Next Article in Journal
Physical Activity and Sociodemographic Profile of Brazilian People during COVID-19 Outbreak: An Online and Cross-Sectional Survey
Next Article in Special Issue
Use of Genomic Information in Health Impact Assessment is Yet to Come: A Systematic Review
Previous Article in Journal
Use of Non-Cancer Medications in New Zealand Women at the Diagnosis of Primary Invasive Breast Cancer: Prevalence, Associated Factors and Effects on Survival
Previous Article in Special Issue
Measuring Burden of Disease Attributable to Air Pollution Due to Preterm Birth Complications and Infant Death in Paris Using Disability-Adjusted Life Years (DALYs)
Open AccessArticle

Health Impact of Air Pollution from Shipping in the Baltic Sea: Effects of Different Spatial Resolutions in Sweden

1
School of Health Systems and Public Health, Faculty of Health Sciences, University of Pretoria, Pretoria 0002, South Africa
2
Occupational and Environmental Medicine, School of Public Health and Community Medicine, Institute of Medicine, University of Gothenburg, Box 100, 405 30 Gothenburg, Sweden
3
Division for Climate Modelling and Air Pollution, Norwegian Meteorological Institute, NO 0313 Oslo, Norway
4
The Environmental Department, City of Gothenburg, PO Box 7012, 402 31 Gothenburg, Sweden
5
Air Quality Research, Finnish Meteorological Institute, PL 503, FI-00101 Helsinki, Finland
6
Department of Infection Diseases, Sahlgrenska Academy, University of Gothenburg, SE-405 30 Goteborg, Sweden
*
Author to whom correspondence should be addressed.
Int. J. Environ. Res. Public Health 2020, 17(21), 7963; https://doi.org/10.3390/ijerph17217963
Received: 7 September 2020 / Revised: 23 October 2020 / Accepted: 27 October 2020 / Published: 29 October 2020
(This article belongs to the Special Issue Health Impact Assessment)

Abstract

In 2015, stricter regulations to reduce sulfur dioxide emissions and particulate air pollution from shipping were implemented in the Baltic Sea. We investigated the effects on population exposure to particles <2.5 µm (PM2.5) from shipping and estimated related morbidity and mortality in Sweden’s 21 counties at different spatial resolutions. We used a regional model to estimate exposure in Sweden and a city-scale model for Gothenburg. Effects of PM2.5 exposure on total mortality, ischemic heart disease, and stroke were estimated using exposure–response functions from the literature and combining them into disability-adjusted life years (DALYS). PM2.5 exposure from shipping in Gothenburg decreased by 7% (1.6 to 1.5 µg/m3) using the city-scale model, and 35% (0.5 to 0.3 µg/m3) using the regional model. Different population resolutions had no effects on population exposures. In the city-scale model, annual premature deaths due to shipping PM2.5 dropped from 97 with the high-sulfur scenario to 90 in the low-sulfur scenario, and in the regional model from 32 to 21. In Sweden, DALYs lost due to PM2.5 from Baltic Sea shipping decreased from approximately 5700 to 4200. In conclusion, sulfur emission restrictions for shipping had positive effects on health, but the model resolution affects estimations.
Keywords: PM2.5; SECA; EMEP model; population exposure; health effects; myocardial infarction; heart attack; stroke; air pollutants PM2.5; SECA; EMEP model; population exposure; health effects; myocardial infarction; heart attack; stroke; air pollutants

