Modeling atmospheric microplastic cycle by GEOS-Chem: An optimized estimation by a global dataset suggests likely 50 times lower ocean emissions

Abstract

•A new atmospheric transport model for microplastics is developed based on GEOS-Chem•Ocean has a 50 times smaller emission (171 [308–764] Gg year−1) than previously thought•Almost all (99%) the microplastic particles are emitted as coarse aerosols (70 μm)•Net land-to-ocean transport is found (25 Gg year−1), much smaller than river discharge Microplastics are tiny plastic particles that are widespread in our environment and pose potential risks to ecosystems and human health. The atmosphere plays a vital role in moving microplastics over long distances and continuously engages in exchanges with the land and ocean. However, identifying the sources of atmospheric microplastics remains challenging. Previous research suggested that the ocean is the main source, but our study, which combines global data and atmospheric models, reveals that the ocean contributes less than initially thought. Road-related sources, such as tire and brake wear and poorly managed plastic waste, also contribute significantly. Our findings underscore the necessity of regulating litter and dumping of plastic waste near roads. Further data collection and laboratory research are needed to better understand the atmospheric microplastic cycle. The atmosphere plays a vital role in microplastic (MP) transport, facilitating continuous exchanges with land and ocean. However, the sources of atmospheric MP remain unclear. Previous studies suggested that the ocean is the primary source, with global emissions reaching up to 8,600 Gg year−1. Here, we use global atmospheric abundance data, a newly developed atmospheric model, and optimal estimation to constrain the atmospheric sources. We find that the global atmospheric MP emissions are 324 (73–1,450) Gg year−1. The ocean source is estimated to have a much smaller global emission (171 [38–764] Gg year−1] than previously believed, followed by road-related sources (115 [26–513] Gg year−1) including the suspension of tire and brake wears and mismanaged plastic waste. We simulate a net land-to-ocean transport by the atmosphere (25 Gg year−1). This highlights the importance of controlling terrestrial sources, and more data are needed to improve our understanding of the atmospheric MP cycle. The atmosphere plays a vital role in microplastic (MP) transport, facilitating continuous exchanges with land and ocean. However, the sources of atmospheric MP remain unclear. Previous studies suggested that the ocean is the primary source, with global emissions reaching up to 8,600 Gg year−1. Here, we use global atmospheric abundance data, a newly developed atmospheric model, and optimal estimation to constrain the atmospheric sources. We find that the global atmospheric MP emissions are 324 (73–1,450) Gg year−1. The ocean source is estimated to have a much smaller global emission (171 [38–764] Gg year−1] than previously believed, followed by road-related sources (115 [26–513] Gg year−1) including the suspension of tire and brake wears and mismanaged plastic waste. We simulate a net land-to-ocean transport by the atmosphere (25 Gg year−1). This highlights the importance of controlling terrestrial sources, and more data are needed to improve our understanding of the atmospheric MP cycle. Plastics are durable, versatile, and ubiquitous in modern life. While their global production has increased from 1,700 Gg year−1 in 1950 to 367,000 Gg year−1 in the 2020s, plastic waste management has not kept up, resulting in more than 40,000–80,000 Gg mismanaged plastic waste (MMPW) globally per year.1Borrelle S.B. Ringma J. Law K.L. Monnahan C.C. Lebreton L. McGivern A. Murphy E. Jambeck J. Leonard G.H. Hilleary M.A. et al.Predicted growth in plastic waste exceeds efforts to mitigate plastic pollution.Science. 2020; 369: 1515-1518https://doi.org/10.1126/science.aba3656Crossref PubMed Google Scholar,2Lebreton L. Andrady A. Future scenarios of global plastic waste generation and disposal.Palgr Commun. 