The research, published in Environmental Advances by Katsushige Uranishi, Shoei Yoshioka and Hikari Shimadera, analyzed PM2.5 chemical composition data collected between 2016 and 2023 at eight sites spread across Kyushu, a region of roughly 36,000 square kilometers surrounded by the Seto Inland Sea, the Sea of Japan, the Yellow Sea and the East China Sea. Six of the sites sit on the coast, including one near Hakata Port, one of the largest international ports in western Japan, and another in a coastal industrial area facing the Seto Inland Sea, one of the busiest maritime corridors in the country. Two sites, Yufu and Miyakonojo, lie about 14 and 32 kilometers from the nearest coastline, farther from major ports and coastal industry, and were included specifically to test how far the atmospheric fingerprint of shipping regulation extends.

The timing of the study matters. The sulfur cap came into force just as the COVID-19 pandemic disrupted economic activity worldwide, so any before-and-after comparison risks conflating a fuel regulation with a collapse in shipping demand. The authors addressed this by extending the observation window to eight years, covering 2016 to 2019 before the rule and 2020 to 2023 after it, which dampens the influence of year-to-year meteorological variability, and by pulling annual port statistics from Japan’s Ministry of Land, Infrastructure, Transport and Tourism to track whether vessel activity itself collapsed in ways that could mimic a regulatory effect.

Because vanadium and nickel, the classic tracers of residual fuel oil combustion, are also emitted by thermal power plants and large industrial facilities, the team did not rely on elemental concentrations alone. Instead they applied positive matrix factorization, a receptor model, to a combined dataset of 3,605 samples and 21 chemical variables, including ions, metallic elements measured by ICP-MS, organic and elemental carbon, and total PM2.5 mass. The model resolved seven source factors, and the solution passed stability checks: no factor swaps appeared in displacement diagnostics, and bootstrap mapping percentages exceeded 92 percent for every factor.

One factor stood out. It was dominated by vanadium, which accounted for 82 percent of the factor’s composition, and nickel at 49 percent, and it showed a vanadium-to-nickel ratio of 3.5, closely matching the value of 3.9 reported for marine vessel emissions in earlier Japanese work. Its contribution was greatest at the coastal sites nearest ports and shipping routes, and directional analysis with bivariate polar plots showed enhancements under winds blowing from nearby ports and maritime source regions. The authors interpret this factor as predominantly, though not exclusively, ship-related heavy oil combustion; stationary heavy oil users in coastal industrial areas cannot be fully excluded, and the paper is explicit that the factor’s entire contribution should not be read as shipping alone.

The temporal signal was unambiguous. Before 2020, the heavy oil combustion factor contributed an average of 0.755 micrograms per cubic meter of PM2.5 across the network; after 2020, that fell to 0.119 micrograms per cubic meter, a drop of 0.636 micrograms per cubic meter. The largest pre-regulation contribution occurred at the Oita coastal site, which faces the heavily trafficked Seto Inland Sea, where the annual mean contribution ranged from 1.35 to 1.82 micrograms per cubic meter before the cap and fell from 1.55 to 0.30 micrograms per cubic meter between the two periods, a reduction of 80.7 percent. The largest overall reduction, 90.9 percent, appeared at the Kagoshima coastal site, while the smallest, at the Saga site, still reached 64.8 percent when measured over 2021 to 2023.

Crucially, the inland sites told the same story. At Yufu and Miyakonojo, contributions fell by 87.0 and 86.9 percent respectively between the pre-regulation and post-regulation periods, and remained 86.0 and 83.8 percent below pre-regulation levels even when 2020, the first pandemic year, was excluded from the comparison. Before the cap, the inland sites received roughly one-third of the heavy oil combustion contribution measured at the neighboring coastal site in Oita, evidence that ship-related particles were being transported regionally rather than staying at the shoreline. Seasonal analysis reinforced this picture: before 2020, contributions at both coastal and inland sites peaked in spring and summer, and after 2020 the enhancement largely vanished in every season.