1. Introduction

Air pollution is one of the world’s most detrimental environmental health risks, with outdoor and indoor air pollution together annually accounting for 4.1 million premature deaths worldwide [1]. Many studies have shown that sulfur dioxide (SO2), nitrogen oxides (NO, NO2), and particulate matter (particles <2.5 µm (PM2.5) and particles <10 µm (PM10)) have strong associations with adverse cardiopulmonary health effects [2,3,4]. Long-term exposure to PM2.5 is highly associated with adverse health effects such as ischemic heart disease, myocardial infarction (MI), and stroke, and mortality [3,5,6,7,8,9,10,11,12]. The multiple sources of air pollution include anthropogenic and natural emissions [13]. One anthropogenic source is shipping emissions, contributing an estimated 13% of global sulfur emissions [14]. PM2.5 from shipping consists of both primary emissions and secondary particles from SO2 [15]. Sulfur in the form of SO2 is one of the key precursors for particulate matter (PM) [16,17], and the reduction in sulfur content in shipping fuel impacts the presence of both PM2.5 and PM10.
Countries across the world have designated sulfur emission control areas (SECAs) with governance from the International Maritime Organization (IMO) [18,19,20]. Studies have shown that shipping emissions increase the presence of SO2 and PM in harbor cities [21,22]. The SECAs stipulated the areas subject to guidelines to reduce the sulfur content in shipping fuel. The gradually tightening marine fuel sulfur (S) regulations (maximum allowed fuel sulfur content decreased from 3.5% to 1.5% in 2006 and to 1.0% in 2010) have significantly decreased the SOx and PM emissions from Baltic Sea shipping and from 2015 onwards, only 0.1%S fuel is allowed in the area. The 2015 change only concerns regional sulfur rules though since low-sulfur fuel has been mandatory in port areas in the EU since 2010. In addition to regional sulfur limits, low-sulfur fuel must be used in EU port areas from 2010 onwards as required by the EU Sulfur Directive, [23,24]. Globally, outside the SECAs (Figure 1), the maximum allowed fuel sulfur content was decreased from 3.5% to 0.5% on 1 January 2020 [24].
In recent decades, the concentrations of air pollutants have decreased in Northern Europe and Sweden [16,21,25,26]. The decrease in particle emissions from shipping due to stricter SECA regulations in the Baltic Sea and the North Sea has been estimated to have led to substantial health improvements, and further restriction will lead to additional benefits [23,27].
Recent studies investigating PM2.5 exposure levels in Sweden and other European countries [28,29] have used only coarse resolutions to assess population exposures. Few studies have used finer population grids or assessed how such refinement may affect the assumptions made about different health outcomes [25]. Therefore, the aim of this study was to investigate how mortality and morbidity rates associated with PM2.5 exposure differed depending on spatial resolutions in Gothenburg, Sweden, before and after stricter sulfur regulations were implemented in 2015. We also examined the health impact in Sweden and its 21 counties before and after sulfur regulations.

2. Materials and Methods

2.1. Modeling and Geocoding

In this paper, three exposure models were used. First, the Ship Traffic Emission Assessment Model (STEAM; [30,31,32]) was used to generate ship emission inventories for consecutive chemical transport modeling steps. STEAM takes as input ship position data from the Automatic Identification System (AIS), which enables tracking of individual vessels with very high spatio-temporal resolution (meters, seconds). These activity data were provided by the Baltic Sea (Helsinki Commission) countries. STEAM models vessel water resistance, power use, and emissions based on technical description of ships [33]. For this work, both regional and local (city-scale, near harbors) emission inventories for ship emissions were generated for 2014, 2015, and 2016 with STEAM. Table S1 shows the summary of ship emissions for both the Baltic Sea and the smaller domain around the Gothenburg area. Figure S1 shows the SOx emissions (kg/cell) from the Baltic Sea in 2016 and Figure S2 shows the difference plot in 2014–2016, around Gothenburg, and illustrates the major shipping lanes.
Further, we have used two chemical transport models (CTM) to assess both total PM2.5 air concentration and the contribution from shipping to PM2.5 in the Baltic Sea. The first CTM, which we will refer to as the regional model, was that of the European Monitoring Evaluation Programme (EMEP), which modeled Europe, including Sweden, in 0.1 × 0.1 (approximately 10 × 10 km) grid squares [34]. The EMEP model has been described in detail previously [35]. The STEAM data were used as input to the EMEP model. The EMEP model was run including all emissions, with high sulfur ship emissions, low sulfur ship emission, and with no ship emissions for the years 2014, 2015, and 2016. The differences “high sulfur - no shipping” and “low sulfur – no shipping” for each year give the shipping contribution with high- and low-sulfur fuel. By calculating the average of the three years for high- and low-sulfur fuel, respectively, the influence of different meteorology between the years was evened out. These models and scenarios have been described in more detail previously [34]. The resulting three-year average model results are named high-sulfur and low-sulfur scenarios throughout the paper. Additionally, we used the population in Sweden for 2015 for all exposure assignments to avoid an effect of possible population fluctuations.
The third model, which we will refer to as the city-scale model is run within the Airviro system [36], an air pollution model system widely used [37,38,39]. Airviro uses the Danard calculations within the Airviro Gauss module (“meteorological pre-processor”). We feed Airviro Gauss with topography, physiography, and weather station data and the module calculates an hour-by-hour wind field over the chosen area with a grid size of 100 × 100 m for the city of Gothenburg, Sweden. The local environmental office in Gothenburg calculated PM2.5 concentrations in this city-scale model with meteorology from 2016 using two scenarios, one with and one without the 2015 SECA enforcement, also in this paper referred to as high-sulfur and low-sulfur scenarios. All land-based emissions and in-harbor area ship emissions in both scenarios were from the emission inventory for the year 2016. The regional EMEP model was used as an input in the city-scale model’s boundary as a transport source, but within the city-scale’s model domain, the regional model grid sources were not used. STEAM emissions were used in both CTMs, but with a different higher-resolution input for the Airviro model. Identical methods for ship emission inventories were used, but the air quality modeling uses EMEP for regional and Airviro for local study. The merging of the concentration fields from the different models was done by projecting the regional model results on to the 100 × 100 m grid of the local model using linear interpolation. The concentration field from the local modeling was then added. The results from this procedure corresponded well with PM2.5 measurements made by the local environmental office in Gothenburg (84–114% of mean annual measurements from permanent stations) [40], without any empirical adjustment of the model results being made.
Data from Eurostat [41] were used to geocode the population of Sweden into 1 × 1 km squares. We also investigated two other resolutions of the spatial distribution of the population in the local Gothenburg analysis. These population datasets included population centroids [42] and the Swedish national population grid (100 m grid size) from Statistics Sweden [43].
A geographical information system in the QGIS program, version 2.18.13, was used in the geocoding and to link air pollution data to inhabitant data. Fixed population data for the year 2015 were used in all calculations to avoid fluctuations due to population changes.