2019; 56https://doi.org/10.1057/s41599-018-0212-7Crossref Scopus (814) Google Scholar MMPW is defined as the plastic waste not properly recycled, incinerated, or buried in landfills but which enters the environment. MMPW is broken down or disaggregated to much smaller microplastics (MPs) (approximately 1 μm–5 mm) or nanoplastics (typically <1,000 nm),3Huang D. Tao J. Cheng M. Deng R. Chen S. Yin L. Li R. Microplastics and nanoplastics in the environment: macroscopic transport and effects on creatures.J. Hazard Mater. 2021; 407: 124399https://doi.org/10.1016/j.jhazmat.2020.124399Crossref Scopus (140) Google Scholar which could cause a long-term ecological impact on both terrestrial and aquatic lives,4Wang W. Gao H. Jin S. Li R. Na G. The ecotoxicological effects of microplastics on aquatic food web, from primary producer to human: a review.Ecotox Environ Safe. 2019; 173: 110-117https://doi.org/10.1016/j.ecoenv.2019.01.113Crossref PubMed Scopus (286) Google Scholar,5Chae Y. An Y.J. Nanoplastic ingestion induces behavioral disorders in terrestrial snails: trophic transfer effects via vascular plants.Environ Sci-Nano. 2020; 7: 975-983https://doi.org/10.1039/c9en01335kCrossref Scopus (94) Google Scholar act as a vector for contaminants,6Ziccardi L.M. Edgington A. Hentz K. Kulacki K.J. Kane Driscoll S. Microplastics as vectors for bioaccumulation of hydrophobic organic chemicals in the marine environment: a state-of-the-science review.Environ. Toxicol. Chem. 2016; 35: 1667-1676https://doi.org/10.1002/etc.3461Crossref PubMed Scopus (297) Google Scholar and even pose a potential threat to human health.7Rubio L. Marcos R. Hernández A. Potential adverse health effects of ingested micro- and nanoplastics on humans. Lessons learned from in vivo and in vitro mammalian models.J. Toxicol. Env. Heal. B. 2020; 23: 51-68https://doi.org/10.1080/10937404.2019.1700598Crossref PubMed Scopus (116) Google Scholar The atmosphere plays an important role in the transport of plastics in the environment, especially MPs smaller than 70 μm.8Brahney J. Mahowald N. Prank M. Cornwell G. Klimont Z. Matsui H. Prather K.A. Constraining the atmospheric limb of the plastic cycle.Proc. Natl. Acad. Sci. USA. 2021; 118e2020719118https://doi.org/10.1073/pnas.2020719118Crossref Scopus (140) Google Scholar,9Allen S. Allen D. Baladima F. Phoenix V.R. Thomas J.L. Le Roux G. Sonke J.E. Evidence of free tropospheric and long-range transport of microplastic at Pic du Midi Observatory.Nat. Commun. 2021; 12: 7242https://doi.org/10.1038/s41467-021-27454-7Crossref Scopus (46) Google Scholar,10Sonke J.E. Koenig A.M. Yakovenko N. Hagelskjær O. Margenat H. Hansson S.V. De Vleeschouwer F. Magand O. Le Roux G. Thomas J.L. A mass budget and box model of global plastics cycling, degradation and dispersal in the land-ocean-atmosphere system.Microplast. nanoplast. 2022; 2: 28https://doi.org/10.1186/s43591-022-00048-wCrossref Google Scholar Atmospheric long-range transport has been suggested as the major pathway of MPs observed in the remote environment such as protected areas of the US,11Brahney J. Hallerud M. Heim E. Hahnenberger M. Sukumaran S. Plastic rain in protected areas of the United States.Science. 2020; 368: 1257-1260https://doi.org/10.1126/science.aaz5819Crossref PubMed Scopus (376) Google Scholar pristine mountain catchments in the French Pyrenees,12Allen S. Allen D. Phoenix V.R. Le Roux G. Jiménez P.D. Simonneau A. Binet S. Galop D. Atmospheric transport and deposition of microplastics in a remote mountain catchment (vol 12, pg 339, 2019).Nat. Geosci. 2019; 12: 679https://doi.org/10.1038/s41561-019-0409-4Crossref Scopus (13) Google Scholar the Tibetan Plateau,13Dong H. Wang L. Wang X. Xu L. Chen M. Gong P. Wang C. Microplastics in a remote lake basin of the Tibetan plateau: impacts of atmospheric transport and glacial melting.Environ. Sci. Technol. 2021; 55: 12951-12960https://doi.org/10.1021/acs.est.1c03227Crossref PubMed Scopus (49) Google Scholar and the polar regions.14Bergmann M. Mützel S. Primpke S. Tekman M.B. Trachsel J. Gerdts G. White and wonderful? Microplastics prevail in snow from the Alps to the Arctic.Sci. Adv. 2019; 5: eaax1157https://doi.org/10.1126/sciadv.aax1157Crossref PubMed Scopus (601) Google Scholar Records of atmospheric MP levels remain sparse, but limited data suggest that atmospheric deposition rates range from 50 to 700 MP m−2 d−1 with atmospheric concentrations of 10−2 to 101 MP m−3.15Allen D. Allen S. Abbasi S. Baker A. Bergmann M. Brahney J. Butler T. Duce R.A. Eckhardt S. Evangeliou N. et al.Microplastics and nanoplastics in the marine-atmosphere environment.Nat. Rev. Earth Environ. 