Identifying and quantifying PM2.5 pollution episodes with a fusion method of moving window technique and constrained Positive Matrix Factorization

Improvements in source apportionment of multiple time-resolved PM2.5 inorganic and organic speciation measurements using constrained Positive Matrix Factorization

The equation of Positive Matrix Factorization (PMF) has been modified to resolve multiple time resolution inputs and applied in numerous field studies. The refined modeling results provide a solution with an increased number of factors and enriched profile features. However, the incorporation of low time-resolved data may retrieve unfavorable mixed factor profiles, introducing high uncertainties into the PMF solution computations. To address this issue, a dual-stage PMF modeling procedure with predefined constraints was proposed. Multiple time-resolved PM2.5 inorganic and organic speciation measurements were collected from autumn of 2022 to summer of 2023 in Taipei, Taiwan. Without using the proposed approach, a mixed factor of vehicle/biomass burning and an unphysically meaningful factor of sodium ion- and ammonium ion-rich were identified. After implementing the proposed approach, a refined number of eight factors with separated and reasonable profiles were retrieved. Over the sampling period, the largest contributor to PM2.5 and organic carbon was vehicle (contribution = 26% and 47%, respectively), while those for secondary inorganic aerosols of SO42-, NO3-, and NH4+ were industry (27%, 25%, and 31%, respectively), highlighting the importance of regulating these two sources. The low vehicle contribution to NO3- may be due to time-lag effects from gas-to-particle conversion, which led to different temporal patterns between NO3- and primary species. Addressing this issue is crucial in future studies for better apportionment of secondary aerosols.

Improving assessment quality of soil natural attenuation capacity at the point and regional scales

Soil natural attenuation capacity (NAC) is an important ecosystem service that maintains a clean environment for organisms in the soil, which in turn supports other services. However, spatially varying indicator weights were rarely considered in the traditionally-used soil NAC assessment model (e.g., ecosystem-service performance model) at the point scale. Moreover, in the spatial simulation of soil NAC, the traditionally-used geostatistical models were usually susceptible to spatial outliers and ignored valuable auxiliary information (e.g., land-use types). This study first proposed a novel soil NAC assessment method based on the ecosystem-service performance model and moving window-entropy weight method (MW-EW) (NACMW-EW). Next, NACMW-EW was used to assess soil NAC in a typical area in Guixi City, China, and further compared with the traditionally-used NACtra and NACEW. Then, robust sequential Gaussian simulation with land-use types (RSGS-LU) was established for the spatial simulation of NACMW-EW and compared with the traditionally-used SGS, SGS-LU, and RSGS. Last, soil NAC's spatial uncertainty was evaluated based on the 1000 realizations generated by RSGS-LU. The results showed that: (i) MW-EW effectively revealed the spatially varying indicator weights but EW couldn't; (ii) NACMW-EW obtained more reasonable results than NACtra and NACEW; (iii) RSGS-LU (RMSE = 0.118) generated higher spatial simulation accuracy than SGS-LU (RMSE = 0.123), RSGS (RMSE = 0.132), and SGS (RMSE = 0.135); and (iv) the relatively high (P[NACMW-EW(u) > 0.57] ≥ 0.95) and low (P[NACMW-EW(u) > 0.57] ≤ 0.05) threshold-exceeding probability areas were mainly located in the south and east of the study area, respectively. It is concluded that the proposed methods were effective tools for soil NAC assessment at the point and regional scales, and the results provided accurate spatial decision support for soil ecosystem service management.

Spatial characterization of HCHO and reapportionment of its secondary sources considering photochemical loss in Taiyuan, China

Formaldehyde (HCHO) plays an important role in atmospheric ozone (O₃) formation. To accurately identify the sources of HCHO, carbonyls and volatile organic compounds (VOCs) were measured at three urban sites (Taoyuan, TY-U; Jinyuan, JY-U; Xiaodian, XD-U) and a suburban site (Shanglan, SL-B) in Taiyuan during a high O₃ period (from July 20 to August 3, 2020). The average mixing ratio of HCHO at XD-U (8.1 ± 2.8 ppbv) was comparable to those at TY-U (7.4 ± 2.1 ppbv) and JY-U (7.0 ± 2.3 ppbv) but higher (p < 0.01) than that at SL-B (4.9 ± 2.3 ppbv). HCHO contributed to 54.3-59.9 % of the total ozone formation potentials (OFPs) of non-methane hydrocarbons (NMHCs) at four sites. The diurnal variation of HCHO concentrations reached a peak value at 12:00-15:00, which may be attributed to the strong photochemical reaction. To obtain more accurate source results of HCHO under the condition of photochemical loss, the initial concentrations of NMHCs were estimated based on photochemical age parameterization and incorporated into the positive matrix factorization (PMF) model (termed IC-PMF). According to the IC-PMF results, secondary formation (SF) contributed the most to HCHO at XD-U (35.6 %) and SL-B (25.1 %), whereas solvent usage (SU) (40.9 %) and coking sources (CS) (36.0 %) were the major sources at TY-U and JY-U, respectively. Compared to the IC-PMF, the conventional PMF analysis based on the observed data underestimated the contributions of SU (100.5-154.2 %) and biogenic sources (BS) (28.5-324.7 %). Further reapportionment of secondary HCHO by multiple linear regression indicated that SU dominated the sources of HCHO at SL-B (28.3 %) and TY-U (41.7 %), while industrial emissions (IE) and CS contributed the most to XD-U (26.6 %) and JY-U (43.0 %) in Taiyuan from north to south, respectively.

County-Based PM2.5 Concentrations’ Prediction and Its Relationship with Urban Landscape Pattern

Satellite top-of-atmosphere (TOA) reflectance has been validated as an effective index for estimating PM2.5 concentrations due to its high spatial coverage and relatively high spatial resolution (i.e., 1 km). For this paper, we developed an emsembled random forest (RF) model incorporating satellite top-of-atmosphere (TOA) reflectance with four categories of supplemental parameters to derive the PM2.5 concentrations in the region of the Yangtze River Delta-Fujian (i.e., YRD-FJ) located in east China. The landscape pattern indices at two levels (i.e., type level and overall level) retrieved from 3-year land classification imageries (i.e., 2016, 2018, and 2020) were used to discuss the correlation between county-based PM2.5 values and landscape pattern. We achieved a cross validation R2 of 0.91 (RMSE = 9.06 μg/m3), 0.89 (RMSE = 10.19 μg/m3), and 0.90 (RMSE = 8.02 μg/m3) between the estimated and observed PM2.5 concentrations in 2016, 2018, and 2020, respectively. The PM2.5 distribution retrieved from the RF model showed a trend of a year-on-year decrease with the pattern of “Jiangsu > Shanghai > Zhejiang > Fujian” in the YRD-FJ region. Our results also revealed that the landscape pattern of farmland, water bodies, and construction land exhibited a highly positive relationship with the county-based average PM2.5 values, as the r coefficients reached 0.74 while the forest land was negatively correlated with the county-based PM2.5 (r = 0.84). There was also a significant correlation between the county-based PM2.5 and shrubs (r = 0.53), grass land (r = 0.76), and bare land (r = 0.60) in the YRD-FJ region, respectively. Three landscape pattern indices at an overall level were positively correlated with county-based PM2.5 concentrations (r = 0.80), indicating that the large landscape fragmentation, edge density, and landscape diversity would raise the PM2.5 pollution in the study region.