2.2. Baseline Population

In 2015, Sweden had a population of approximately 9.5 million registered residents in 21 counties. Stockholm was the largest county, with a population over 2.1 million inhabitants, whereas Gotland County had the fewest inhabitants (<60,000; Table 1). The population of Gothenburg, the largest city in Västra Götaland County, was approximately 650,000 in 2015.

2.3. Exposure Assessment

We calculated the population-weighted exposure to PM2.5 by combining the regional model EMEP and the Eurostat data. All inhabitants were assigned exposure to PM2.5 in each 1-km square in both high- and low-sulfur scenarios for each of the 21 counties (Table 1), and the population-weighted means were calculated. For the Gothenburg analysis, we compared the EMEP model with the city-scale model using both high- and low-sulfur scenarios for the three population grids.

2.4. Exposure–Response Function and Mortality Calculation

Two linear exposure–response (ER) functions were used for all-cause mortality from long-term exposure to PM2.5. The first is from the European Study of Cohorts for Air Pollution Effects (ESCAPE) project with a relative risk of 1.07 (95% CI 1.02–1.13) per 5 µg/m3 increase in PM2.5 for all-cause mortality [12]. The second ER function was derived from the Health Risks of Air Pollution in Europe (HRAPIE), with a relative risk of 1.062 (95% CI 1.04–1.08) per 10 µg/m3 increase in PM2.5 for all-cause mortality [44]. The two ER functions provide a comparative assessment of the all-cause mortality from shipping emissions. The ESCAPE ER function is more recent and estimates a 14% increased risk of all-cause mortality per 10 µg/m3 of PM2.5 exposure versus the older HRAPIE ER function of 6.2% per 10 µg/m3 of PM2.5 that has been used in most previous health impact assessments [45]. The premature deaths attributable to PM2.5 can be seen as those deaths that could have been avoided if PM2.5 concentrations from shipping had been reduced to 0 µg/m3 everywhere in Sweden (calculation found in the Supplementary Information).
There are 21 counties in Sweden and the majority of the population is located in the southern part of the country. We used the 2015 natural (non-accidental) mortality for those aged ≥30 years for each county [43].

2.5. Years of Life Lost (YLL) Calculations

We used data life tables from Statistics Sweden (SCB) and Eurostat [41] to calculate the YLL for all counties in Sweden and age-specific death and survival rates from the life tables to estimate the corresponding reduction in life expectancy from PM2.5 exposure from shipping for the average person in the Swedish population aged ≥30 years, per year.
The YLL calculation was used to determine disability-adjusted life years (DALYs) as a measure of the estimated burden of disease in Sweden, its 21 counties, and the city of Gothenburg.

