Tracing Sources of PM10 in Central Taiwan by using Chemical Characteristics and Pb Isotope Ratios: Local Versus Long-range Transport

Central Taiwan is among the most heavily polluted regions in Taiwan because of a complex mixing of local emissions from intense anthropogenic activities with natural dust. Long-range transport (LRT) of pollutants from outside Taiwan also contributes critically to the deterioration of air quality, especially during the northeast monsoon season. To identify the sources of particulate matter <10 μm (PM10) in central Taiwan, this study performed several sampling campaigns, including three local events, one LRT event, and one dust storm event, during the northeast monsoon season of 2018/2019. The PM10 samples were analyzed for water-soluble ion and trace metal concentrations as well as Pb isotope ratios. Local river sand/soil samples were also collected and analyzed to constrain chemical/isotopic signatures of natural sources. The Pb isotope data were interpreted together with the enrichment factors of trace metals in PM10, and reanalysis data sets were used to delineate the sources of PM10 in central Taiwan. Our results suggested that PM10 was predominantly contributed by oil combustion and oil reneries during the local events (48%–88%), whereas the lowest contributions were from coal combustion (<21%). During periods of high wind speed, the contribution from natural sources increased signicantly from 7% to 31%. Moreover, the Pb isotopic signals of PM10 collected during the LRT event conrmed the impact of LRT from Mainland China, and the chemical characteristics of the PM10 signicantly differed from those of the PM10 collected during local events. This study demonstrates the robustness of using a combination of Pb isotopic compositions and enrichment factors in PM10 for source apportionment in complex and heavily polluted areas. had the Coal in and Pb and had the highest Pb isotope ratios. River was characterized by the unity of EFs and moderate Pb isotope ratios. This study also included a ternary mixing model of Pb isotopes for the source apportionment of PM 10 in central Taiwan. The contribution of PM 10 was dominated by oil combustion and oil reneries during local events (48%– 88%), whereas the contributions from coal combustion were lower (<21%). The contributions from river sand/soil were increased from 7% to 31% when the wind speed became high. All of these results were supported by the EFs of PM 10 and the reanalysis dataset. By combining Pb isotope ratios, EFs, and the reanalysis dataset, this study improved the constraints of PM 10 sources during different events. Moreover, we demonstrated a multi-tracer approach to understanding transportation and the contributions of PM 10 from various sources, which serves as a powerful tool for delineating complex atmospheres impacted by complex emission sources.


