Use of multi-pollutant air sensor data and geometric non-negative matrix factorization for source apportionment of air pollution burden in Curtis Bay, Baltimore, USA
🤖 AI Summary
This study addresses the unreliability and poor scalability of source apportionment for air pollution in Curtis Bay, Baltimore, USA. We propose a multi-pollutant source apportionment framework based on geometric nonnegative matrix factorization (NMF), leveraging high-temporal-resolution sensor data (PM₂.₅, PM₁₀, black carbon, CO, NOₓ) integrated with regression modeling and sensitivity analysis to achieve stable and identifiable source separation. The method overcomes key limitations of conventional NMF—including non-uniqueness and poor reproducibility—under large-scale, real-world monitoring conditions, significantly enhancing the geometric identifiability and robustness of the source contribution matrix. Three physically interpretable sources are consistently resolved: Source 1 (coal terminal–dominated, primarily emitting fine particles), Source 2 (traffic-dominated, contributing black carbon and gaseous pollutants), and Source 3 (coal terminal–associated coarse particles). Results are validated via wind-direction correlation and case-study analysis of representative pollution events, demonstrating strong environmental interpretability and actionable insights for regulatory policy.
Application Category
📝 Abstract
Air sensor networks provide hyperlocal, high temporal resolution data on multiple pollutants that can support credible identification of common pollution sources. Source apportionment using least squares-based non-negative matrix factorization is non-unique and often does not scale. A recent geometric source apportionment framework focuses inference on the source attribution matrix, which is shown to remain identifiable even when the factorization is not. Recognizing that the method scales with and benefits from large data volumes, we use this geometric method to analyze 451,946 one-minute air sensor records from Curtis Bay (Baltimore, USA), collected from October 21, 2022 to June 16, 2023, covering size-resolved particulate matter (PM), black carbon (BC), carbon monoxide (CO), nitric oxide (NO), and nitrogen dioxide (NO2). The analysis identifies three stable sources. Source 1 explains > 70% of fine and coarse PM and ~30% of BC. Source 2 dominates CO and contributes ~70% of BC, NO, and NO2. Source 3 is specific to the larger PM fractions, PM10 to PM40. Regression analyses show Source 1 and Source 3 rise during bulldozer activity at a nearby coal terminal and under winds from the terminal, indicating a direct coal terminal influence, while Source 2 exhibits diurnal patterns consistent with traffic. A case-study on the day with a known bulldozer incident at the coal terminal further confirms the association of terminal activities with Sources 1 and 3. The results are stable under sensitivity analyses. The analysis demonstrates that geometric source apportionment, paired with high temporal resolution data from multi-pollutant air sensor networks, delivers scalable and reliable evidence to inform mitigation strategies.
Problem
Research questions and friction points this paper is trying to address.
Identifying pollution sources using multi-sensor data and geometric factorization method
Analyzing air quality in Curtis Bay with 451,946 sensor records across pollutants
Attributing pollution to coal terminal activities and traffic patterns reliably
Innovation
Methods, ideas, or system contributions that make the work stand out.
Geometric non-negative matrix factorization for source apportionment
Analyzing multi-pollutant sensor data with high temporal resolution
Identifying pollution sources through geometric source attribution matrix
🔎 Similar Papers
No similar papers found.