2022; 3: 393-405https://doi.org/10.1038/s43017-022-00292-xCrossref Scopus (34) Google Scholar Two categories of sources are hypothesized as the major contributors to atmospheric MPs: (1) marine sources, such as aerosolization of marine plastics caused by sea spray, and (2) terrestrial sources, such as the resuspension of road-related plastic particles from tires, brakes, and road surfaces; movement of previously deposited or discharged plastics in soils and agriculture lands; and direct atmospheric emissions from human activities occurring around population centers.8Brahney J. Mahowald N. Prank M. Cornwell G. Klimont Z. Matsui H. Prather K.A. Constraining the atmospheric limb of the plastic cycle.Proc. Natl. Acad. Sci. USA. 2021; 118e2020719118https://doi.org/10.1073/pnas.2020719118Crossref Scopus (140) Google Scholar While most bottom-up inventories focus on plastic emissions to the total environment, the inverse modeling method is often applied to constrain the atmospheric emissions with observed data. For example, Brahney et al.8Brahney J. Mahowald N. Prank M. Cornwell G. Klimont Z. Matsui H. Prather K.A. Constraining the atmospheric limb of the plastic cycle.Proc. Natl. Acad. Sci. USA. 2021; 118e2020719118https://doi.org/10.1073/pnas.2020719118Crossref Scopus (140) Google Scholar suggest the current global atmospheric emissions of MPs as 8,600 (0–22,000) Gg year−1 mainly from the ocean source, and Long et al.16Long X. Fu T.M. Yang X. Tang Y. Zheng Y. Zhu L. Shen H. Ye J. Wang C. Wang T. Li B. Efficient atmospheric transport of microplastics over Asia and adjacent oceans.Environ. Sci. Technol. 2022; 56: 6243-6252https://doi.org/10.1021/acs.est.1c07825Crossref Scopus (11) Google Scholar found a large tire dust source (280 Gg yea−1) over the continent of Asia. Residence time is the average time for MP particles to stay in the atmosphere before their removal by dry and wet deposition, which is calculated as the ratio of its atmospheric abundance to total emission flux. The wet deposition of an MP is similar to other aerosol components such as dust and sea salt,11Brahney J. Hallerud M. Heim E. Hahnenberger M. Sukumaran S. Plastic rain in protected areas of the United States.Science. 2020; 368: 1257-1260https://doi.org/10.1126/science.aaz5819Crossref PubMed Scopus (376) Google Scholar while the size, shape, and chemical compositions are important factors influencing the dry deposition velocity of MP particles. The shapes can be pellets, fibers, and sheets with sizes ranging from 10−6 to 10−3 m.15Allen D. Allen S. Abbasi S. Baker A. Bergmann M. Brahney J. Butler T. Duce R.A. Eckhardt S. Evangeliou N. et al.Microplastics and nanoplastics in the marine-atmosphere environment.Nat. Rev. Earth Environ. 2022; 3: 393-405https://doi.org/10.1038/s43017-022-00292-xCrossref Scopus (34) Google Scholar The density ranges, approximately from 0.9 to 1.4 g cm−3, may have an overall smaller influence on deposition velocity than the other factors. Therefore, aerodynamic diameters are often used, as they are directly associated with the residence time of MPs in the atmosphere. In modeling studies, aerodynamic sizes of different MP particles are considered with a residence time from hours to days.8Brahney J. Mahowald N. Prank M. Cornwell G. Klimont Z. Matsui H. Prather K.A. Constraining the atmospheric limb of the plastic cycle.Proc. Natl. Acad. Sci. USA. 2021; 118e2020719118https://doi.org/10.1073/pnas.2020719118Crossref Scopus (140) Google Scholar,16Long X. Fu T.M. Yang X. Tang Y. Zheng Y. Zhu L. Shen H. Ye J. Wang C. Wang T. Li B. Efficient atmospheric transport of microplastics over Asia and adjacent oceans.Environ. Sci. Technol. 