2.6. Morbidity Calculations

MI and stroke were the health outcomes considered for the morbidity estimation, because heart disease and stroke are important health outcomes and the most common reasons for premature deaths attributed to exposure to air pollution [7,46]. The relative risk per 5 µg/m3 increase in PM2.5 used for MI was 1.13 (95% CI 0.98–1.30) [47], and for stroke it was 1.19 (95% CI 0.88–1.62) [48]. The Swedish National Board of Health and Welfare’s statistical database provided all the data regarding MI and stroke events in Sweden, its counties, and the city of Gothenburg [43]. The relative risk was calculated from the average population exposure and the number of first-time incidents for the year in question. We avoided double counting by subtracting the number of deaths caused by MI and stroke that year from the number of first-time incidents.

3. Results

3.1. Exposures in Gothenburg

The city-scale model estimated a total mean PM2.5 concentration of 6.92 µg/m3 in the high-sulfur scenario and 6.82 µg/m3 in the low-sulfur scenario. In comparison, the regional model recorded total mean PM2.5 concentration levels of approximately 4.9 µg/m3 in the high-sulfur scenario and 4.6 µg/m3 in the low-sulfur scenario. Using both models, the total PM2.5 concentration levels were below both the European annual air quality guideline of 25 µg/m3 and the World Health Organization (WHO) recommended annual average of 10 µg/m3 [8]. The average population exposure in Gothenburg to PM2.5 from Baltic shipping was approximately 1.5 µg/m3 in the high-sulfur scenario and 1.6 µg/m3 in the low-sulfur scenario for all the different population grids for the city-scale model. Using the regional model, the average population exposures for the different population grids were approximately 0.5 µg/m3 in the high-sulfur scenario and 0.3 µg/m3 in the low-sulfur scenario (Table 2).
The mean population exposure-from-shipping difference before and after the revised SECA regulations was smaller in the city-scale model (Figure 2) than in the regional model (Figure 3). Exposure concentrations in the city-scale model show a clearer gradient from the areas closer to the shipping source as the exposure levels enter the inland areas, as opposed to the regional model (Figure S3). The choice of population dataset (data centroids, 1 km grid or 100 m grid) did not affect the mean population exposure in either the regional or city-scale models.

3.2. Exposure in Sweden

In Sweden as a whole, the mean population exposure levels of PM2.5 from Baltic shipping in low- and high-sulfur scenarios were 0.35 and 0.23 µg/m3, respectively. Generally, the highest exposure levels were in the southern parts of Sweden, especially in counties bordering the Baltic Sea (Figure 4). Skåne County recorded the highest contribution from shipping emissions to PM2.5 exposure in both the low-sulfur scenario (0.73 µg/m3) and the high-sulfur scenario (0.54 µg/m3), while Jämtland County recorded the lowest population-weighted levels in both scenarios, with 0.04 µg/m3 in the low-sulfur scenario and 0.02 µg/m3 in the high-sulfur scenario (Table 1).
The population density in Sweden varied from more than 260 inhabitants per km2 in Stockholm to 1 inhabitant per km2 in sparsely populated counties in north-western Sweden (Supplementary Figure S4).

3.3. Mortality

Levels of exposure to PM2.5 from shipping in the Baltic Sea and the consequent health effects decreased in both the city-scale and the regional models after the implementation of stricter SECA regulations in 2015. In Gothenburg, the city-scale model showed estimated premature deaths due to PM2.5 from shipping emissions decreasing from 97 to 90, while the regional model showed a decrease from 32 to 21. In the high-resolution model, the estimated YLL due to PM2.5 from shipping in Gothenburg also dropped from 1294 to 1206 and the average reduction in life expectancy fell from 17 hours to 16 per person. In the regional model, YLL decreased from 433 to 283 overall and from approximately 6 hours per person to 4.
Table 3 illustrates the estimated premature deaths from exposure to PM2.5 from Baltic Sea shipping emissions in the high- and low-sulfur scenarios with the stricter 2015 SECA sulfur fuel regulations using ESCAPE and HRAPIE ER functions. Stockholm County, the most densely populated, experienced a 44% reduction in premature deaths due to PM2.5 from shipping after the sulfur regulations were implemented. Other highly populated counties such as Skåne and Västra Götaland experienced 26% and 35% reductions, respectively. In Sweden, YLL decreased from 4161 in the high-sulfur scenario to 2680 in the low-sulfur scenario, a reduction of 36%.