Introduction
Atmospheric particulate matter (PM) is derived from both natural and anthropogenic sources. It affects the atmospheric condition by reducing visibility and deteriorating air quality, and it even changes the Earth's surface albedo, which in uences regional climate changes [1,2]. In addition, high concentrations of PM, the main carrier of heavy metals, can cause lung function decline and increase the risk of respiratory and cardiovascular diseases [3][4][5]. For example, chronic exposure to Pb can be harmful to the neural system, leading to lower memory capabilities and even inducing cancer [6,7]. Thus, studying the sources of PM is vital for strategically mitigating such pollution.
In the recent three decades, emissions of anthropogenic PM and heavy metals into the atmosphere have been increasing in East Asia (e.g., PM 10 emission increased from 20Tg in 1990 to 27 Tg in 2010) as a result of growing economies and the rapid development of cities [8]. According to available global estimates, China is a major contributor of global PM and heavy metal emissions [8][9][10]. Despite a recent declining trend, atmospheric emissions are still the highest among Asian countries [9,11,12]. The in uences of industrial aerosols from China through long-range transport (LRT) have been recorded in many areas, such as South Korea [13], Japan [14,15], the North Paci c [16], and even the United States [17,18] and Canada [19]. Taiwan, as a neighboring country approximately 100 miles away, has continued to receive airborne PM from China (including particles derived from anthropogenic activities and dust storms), particularly during the northeastern monsoon season [20][21][22][23].
In addition to LRT, the coastal area of central Taiwan is also impacted by intense local anthropogenic activities, such as emissions from industrial parks, coal-red power plants, petrochemical complexes, and vehicle exhaust. Studies have conducted detailed investigations into the chemical characteristics of PM < 2.5 µm (PM 2.5 ) and PM < 10 µm (PM 10 ) in this region [24][25][26]. Hsu et al. [25] found that average PM 10 concentrations were high (76.4 ± Page 3/18 22 µg m − 3 ) during winter, with an annual PM 10 concentration of 52.4 ± 27.2 µg m − 3 . Moreover, extremely high PM 10 episodes (> 125 µg m − 3 ) have also occurred occasionally during the northeast monsoon season [24,27].
The PM 10 in central Taiwan was estimated using a receptor model derived from soil dust, crustal materials, coal combustion, oil combustion, and tra c emissions [25]. Although several studies have investigated the characteristics of PM and the contributions from local sources in central Taiwan, remarkably few has attempted to distinguish the chemical characteristics and Pb isotope ratios of PM 10 from local sources and LRT.
The Pb isotope ratio is an essential tool for tracking pollution sources in the atmosphere. Pb exists in both natural and anthropogenic sources and has four naturally occurring isotopes: 204 Pb is a non-radiogenic nucleus, whereas 206 Pb, 207 Pb, and 208 Pb are radiogenic end products from the decay series of 238 U, 235 U, and 232 Th, respectively.
Pb is produced and released into the environment through human activities that use various ore minerals (e.g., Pb ore) with distinct Pb isotope ratios formed under varying geological conditions. These isotope ratios do not fractionate during industrial processes, making them a promising tool for identifying pollution sources [28,29]. This technique has been deployed to study PM sources in the atmosphere [14,18,[30][31][32][33] as well as the sources of particles in ice cores [29,34,35].
Although some efforts have been made to study the Pb isotope ratios of PM in northern Taiwan [20,23], the information available on Pb isotopic variations in PM 10 remains limited. For instance, the constraints on local end-members and variations in isotopic signals over time and during different events have not been thoroughly studied. In the present study, PM 10 was collected in central Taiwan during events of different types, including local events, an LRT event, and a dust storm (Fig. 1). Details about the sampling sites and period, wind speed and direction, temperature, precipitation, and relative humidity are summarized in Table 1. The main purposes of this study were, for the rst time, (1) to characterize the chemical properties and Pb isotope ratios of local pollutants under different wind conditions (on daily basis) in central Taiwan,and (2) to identify the possible source of PM 10 during each event using the chemical characteristics and Pb isotope ratios. Pb isotope compositions of potential PM 10 sources in Taiwan