2022; 56: 6243-6252https://doi.org/10.1021/acs.est.1c07825Crossref Scopus (11) Google Scholar Given their importance, the sources of atmospheric MPs and their residence times remain unclear. This study develops an atmospheric transport model for MPs based on the Goddard Earth Observing System (GEOS)-Chem model to determine the most likely sources and residence time of atmospheric MPs. The global atmospheric MP emissions are determined to be 324 (73–1,450) Gg per year. Marine plastic aerosolization is still considered a significant source, but with considerably lower global emissions (171 [38–764] Gg per year), followed by road-related sources (115 [26–513] Gg per year), which include tire and brake wear suspension and improperly managed plastic waste. The findings reveal a predominantly coarse size distribution for these sources, with almost all (99%) of the MP particles classified as very coarse aerosols (70 μm). We calculate a net land-to-ocean transport of atmospheric MP at 25 Gg per year, a considerably smaller amount than riverine discharge and erosion of coastal waste to the ocean (approximately 1,000 Gg per year). Our findings underscore the necessity of regulating litter and dumping of plastic waste near roads. To enhance our comprehension of the atmospheric MP cycle, more data on atmospheric abundance in unexamined regions and direct measurements of emissions from various source categories are required. The GEOS-Chem model considers the emissions, transport, and deposition of MPs in the atmosphere and includes six aerosol-like MP tracers with aerodynamic sizes ranging from 0.3 to 70 μm and residence times ranging from 0.04 to 6.5 days following Brahney et al.8Brahney J. Mahowald N. Prank M. Cornwell G. Klimont Z. Matsui H. Prather K.A. Constraining the atmospheric limb of the plastic cycle.Proc. Natl. Acad. Sci. USA. 2021; 118e2020719118https://doi.org/10.1073/pnas.2020719118Crossref Scopus (140) Google Scholar This size range mainly belongs to MPs with the smallest one to nanoplastics, but we call both MPs for simplicity. We lump all shapes of MP particles into these six bins, as the aerodynamic size is the most important influencing factor for their residence time, and pellets also dominate the air samples among other shapes such as fragments and fibers.16Long X. Fu T.M. Yang X. Tang Y. Zheng Y. Zhu L. Shen H. Ye J. Wang C. Wang T. Li B. Efficient atmospheric transport of microplastics over Asia and adjacent oceans.Environ. Sci. Technol. 2022; 56: 6243-6252https://doi.org/10.1021/acs.est.1c07825Crossref Scopus (11) Google Scholar Possible sources include aerosolized marine plastic, traffic-related sources, resuspension of MMPW and agricultural plastic waste, and generic sources associated with residential activities. The model results are compared with available observed data for atmospheric MPs (both atmospheric concentrations and deposition fluxes). We adopt an optimal estimation approach following Brahney et al.8Brahney J. Mahowald N. Prank M. Cornwell G. Klimont Z. Matsui H. Prather K.A. Constraining the atmospheric limb of the plastic cycle.Proc. Natl. Acad. Sci. USA. 2021; 118e2020719118https://doi.org/10.1073/pnas.2020719118Crossref Scopus (140) Google Scholar and Long et al.16Long X. Fu T.M. Yang X. Tang Y. Zheng Y. Zhu L. Shen H. Ye J. Wang C. Wang T. Li B. Efficient atmospheric transport of microplastics over Asia and adjacent oceans.Environ. Sci. Technol. 2022; 56: 6243-6252https://doi.org/10.1021/acs.est.1c07825Crossref Scopus (11) Google Scholar and incorporate all available observational data over the global land and the ocean environment.15Allen D. Allen S. Abbasi S. Baker A. Bergmann M. Brahney J. Butler T. Duce R.A. Eckhardt S. Evangeliou N. et al.Microplastics and nanoplastics in the marine-atmosphere environment.Nat. Rev. Earth Environ. 2022; 3: 393-405https://doi.org/10.1038/s43017-022-00292-xCrossref Scopus (34) Google Scholar Due to the close coupling between emission strength and size distribution (i.e., residence time), they are optimized simultaneously in the optimization procedure (more details in the experimental procedures). We find a global total atmospheric MP emission of 324 Gg year−1, with terrestrial and marine sources contributing 154 and 171 Gg year−1, respectively (Figure 1). Among the terrestrial sources, the road-related source is dominant (115 Gg year−1), followed by agriculture dust (38 Gg year−1), residential sources (0.80 Gg year−1), and MMPW resuspension (0.11 Gg year−1). The spatial pattern of ocean emissions follows that of the modeled surface ocean plastic mass, scaled by the wind speed and sea surface temperature (see experimental procedures).17Jaeglé L. Quinn P.K. Bates T.S. Alexander B. Lin J.T. Global distribution of sea salt aerosols: new