3.4. Morbidity

In Gothenburg, using the average population exposure from the city-scale model, extra MI cases due to PM2.5 from Baltic Sea shipping decreased from 56 in the high-sulfur scenario to 52 in the low-sulfur scenario. Using the regional model, cases of MI decreased from 19 to 12. The estimated extra stroke cases dropped from 68 to 63 cases in the city-scale model and from 23 to 15 in the regional model.
Table 4 shows the estimated excess fatal MI and stroke cases due to PM2.5 exposure from shipping in the Baltic for Sweden and its counties for the high- and low-sulfur scenarios. MI cases in Sweden were 184 in the high-sulfur scenario and 118 in the low-sulfur scenario; stroke cases decreased from 274 to 177. The most populated counties recorded larger reductions in both MI and stroke from the high- and low-sulfur scenarios, and Stockholm County recorded the highest reductions. Reductions in Skåne, however, with the highest recorded exposure levels in both high- and low-sulfur scenarios (Table 1), were not as high as in other counties.
Using the city-scale model, the total burden of disease due to PM2.5 from Baltic Sea shipping in Gothenburg decreased from approximately 1374 to 1286 DALYs. Using the regional model, DALYs decreased from 513 to 363 over the same period. In Sweden, DALYs lost due to PM2.5 from Baltic Sea shipping decreased from approximately 5700 to approximately 4200 (Table 4).

4. Discussion

The stricter regional regulations of 2015 reduced the percentage of sulfur dioxide (a precursor of PM2.5) [16] in shipping fuel, and thereby decreased population exposure to PM2.5 from shipping. The largest reductions in PM2.5 concentrations were in the southern parts of Sweden. Skåne recorded the highest PM2.5 exposure levels due to its location next to the entrance to the Baltic Sea, next to major shipping lanes, while inland counties had lower levels of exposure to shipping emissions. The results for Gothenburg show that PM2.5 exposure from Baltic Sea shipping emissions was markedly higher in the city-scale model than in the regional model. The decrease in exposure between the high-sulfur and low-sulfur scenarios, however, was found to be larger in the regional model. Average population-weighted exposure levels were not affected by differences in population grid size (i.e., centroids, 1 × 1 km, and 100 × 100 m). This study indicates that the reductions in exposure levels after stricter sulfur regulations for shipping fuel in the Baltic Sea region have considerably decreased adverse health effects, with reductions in mortality and burden of disease from MI and stroke. However, estimations of these health effects were highly dependent on the exposure model used.

4.1. Exposure Assessment

When investigating the use of different resolution models to determine shipping emission exposure levels, we focused on the city of Gothenburg located in Västra Götaland County. In the city-scale model, shipping emissions accounted for 23% of the total PM2.5 concentration levels, while in the regional model they accounted for only 10%. The city-scale model produced a mean estimated population exposure approximately three times higher than the regional model (Table 2). Higher-resolution models have previously been shown to result in higher mean concentrations compared to lower resolution [49]. The estimated mean PM2.5 population-weighted exposure level of 1.5 µg/m3 in the city-scale model was relatively high compared with previous studies [5,24]. Those studies, however, made predictive estimates from data obtained prior to 2015 and modeled only local shipping as a separate source, not considering long-range emissions from outside the city-scale model domain. The long-range emissions could also be contributed from the North Sea, where stricter sulfur fuel regulations were simultaneously implemented in 2015, therefore a possible source of PM2.5 levels within the Swedish territory.
The differences between the regional- and city-scale models were highest closer to the sea and around the harbor inlet and were lower further inland, as illustrated in Figure 2 and Figure 3, and Figure S3. This is because the city-scale model more accurately captures emissions and concentration decays from the shipping routes, which are more important near strong sources. The city-scale model also provides a better estimation of how emission levels change from the source to actual population exposure. The regional model provides only average exposure levels over large grid squares, this difference illustrates the importance of using a city-scale model close to major sources of emissions. This is a smaller problem when calculating average county exposures, since average distances from shipping routes are larger. Sofiev et al. [14] encourages the use of higher-spatial resolution models that more accurately assess proximal exposure levels to at-risk communities.
We also found that different population resolutions did not influence mean population exposures. The size of the population grid (i.e., centroid, 1 km, or 100 m) had only a negligible effect on mean population exposures.
The regional model shows a greater reduction in PM2.5 from shipping emissions after SECA enforcement than the city-scale model (Figure 3a,b). This is because regional model concentrations are spread out over a larger area and it is not possible to reproduce the fine structure of maximum and minimum concentrations within the city with the 10 × 10 km squares equally as well as with the 100 × 100 m squares in the city-scale model. Other reasons for the low difference in the city-scale model could be that shipping in European Union port areas has already been using 0.1% fuel since 2010 [23,24], thus only the ship emissions at open sea outside the Gothenburg harbor area were impacted by the SECA regulation in 2015, and the much lower contribution from shipping in the harbor will affect the large grids in the regional model less. In addition, the city-scale model gives a more accurate concentration gradient from the source of the pollution (i.e., shipping) to the city and its inhabitants.
Due to the sensitivity, mostly from meteorological factors, high-resolution modeling is not always preferred for exposure modeling [50,51]. A study by Lui and Zhang [50] showed little or no improvements with model resolutions finer than the meteorological input data. Additionally, some studies have found that finer-resolution models do not necessarily improve model performance and so coarse resolution can still be favored [51]. However, Schaap et al. [52] showed marked improvements for short-lived species as a function of resolution, but the improvement was shown to be limited by the resolution of the input data. Furthermore, city-scale modeling can show a clearer pollution diffusion tendency with much higher concentrations in polluted areas than in the nearby areas, whereas with a coarser resolution, the spatial discrepancies are averaged out, making the distinction in the areas more difficult to determine [53]. City-scale modeling has also been found to be preferable in determining population-weighted exposure in epidemiological studies [49].
In the southern part of Sweden, more densely populated counties experienced a moderate reduction in exposure levels due to their location by the coast. Skåne recorded the highest PM2.5 exposure levels in both low- and high-sulfur scenarios. The county is located at the only entry and exit point for ships to and from the inner part of the Baltic Sea. Denmark, which is similarly located, also recorded the highest PM2.5 mean concentration in a similar study by Barregard et al. [28]. Counties located inland, such as Norrbotten and Jämtland, had minimal impact from shipping emissions due to their distance from the source of interest.

