Source Apportionment and Health Risk Assessment of PM2.5-bound Elements in Windsor, Ontario, Canada
Bibliographic record
Abstract
Windsor, Ontario, Canada frequently experiences poor air quality due to local emissions and transboundary pollution inputs. This study investigates (1) the ambient concentration levels of PM2.5 mass and PM2.5-bound elements, (2) major sources of PM2.5-bound elements and source contributions, (3) sensitivity of Positive Matrix Factorization (PMF) modeling to input concentration data, (4) human health risks from inhalation exposure to PM2.5-bound elements, (5) major contributors to elemental concentrations and health risks, and (6) temporal variability of concentrations, source contributions, and health risks. Hourly concentrations of PM2.5 mass, black carbon (BC), brown carbons (BrCs), and 24 PM2.5-bound elements were continuously monitored at the Windsor West station during April 2021 ─ April 2023. USEPA’s PMF model was utilized to identify sources and quantify their contributions. USEPA’s health risk assessment approach was used to estimate total and individual lifetime cancer risks (CRs) and chronic hazard quotients (HQs) due to inhalation exposure to six and eleven elements, respectively. The two-year average PM2.5 mass concentration was 9.2 μg/m³, slightly exceeding the Canadian Ambient Air Quality Standards of 8.8 μg/m³. The total elemental concentration was 1.4 μg/m³, which accounted for 15% of PM2.5 concentration. Five PM2.5-bound element sources were resolved by the PMF modeling, (1) coal/heavy oil burning (33% of total elemental concentrations), (2) vehicular exhaust (28%), (3) metal processing (20%), (4) crustal dust (16%), and (5) vehicle tire and brake wear (3%). The three traffic-related sources (i.e., vehicular exhaust, crustal dust, and vehicle tire and brake wear) and two industrial sources (i.e., coal/heavy oil burning and metal processing) contributed nearly equally (47% vs. 53%) to total PM2.5-bound element concentration. The sensitivity analysis of the PMF modeling yielded: (1) Treatment of concentrations below the method detection limits (MDLs), i.e., leaving as is vs. replacing with ½ MDLs had negligible effects on source identification, source contribution estimations, and model performance. (2) The modeling results are not sensitive to excluding BrCs concentrations, because they are strongly correlated with BC concentrations, which were already included in the PMF modeling. (3) Conducting PMF separately for episodic events is beneficial for identifying unique sources associated with these events and improving model performance. Both the total CR (4.1×10⁻⁵) and total HQ (0.82) remained below the USEPA acceptable thresholds of 10⁻⁴ and 1, respectively. Among the five sources identified by PMF modeling, metal processing was the largest contributor to total CR (52%) and total HQ (60%), followed by coal/heavy oil combustion (19% and 16%) and vehicular exhaust (19% and 12%), and the remaining two sources, crustal dust and vehicle tire and brake wear (10% and 12%). The seasonal PM2.5 concentrations were highest in Winter, followed by Summer, Spring, and Fall. The seasonal variability of total CR and HQ was small. The hour-of-day PM2.5 showed higher concentration in the early morning and lower in the afternoon. Neither the seasonal nor the diurnal trends of most elements are similar to that of the PM2.5 mass, calling for the monitoring of PM2.5-bound elements. Among the 24 PM2.5-bound elements, the top five elements (S, Si, Fe, K, and Ca, from most to least abundant) combined contributed 95% of total elemental concentrations. However, the elements contributing most to concentrations are not the largest contributors to health risks due to different toxicities among the elements. Specifically, Cd was the largest contributor to both total cancer risk (62%) and total hazard quotient (73%) due to its high inhalation unit risk and low reference concentration, while Cd only contributed 0.5% of total elemental concentrations. Metal processing as a source contributed over half of total CR and total HQ but only 20% of elemental concentrations. Emission control measures should consider major contributors to ambient concentrations and those to health risks.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".