constraints from in situ and remote sensing observations.Atmos. Chem. Phys. 2011; 11: 3137-3157https://doi.org/10.5194/acp-11-3137-2011Crossref Scopus (384) Google Scholar The highest emissions are in the subtropical centers of the five gyres (i.e., North Pacific, South Pacific, North Atlantic, South Atlantic, and the Indian Ocean), where convergent ocean circulation causes an accumulation of floating plastic particles (Figure 1B).18Peng Y. Wu P. Schartup A.T. Zhang Y. Plastic waste release caused by COVID-19 and its fate in the global ocean.Proc. Natl. Acad. Sci. USA. 2021; 118e2111530118https://doi.org/10.1073/pnas.2111530118Crossref Scopus (156) Google Scholar Road-related emissions follow that of traffic activities with higher emissions in North America, West Europe, East Asia, and South Asia (Figure 1A). The other three categories have much smaller emissions compared with these two. The optimizer can also infer the size distribution of emissions. The GEOS-Chem model simulates the mass concentrations of six MP bins that have distinct atmospheric residence times. A unit mass of emissions with different sizes thus has different travel distances and a varied impact on the MP concentrations at receptors. By searching within the possible combinations for the size distributions of emissions to minimize the model-observation difference, the optimizer suggests that almost all (99% in mass) MP particles are emitted as very coarse aerosols (70 μm), reflecting the fact that most of them are from the ocean and road-related sources. The fractions from smaller sizes (0.3–35 μm) are much smaller, with negligible contributions. Indeed, these sources are related to mechanical processes forced by strong wind events or wind/wave breaking of sea surface spray. The size distribution of emissions reflects the influence of inertial or cohesive forces on roadside and ocean surfaces. The results are also consistent with observed size distributions of MPs in the marine atmospheric boundary layer (MABL or MBL), with mass concentrations often dominated by sizes >50 μm.19Wang X. Li C. Liu K. Zhu L. Song Z. Li D. Atmospheric microplastic over the South China sea and East Indian ocean: abundance, distribution and source.J. Hazard Mater. 2020; 389121846https://doi.org/10.1016/j.jhazmat.2019.121846Crossref Scopus (118) Google Scholar,20Ding Y. Zou X. Wang C. Feng Z. Wang Y. Fan Q. Chen H. The abundance and characteristics of atmospheric microplastic deposition in the northwestern South China Sea in the fall.Atmos. Environ. 2021; 253118389https://doi.org/10.1016/j.atmosenv.2021.118389Crossref Scopus (48) Google Scholar This size distribution pattern contrasts with combustion processes, which are typically associated with smaller or finer particles like sulfate and black carbon. The residential sources indeed may have a smaller particle size, but our study could not sufficiently constrain it due to the very low overall global emissions. This is generally consistent with previous studies.8Brahney J. Mahowald N. Prank M. Cornwell G. Klimont Z. Matsui H. Prather K.A. Constraining the atmospheric limb of the plastic cycle.Proc. Natl. Acad. Sci. USA. 2021; 118e2020719118https://doi.org/10.1073/pnas.2020719118Crossref Scopus (140) Google Scholar,16Long X. Fu T.M. Yang X. Tang Y. Zheng Y. Zhu L. Shen H. Ye J. Wang C. Wang T. Li B. Efficient atmospheric transport of microplastics over Asia and adjacent oceans.Environ. Sci. Technol. 2022; 56: 6243-6252https://doi.org/10.1021/acs.est.1c07825Crossref Scopus (11) Google Scholar For example, Brahney et al.8Brahney J. Mahowald N. Prank M. Cornwell G. Klimont Z. Matsui H. Prather K.A. Constraining the atmospheric limb of the plastic cycle.Proc. Natl. Acad. Sci. USA. 2021; 118e2020719118https://doi.org/10.1073/pnas.2020719118Crossref Scopus (140) Google Scholar considered three cases with different prescribed size distributions and found that the best case has 85% global ocean emissions in the 70 μm bin. Long et al.16Long X. Fu T.M. Yang X. Tang Y. Zheng Y. Zhu L. Shen H. Ye J. Wang C. Wang T. Li B. Efficient atmospheric transport of microplastics over Asia and adjacent oceans.Environ. Sci. Technol. 