Anthropogenic sources
The major anthropogenic sources of Pb are emissions from oil combustion or oil re neries, coal combustion, and high-temperature industrial processes (e.g., steel plants). Taiwan is an island with limited energy resources and mainly relies on imported resources from other countries (up to 98%). According to Taiwan's Bureau of Energy, oil, coal, and natural gas accounted for 48%, 29%, and 15%, respectively, of Taiwan's total primary energy consumption in 2018 [37]. Since 2012, Taiwan has been importing crude oil from Saudi Arabia (31%), Kuwait (21%), and others in similar proportions. Taiwan has been imported coal mainly from Australia (35-50%) and Indonesia ( However, these studies have not provided a quantitative estimation of the contribution of local dust to PM 10 in central Taiwan. Recently, Hsu et al. [25] attempted to estimate this contribution by using a chemical mass balance model. Herein, we provide an alternative method for calculating the contribution based on the Pb isotope ratios of PM 10 and potential end-members. To constrain the isotopic signals of natural materials and further estimate their contribution to PM 10 , we analyzed river sand/soil samples from the Choshui River catchment ( Fig. 1) for their metal concentrations and Pb isotope ratios. Because river sand/soil in downstream areas could be in uenced by human activities, sampling was conducted in both the downstream (group A) and upstream (group B) segments of the Choshui River catchment (Fig. 1). The EFs of this river sand/soil were calculated relative to the upper continental crust [42], and the results are presented in Supplementary Fig. S2. Although the EFs of Cd were higher (EF Cd = 9.9 ± 8.6) downstream than upstream (EF Cd = 0.5 ± 0.2), most EF values were close to unity for river sand/soil collected from the upstream and downstream segments of the Choshui River catchment. This indicates that the element contents of this river's sand and soil are similar to the average composition in the upper continental crust. All the river sand/soil samples analyzed in this study exhibited higher   (Table 3). These ratios were close to or within the range of those reported for commercial oil products in Taiwan, suggesting that oil combustion and oil re neries are the main sources of Pb during low-wind-speed conditions. The EFs of trace metals in PM 10 at each site were plotted against distance to a petrochemical complex to illustrate the spatial distribution pattern of trace metals in this region. As depicted in Fig. 3, the highest EFs of V (> 78), Ni (> 108), Cu (> 1,056), and Mo (> 1,478) were found at sites #4 and #5, which were the sites nearest to the petrochemical complex. V and Ni have been widely used as indicator elements of oil combustion [9,43], and Mo is regarded as a tracer for heavy oil combustion [44,45]. The combination of low Pb isotope ratios and elevated EFs of V, Ni, and Mo con rmed that emissions from oil combustion were the primary sources of PM 10 during the low-wind-speed event. Under this low-wind-speed condition, PM 10 at sites #4 and #6 exhibited slightly elevated Pb isotope ratios, indicating the presence of Pb sources other than oil combustion (Fig. 2a). For instance, site #6, the southernmost site in the study region, exhibited high Pb isotope ratios ( 206 Pb/ 207 Pb = 1.161 and 208 Pb/ 207 Pb = 2.434) and relatively high EFs for Mn, Zn, As, Cd, and Pb (Fig. 3). An earlier study found that emissions from steel industry was the dominant source of PM 2.5 during the cold season in Chiayi County, Taiwan (where site #6 is located) because of the high content of Ca, Ti, Pb, and Fe [46]. After investigating the metal compositions of lterable stack total suspended particles emitted from industrial activities in Taiwan Fig. 2a). Relative to the low-wind-speed event, the EFs were reduced during the moderatewind-speed event and reached their lowest values during the high-wind-speed event (Fig. 3). The higher Pb isotope ratios suggested that Pb originated from either river sand/soil or coal combustion because of the overlap of ratios between Indonesian coal and local river sand/soil. This overlap of isotopic signals may hamper the explanation of the source. However, these two potential sources can be differentiated using EF values. As illustrated in Figs. 2a and 3, PM 10 at sites #4 and #5 during the high-wind-speed event had the highest Pb isotope ratios and the lowest EF of Pb (< 10), indicating that these high ratios were mainly contributed by river sand/soil rather than by other sources. The high concentrations of Al (3,107 and 3,656 ng/m 3 , respectively) in PM 10 at these two sites also support the increased contribution of crustal materials during the high-wind-speed event, con rming that natural dust is the dominant source of PM 10 in central Taiwan during the northeast monsoon season.
In summary, PM 10 Fig. S4). The European Centre for Medium-Range Weather Forecasts (ECMWF) Atmospheric Composition Reanalysis 4 (EAC4) reanalysis dataset reveals that a nitric acid plume formed in Southeast China and gradually moved northeastward to Taiwan (Supplementary Fig. S5). In this study, PM 10 samples were collected on October 4, 2019. As discussed previously, the high concentrations of ionic species (sulfate, nitrate, and ammonium) and heavy metals (V, Ni, As, Cd, and Pb) indicated that the air quality became increasingly polluted (Table S1). PM 10  According to relevant studies, coal combustion became the major source of Pb in the atmosphere after the phasing-out of leaded gasoline, leading to more radiogenic isotope ratios being observed for particles in the atmosphere in China [50,51]. High Pb isotopic ratios were also reported for aerosols in areas with high levels of industrial emissions [52][53][54][55][56] (Fig. 2b). Thus, the high Pb isotope ratios in our PM 10 Fig. 2b).
During the dust storm event, the air mass back trajectory revealed that the air parcels primarily originated from Northern China and passed through Shanghai, a megacity and economic center in East China, before reaching Taiwan ( Supplementary Fig. S4). The dust-mixing ratio estimated by the EAC4 reanalysis indicated that the dust storm was mainly derived from Northern China, with the dust plume gradually being transported to East China ( Supplementary Fig. S6). When the dust storm passed through Shanghai, a nitric acid plume derived from East China gradually moved southward to the Taiwan Strait ( Supplementary Fig. S7). During this dust event, PM 10 samples were collected for 2 days consecutively at site #5. The results indicated that PM 10 Fig. 2b). The EFs of most metals in the PM 10 collected during the dust storm event were low (< 343, Fig. 3), indicating that the PM 10 was a mixture of pollutants and natural dust. This was further supported by the elevated Estimating the relative contribution to PM 10 Because distinctive Pb isotope ratios were found for each end-member, we were able to estimate their relative contributions to PM 10 . We adopted a ternary mixing model to calculate the relative contributions to PM 10 using the following equations: where R PM10 is the observed 206 Pb/ 207 Pb (or 208 Pb/ 207 Pb) ratio of PM 10 , Ri is the assumed 206 Pb/ 207 Pb (or 208 Pb/ 207 Pb) ratio for each end-member, and is the fraction of the contribution of each end-member.
The end-members discussed in this study (i.e., oil combustion and oil re neries, coal combustion, and river sand/soil) were assigned accordingly, as shown in Fig. 2. We assumed that the end-member oil combustion and oil re neries were the average isotopic signatures of the oil products (gasoline and diesel) investigated in Taiwan and assumed that the end-member coal combustion was the average isotopic signature of Australian coal since Australian coal occupied half of the coal imported into Taiwan. We also assumed that the end-member natural material was the average isotopic signature of river sand/soil collected in this study.
A Bayesian mixing model, MixSIAR [57,58], was employed to calculate each end-member contribution to PM 10 .
This model incorporates the uncertainties of the isotopic signatures for each source, and it was successfully applied to assess Pb sources in isotopic mixtures quantitatively [59]. The model results revealed the contribution from oil combustion to be predominant (48-88%) during the local events, whereas coal combustion made a minor contribution (< 21%) in central Taiwan (Supplementary Table S2, Fig. 4). On the other hand, contributions from river sand/soil increased signi cantly from 7-66% during the high-wind-speed events and were highly variable among sites. The high proportions of oil combustion suggested an approach for mitigating emissions in the future.
During the LRT and dust storm events, the Pb isotope ratios of PM 10 might have been overprinted by pollutants derived from China (as shown in Fig. 2b), and further complicated the contribution estimations. In such cases, PM 10 was assumed to be dominated by the LRT of pollutants, and the approach presented here should be able to delineate the PM 10 sources if the primary sources of Pb can be better constrained.

