles (UAVs) to monitor greenhouse gas emissions from point sources. These strategies aim to enhance the accuracy and efficiency of data collection in various atmospheric conditions. 
• The second part of the research presents methodologies for quantifying CO2 emissions from municipal solid waste incinerators using UAVs. The research compares the mass balance method and the inverse Gaussian method to determine the most effective approach for accurate emission measurements.