2022; 56: 6243-6252https://doi.org/10.1021/acs.est.1c07825Crossref Scopus (11) Google Scholar also found that 87% terrestrial emissions and 90% ocean emissions are larger than 50 μm. One drawback of this study is that we use only bulk mass concentrations of observed MPs in the cost function of our optimizer (see experimental procedures), as many studies did not report size distributions or used inconsistent definition of size categories.9Allen S. Allen D. Baladima F. Phoenix V.R. Thomas J.L. Le Roux G. Sonke J.E. Evidence of free tropospheric and long-range transport of microplastic at Pic du Midi Observatory.Nat. Commun. 2021; 12: 7242https://doi.org/10.1038/s41467-021-27454-7Crossref Scopus (46) Google Scholar In addition, many of these studies adopted Fourier transform infrared spectroscopy to identify polymers. This technique is typically limited to size ranges greater than ∼30 μm, which thus causes an underestimation of MP particles smaller than this size.21Allen S. Allen D. Moss K. Le Roux G. Phoenix V.R. Sonke J.E. Examination of the ocean as a source for atmospheric microplastics.PLoS One. 2020; 15e0232746https://doi.org/10.1371/journal.pone.0232746Crossref Scopus (125) Google Scholar Nevertheless, our optimizer can greatly benefit from more consistent observations with detailed size distributions by using a cost function considering such information. We find a global terrestrial emission source of 154 Gg year−1. Figure 2 compares the model results driven by the optimized emission inventory against these observations. These data include atmospheric deposition fluxes for US national parks and European and Asian coastal cities. Overall, the modeled average deposition flux over all the sites is 3.6 ± 3.8 kg km−2 year−1, which is not significantly different from the observations (3.5 ± 3.2 kg km−2 year−1) (Figures 2C and 2D). The measured atmospheric deposition flux over US national parks (2.8 ± 2.3 kg km−2 year−1), which are remote and mountainous regions, is slightly lower than the data over urban areas in Asia (3.5 ± 2.8 kg km−2 year−1). Our model simulates a lower deposition flux of 1.5 ± 1.2 kg km−2 year−1 in US national parks but at a comparable level (3.9 ± 2.3 kg km−2 year−1) over Asia. The model (14 ± 15 ng m−3) also reproduces the observed (18 ± 24 ng m−3) atmospheric MP concentrations over land-based sites. Indeed, our model cannot fully simulate the variability in the observations (e.g., Figure 2D). The optimized terrestrial emission and model performance are similar to those of Brahney et al.8Brahney J. Mahowald N. Prank M. Cornwell G. Klimont Z. Matsui H. Prather K.A. Constraining the atmospheric limb of the plastic cycle.Proc. Natl. Acad. Sci. USA. 2021; 118e2020719118https://doi.org/10.1073/pnas.2020719118Crossref Scopus (140) Google Scholar (183 Gg year−1), even though we include observations not only over US national parks, which were used by that study, but also data from other continents including atmospheric concentrations and deposition fluxes. This, on one hand, suggests that our estimate for terrestrial sources is quite robust, i.e., not sensitive to a specific subset of data. On the other hand, this may mean that both our optimized results are local optimums due to the lack of size distribution information, which is the key to further improve the model performance. The optimized results suggest that road-related (115 Gg year−1) and agricultural (38 Gg year−1) sources contribute the most to total terrestrial emissions, while the contributions from MMPW and residential sectors are much smaller. This is generally consistent with Brahney et al.8Brahney J. Mahowald N. Prank M. Cornwell G. Klimont Z. Matsui H. Prather K.A. Constraining the atmospheric limb of the plastic cycle.Proc. Natl. Acad. Sci. USA. 2021; 118e2020719118https://doi.org/10.1073/pnas.2020719118Crossref Scopus (140) Google Scholar and Long et al.,16Long X. Fu T.M. Yang X. Tang Y. Zheng Y. Zhu L. Shen H. Ye J. Wang C. Wang T. Li B. Efficient atmospheric transport of microplastics over Asia and adjacent oceans.Environ. Sci. Technol. 