Concluding Remarks
This study investigated the chemical characteristics and sources of PM 10 during various events of the 2018/2019 northeast monsoon season in central Taiwan. High concentrations of ionic species were found during the lowwind-speed local event, suggesting that local pollutants accumulated when the atmospheric condition stagnated.
By contrast, high concentrations of crustal elements (e.g., Al, Fe, and Ti) were found during the high-wind-speed local event, suggesting the enhanced in uence of river soil/sand. By employing EFs and Pb isotope ratios, we revealed an evident variation in the contributions from local sources. For instance, oil combustion was enriched in V and Ni and had the lowest Pb isotope ratios. Coal combustion was enriched in As, Sb, and Pb and had the highest Pb isotope ratios. River sand/soil was characterized by the unity of EFs and moderate Pb isotope ratios.
This study also included a ternary mixing model of Pb isotopes for the source apportionment of PM 10 in central Taiwan. The contribution of PM 10 was dominated by oil combustion and oil re neries during local events (48%-88%), whereas the contributions from coal combustion were lower (<21%). The contributions from river sand/soil were increased from 7% to 31% when the wind speed became high. All of these results were supported by the EFs of PM 10 and the reanalysis dataset. By combining Pb isotope ratios, EFs, and the reanalysis dataset, this study improved the constraints of PM 10 sources during different events. Moreover, we demonstrated a multi-tracer approach to understanding transportation and the contributions of PM 10 from various sources, which serves as a powerful tool for delineating complex atmospheres impacted by complex emission sources.

Methods
Sampling site and PM 10 collection PM 10 sampling was conducted at six sites located in the rural-fringe areas of Changhua, Yunlin, and Chiayi Counties in central Taiwan (Fig. 1), where the air quality is reported to be among the worst in Taiwan [60]. Several obvious anthropogenic sources of PM exist, such as a coal-red plant (one of the largest in the world) in Taichung City, Changhua coastal industrial park in Changhua County, and a petrochemical complex (including a coal-red power plant) in Yunlin County. Moreover, river dust in the Choshui River catchment is the major source of PM derived from natural materials. Details about the sampling sites, sampling period, wind speed, wind direction, temperature, precipitation, and relative humidity are summarized in Table 1 Nov 25, 2018. After collection, the lter was transferred into a plastic envelope and delivered to the conditioning room within a few hours. Furthermore, a total of 13 river dust samples were collected from eight sites (Fig. 1) to constrain the chemical and isotope signals of the natural source. The dust samples were collected in both upstream and downstream segments of the Choshui River catchment with plastic bags, dried at 45ºC in an oven, ground and sieved through a 75 µm sieve, and stored in a desiccator.