The port statistics help rule out a pandemic-driven explanation. At Oita Port, vessel calls and total gross tonnage in 2020 stood at 94.4 and 93.0 percent of their 2019 levels, yet the heavy oil combustion contribution fell by more than 80 percent. At Miyazaki Port, calls dipped only to 95.8 percent of the 2019 level while gross tonnage actually rose to 105.5 percent, even as the factor fell by roughly 86 percent at the nearby sites. By 2022 and 2023, tonnage at Oita had recovered to 100.5 and 102.4 percent of 2019 levels, Miyazaki had climbed to 117.6 and 122.3 percent, and vessel calls at Beppu Port, near the inland Yufu site, exceeded 2019 levels, while the atmospheric signal stayed low throughout. The authors caution that these statistics capture vessel activity rather than fuel consumption, engine load or actual emissions, and that ships transiting corridors such as the Seto Inland Sea without calling at the selected ports are not counted, so pandemic-related changes in overall maritime activity cannot be completely excluded.

Supporting evidence came from ambient trace metals. Vanadium concentrations fell by approximately 72 to 83 percent across all eight sites between 2019 and 2020, with the coastal sites near ports showing reductions of about 80 and 77 percent. Nickel behaved less consistently, even rising at one coastal site in 2020 while vanadium dropped sharply, suggesting additional industrial influences and confirming vanadium as the more robust indicator of the fuel switch. The study also situates its results in a wider East Asian context: earlier work estimated that the sulfur regulation cut ship-related PM2.5 contributions by about 47 percent on average in Busan, Korea, and reduced vanadium concentrations in Shanghai by 74 percent, while a 2025 estimate for Shanghai put reductions in ship-emitted vanadium and nickel at 91.6 and 46.7 percent. The Kyushu reduction rates are consistent with this regional pattern, though the authors note the numbers cannot be compared directly because the studies measured different quantities.

What does the shipping cleanup mean for overall air quality in Kyushu? Total observed PM2.5 averaged 14.97 micrograms per cubic meter in 2016 to 2019 and 11.64 micrograms per cubic meter in 2020 to 2023, a decline of 3.33 micrograms per cubic meter. The heavy oil combustion factor accounts for about 19.1 percent of that improvement, or 18.7 percent using a sensitivity calculation based on PMF-reconstructed mass, a measurable but not dominant share. The largest PM2.5 source throughout the period was secondary sulfate, at 38 percent of the total, whose gradual decline the authors attribute largely to reduced transboundary transport from the Asian continent as Chinese emission controls continued, with possible contributions from volcanic sulfur dioxide at Sakurajima that the model did not separately quantify. The study’s broader lesson is methodological: long-term chemical composition measurements combined with receptor modeling can isolate the fingerprint of a specific emission regulation even when it overlaps with an economic shock, and the authors suggest the same framework could evaluate future shipping rules and other large-scale air quality policies in coastal regions worldwide.

Subject of Research: The impact of the IMO 2020 Sulfur Cap on ship-related PM2.5 across coastal and inland areas of Kyushu, Japan

Article Title: Regional-scale impacts of the IMO 2020 Sulfur Cap on ship-related PM 2.5 across coastal and inland areas of Kyushu, Japan

Article References: Uranishi, K., Yoshioka, S., & Shimadera, H. (2026). Regional-scale impacts of the IMO 2020 Sulfur Cap on ship-related PM2.5 across coastal and inland areas of Kyushu, Japan. Environmental Advances, Article 100773. https://doi.org/10.1016/j.envadv.2026.100773

Image Credits: AI Generated

DOI: 10.1016/j.envadv.2026.100773

Keywords: IMO 2020 Sulfur Cap, PM2.5, ship emissions, positive matrix factorization, Kyushu, vanadium, nickel, heavy oil combustion, air quality, marine fuel sulfur, coastal pollution, long-range transport