2022; 56: 6243-6252https://doi.org/10.1021/acs.est.1c07825Crossref Scopus (11) Google Scholar but we estimate a lower emission from agricultural sources than Brahney et al.8Brahney J. Mahowald N. Prank M. Cornwell G. Klimont Z. Matsui H. Prather K.A. Constraining the atmospheric limb of the plastic cycle.Proc. Natl. Acad. Sci. USA. 2021; 118e2020719118https://doi.org/10.1073/pnas.2020719118Crossref Scopus (140) Google Scholar (69 Gg year−1). We scale the MP emissions of this source by the fraction of agricultural land use and dust emission fluxes, which are quite sporadic but had no observations available nearby (see experimental procedures; Figure 1C). We thus call for more observations near these source regions to better constrain this source. The residential sector contains sources that include synthetic fibers from clothing, artificial turf, personal care, and cosmetic products.22Wang T. Li B. Zou X. Wang Y. Li Y. Xu Y. Mao L. Zhang C. Yu W. Emission of primary microplastics in mainland China: invisible but not negligible.Water Res. 2019; 162: 214-224https://doi.org/10.1016/j.watres.2019.06.042Crossref Scopus (102) Google Scholar We scale the spatial pattern of this source by population density following Brahney et al.8Brahney J. Mahowald N. Prank M. Cornwell G. Klimont Z. Matsui H. Prather K.A. Constraining the atmospheric limb of the plastic cycle.Proc. Natl. Acad. Sci. USA. 2021; 118e2020719118https://doi.org/10.1073/pnas.2020719118Crossref Scopus (140) Google Scholar We find that the relatively high MP deposition flux over US national parks with low population density drives down the optimized emission for this category. It is surprising to find that the aerosolization of MMPW is a negligible source. MMPW emissions are based on MMPW generation (see experimental procedures), which has a larger emission from populous developing countries such as India and China (Figure 1E).1Borrelle S.B. Ringma J. Law K.L. Monnahan C.C. Lebreton L. McGivern A. Murphy E. Jambeck J. Leonard G.H. Hilleary M.A. et al.Predicted growth in plastic waste exceeds efforts to mitigate plastic pollution.Science. 2020; 369: 1515-1518https://doi.org/10.1126/science.aba3656Crossref PubMed Google Scholar Wind speed and soil moisture also influence the emissions analogous to dust emissions, which makes higher emissions over dry and windy climate regions such as the Middle East (Figure 1D). Developed countries in Western Europe and North America have generally lower MMPW emissions due to better waste management practices.1Borrelle S.B. Ringma J. Law K.L. Monnahan C.C. Lebreton L. McGivern A. Murphy E. Jambeck J. Leonard G.H. Hilleary M.A. et al.Predicted growth in plastic waste exceeds efforts to mitigate plastic pollution.Science. 2020; 369: 1515-1518https://doi.org/10.1126/science.aba3656Crossref PubMed Google Scholar,23Jambeck J.R. Geyer R. Wilcox C. Siegler T.R. Perryman M. Andrady A. Narayan R. Law K.L. Plastic waste inputs from land into the ocean.Science. 2015; 347: 768-771https://doi.org/10.1126/science.1260352Crossref PubMed Scopus (6050) Google Scholar However, such a pattern is not well supported by the available observation data, and the optimizer generates a low weighting factor for this source category. An alternative explanation is that the optimized road-related source does not only include plastic emissions from tire and brake wears but also the aerosolization of MMPW dumped and/or littered roadside, which consists of two-thirds of all MMPW.24Kawecki D. Nowack B. Polymer-specific modeling of the environmental emissions of seven commodity plastics as macro- and microplastics.Environ. Sci. Technol. 2019; 53: 9664-9676https://doi.org/10.1021/acs.est.9b02900Crossref PubMed Scopus (116) Google Scholar Exposure to the environment including ultra-violet radiation, high temperatures, and being smashed by vehicles further break it down into microscopic fragments before being brought into the atmosphere by the energetic flows and turbulence caused by traffic.8Brahney J. Mahowald N. Prank M. Cornwell G. Klimont Z. Matsui H. Prather K.A. Constraining the atmospheric limb of the plastic cycle.Proc. Natl. Acad. Sci. USA. 2021; 118e2020719118https://doi.org/10.1073/pnas.2020719118Crossref Scopus (140) Google Scholar This is also supported by the fact that road dust consists of multiple chemical compositions rather than just tire materials (natural or synthetic rubber).25Dehghani S. Moore F. Akhbarizadeh R. Microplastic pollution in deposited urban dust, Tehran metropolis, Iran.Environ. Sci. Pollut. Res. 2017; 24: 20360-20371https://doi.org/10.1007/s11356-017-9674-1Crossref PubMed Scop