Chemical analysis
The PTFE lter was measured for particle mass concentration with a microbalance after equilibration at 25 ± 1.5ºC under a relative humidity of 40% ± 5% for 24 h. Next, 1/9 of the lter was cut using a ceramic cutter, and the aliquot was then digested with an acid mixture of 9 mL of concentrated HNO 3 , 3 mL of concentrated HCl, and 3 mL of concentrated HF using a microwave (CEM Corp., Matthews, NC, USA). A two-stage heating procedure was employed; rst, the mixture was heated to 170ºC over 20 min and maintained at this temperature for 10 min, followed by another round of heating to 200ºC over 7 min and maintaining for 10 min. After cooling, the solution was transferred into a PFA beaker and evaporated. The dried sample was re-dissolved with an acid mixture of 2 mL of concentrated HNO 3 and 1 mL of concentrated HCl, and nally diluted to 50 mL with Milli-Q water.
Ultrapure concentrated acids and Milli-Q water were used for sample preparation in this study. For dust samples, 100 mg of dust was weighed and digested following the aforementioned procedure. A total of 23 elements, including major and trace metal concentrations, were analyzed by using Q-ICP-MS (Agilent 7700X) with internal standards (Sc, Y, Rh, Tb, Lu, and Bi) to monitor the instrumental drift and matrix effect (NIEA M105); the analytical precision was greater than 10% (RSD). The method detection limit for each element was determined by eld blanks, as presented in Supplementary

Pb isotope analysis
An aliquot of the digested sample was transferred into an acid-cleaned PFA beaker. This solution was evaporated and re-dissolved with 2 mL of 2M HNO 3 and 0.07M HF. The Pb fraction was extracted using Sr-spec ion exchange resin (Eichrom Technologies Inc., Lisle, IL, USA) following the steps modi ed from Pin et al. [61]. An international reference material (NIST 1648a) was also processed for each puri cation batch to assess the column chemistry performance. Pb was puri ed under a Class-10 laminar ow bench in a Class-10,000 clean room. The total procedural blanks for Pb were < 80 pg. Pb isotope ratios were measured using HR-MC-ICP-MS (Neptune PLUS, Thermo-Fisher Scienti c) at the Institute of Earth Sciences, Academia Sinica (AS-IES). Standards and samples were doped with thallium (Tl) (NIST 997; 203 Tl/ 205 Tl ratio of 0.418673) to correct for instrumental mass fractionation [62]. Two standard reference materials (NIST SRM 981 and NIST SRM 1648a) were analyzed to assess the accuracy and long-term precision of the analytical protocol developed at AS-IES. The measured Pb isotope ratios for these international reference materials were in good agreement with the recommended values and are listed in Supplementary Table S6. The analytical uncertainties (2SD) for 206 Pb/ 207 Pb and 208 Pb/ 207 Pb were ± 0.0002 (n = 69) and ± 0.0003 (n = 69), respectively. Pb isotope ratios were reported as 206 Pb/ 207 Pb and 208 Pb/ 207 Pb in this study.

Enrichment factor
The EF has been widely used to examine the contributions from natural and anthropogenic sources in aerosols [63][64][65]. The EF of elements was calculated using Eq. (3): where (Xi/Al) PM is the concentration ratio of element X to Al in PM, and (Xi/Al) Crust is the concentration ratio of element X to Al in the upper continental crust [42]. In general, EF values close to unity indicate the predominance of crustal sources; EF values ≥ 5 indicate a signi cant contribution from noncrustal sources; and EF values higher than 10 indicate essentially anthropogenic origins [24,64,66].

Reanalysis datasets
The EAC4 global reanalysis dataset provided by the Copernicus Atmosphere Monitoring Service (CAMS) was applied to provide additional constraints on the sources and transportation of pollutants in this study. EAC4 reanalysis combines model data with global observations into a globally complete and consistent dataset; the dataset used in this study was generated using CAMS information (2020) [67]. The spatial resolution of the dataset was 0.75° latitude by 0.75° longitude, with a temporal resolution of 3 h.
In addition, air mass back trajectory analysis was used to track the origins of the air parcels transported to the study sites. Back trajectories were calculated using the HYSPLIT model maintained by the US National Oceanographic and Atmospheric Administration with a spatial resolution of 1° latitude by 1° longitude in the meteorological dataset [68].