PhD Thesis Presentation - An Integrated Framework for Atmospheric VOC Monitoring: From Reliable Sensor Calibration and Precise Smart Sampling to Efficient Source Apportionment
Supporting the below United Nations Sustainable Development Goals:支持以下聯合國可持續發展目標:支持以下联合国可持续发展目标:
Volatile organic compounds (VOCs) are critical ozone precursors. While low-cost Photoionization Detectors (PIDs) offer high-frequency screening, their field deployment is hindered by environmental interferences, lack of chemical speciation, and spatial limitations in source tracing. To overcome these bottlenecks, this dissertation establishes an integrated framework spanning "Precise Sampling — Reliable Calibration — Efficient Source Apportionment."
First, to achieve reliable calibration, a physics-driven Inter-Sensor Consistency Model (ISCM) was developed to eliminate temperature and humidity interferences. By decoupling baseline and sensitivity corrections, the ISCM ensured robust in-situ network consistency and long-term data reliability under fluctuating ambient conditions.
Second, for efficient source apportionment, an Enhanced Nonparametric Trajectory Analysis (NTA) framework was constructed. By incorporating Kriging spatial interpolations for wind and concentration fields alongside a novel spatial weight-reducing function, the model eliminated traditional spatial blurring. Its localized source-pointing accuracy was validated via cross-validation and independent receptor chemical fingerprinting.
Third, to enable precise sampling, a sensor-triggered smart canister sampler was engineered. Integrating real-time O3 and TVOC sensors with active humidity controls, the system successfully captured transient, high-value episodic plumes. Offline analysis confirmed significantly elevated Ozone Formation Potential (OFP) during triggered events compared to traditional scheduled sampling.
Finally, a future roadmap is presented, featuring PID-methane sensor fusion and dual-track portable gas chromatography (GC) integration. Overall, this dissertation successfully bridges low-cost sensor networks with high-fidelity laboratory analysis, providing a scalable, scientifically rigorous toolkit for targeted VOC mitigation.