4.2. Mortality

The number of premature deaths attributed to exposure to PM2.5 from shipping emissions in city-scale model of Gothenburg in the low-sulfur scenario was 90 and 21 in the high-sulfur scenario. The number of premature deaths recorded in this study is lower than the estimated 140 premature deaths recorded by Tang et al. [25]. That study, however, predicted mortality based on 2012 data and assessed health impacts using the ALPHA-RiskPoll methodology, which includes other pollutants (NO2 and O3) as well as PM2.5. In the city-scale model in the present study, YLL per person calculated lower figures in the low-sulfur scenario in Gothenburg (approximately 16 hours YLL per person) than the just over 24 hours YLL per person calculated by Tang et al. [25].
In Sweden, there were an estimated 4161 YLL in the high-sulfur scenario, which were decreased by 36% to 2680 in the low-sulfur scenario due to the reduction in PM2.5. Areas with a high population density experienced the most significant reductions in YLL from exposure to PM2.5 from shipping. These reductions were also reflected in the number of premature deaths attributed to exposure to shipping emissions in individuals aged ≥30 years. Another study using country-level data observed a similar relative reduction in YLL from 4092 in a high-sulfur scenario to 2635 in a low-sulfur scenario [28]. The slight difference could be explained by our cutoff age of 30 years versus that study’s cutoff of 25; the number of cases also differed slightly due to a change in statistics and registries.

4.3. Morbidity

MI and stroke were chosen as indicators of the effects of emissions from Baltic shipping on morbidity in Sweden and its counties because they are serious health outcomes with strong evidence of associations with PM2.5 exposure [2,46,54]. Results indicated a reduction of 36% in first-time cases of both MI and stroke, corresponding to 65 annual cases of MI and almost 100 cases of stroke from the high- to low-sulfur scenarios. These reductions, although not large considering the population of Sweden, show that the stricter sulfur regulations implemented in 2015 have contributed to a reduction in morbidity. Furthermore, the counties located in the south experienced a relatively large reduction in first-time cases of MI and stroke. It can be assumed that near major shipping lanes and ports in the more densely populated coastal regions, the health impacts of PM2.5 from shipping are greater and the benefits of reduced emissions are larger [14].
The burdens of disease of MI and stroke differed between the high- and regional exposure models for Gothenburg. The number of DALYs decreased by 88 in the city-scale model and 150 in the regional model. Over 5700 DALYs due to MI and stroke in Sweden were attributed to PM2.5 in the high-sulfur scenario. This decreased by approximately 1400 DALYs in the low-sulfur scenario after the 2015 implementation of stricter SECA regulations. Stroke contributed more to the burden of disease than MI. A reduction in stroke cases relieves the health system of high healthcare expenses [55]. By implementing stricter emission regulations, Sweden and its counties experienced a relative reduction in DALYs that shows the benefits of such regulation. This is particularly significant because Sweden’s health system is funded by taxes [56]. A reduction in healthcare costs in one area increases funds available for other sectors in the health system. The stricter SECA regulations have also resulted in better overall health outcomes as predicted by Jonson et al. [23], and Brandt et al. [27].
Other European studies have shown that shipping emissions affect mortality and morbidity in surrounding populations [12,14,28]. Air pollution from shipping, however, affects not only the countries around the Baltic region, but also, through long-range (air current) transmission of PM2.5, areas further away [16]. The weather patterns in Northern Europe, including Sweden, mean that people spend over 80% of their time indoors during the winter. Since infiltration of outdoor air pollution is in part prevented by the filtration of air in homes and workplaces [57], it is possible that the population would be more affected during the warmer months when people spend more time outdoors. However, studies have also indicated that outdoor exposure still infiltrates indoor microenvironments such as homes and workplaces and puts individuals at risk of exposure in spite of spending a significant amount of time indoors [58,59]. Reducing air pollution from ship emissions will not only make the warmer months safer for exposed populations, but the reduced contribution of these emissions to climate change will likely influence the atmospheric chemistry and encourage governing bodies to improve on these beneficial regulations [14,25].
This study aimed to compare and contrast high- and low-spatial resolution models to account for and provide tangible results showing the differences between the two resolutions and their effects on estimated health outcomes calculated though DALYs. We found the city-scale model to be more likely to define concentrations of population exposure accurately. Although its practicality on a larger scale may be a challenge, similar methods can be used in cities that border the harbor. Using different grids for geocoding the population did not affect exposure estimations.

5. Limitations and Recommendations

Limitations to this study include a possible underestimation of the health effects associated with air pollution from shipping since the EMEP scenarios only considered emission changes in the Baltic Sea. The effects of the corresponding emission reductions in the North Sea have also contributed to the reductions in PM2.5 levels in Sweden but are not considered here. Additionally, the health effects of air pollution may have been underestimated since we only calculated effects of PM2.5 and did not include the effects of other air pollutants such as NOx, SOx, PM10, and O3 in this work, and the effects of PM2.5 were also, therefore, overestimated. Additionally, other health effects associated with air pollution such as chronic obstructive pulmonary disease [60], lower respiratory infections, type 2 diabetes [61], lower birth weight, or short gestation [62] were not considered in this study. Despite the inclusion in the ESCAPE project of several European countries, use of the ESCAPE cohort’s RR in studies to show the association between air pollution (namely PM2.5) and all-cause mortality [63] has drawn some criticism. Some studies used in the ESCAPE project show a moderate association between exposure to PM2.5 and all-cause mortality; others did not consider confounding factors that could have influenced the overall relative risk. For this reason, the results of this study regarding the premature deaths associated with PM2.5 may be somewhat uncertain. In addition, the study’s relative risks were derived from studies using land regression models and exposure assessments. We only estimated the change due to the stricter fuel restrictions from 1% to 0.1% in 2015 and could not analyze the effects of the earlier stepwise reductions. The populations for Sweden and the city of Gothenburg in both the high- and low-sulfur scenarios were used in the calculations with no adjustment for small population changes, which may have marginally affected the results. Further, the method of exposure modeling may have underestimated the exposure concentrations. The influence of meteorological conditions known to influence air pollution concentration levels in the atmosphere [64] was not accounted for and may also have influenced the estimation of adverse health effects. The study did, however, make estimations using empirical data and similar studies conducted in other coastal cities such as Stockholm and Malmö could contribute to our knowledge of the effects of pollution prevention measures.

6. Conclusions

The reduction in maximum sulfur content in shipping fuels led to decreased emissions and population exposures to PM2.5 from shipping in Sweden. Exposure estimations differed between the regional model and the city-scale model, with higher estimated population exposure in the latter, greatly influencing the health impact estimation. Interestingly, the spatial resolution of the population data did not have an impact on population-weighted PM2.5 exposure levels. Both the regional and the city-scale models showed a clear reduction after the implementation of the stricter 2015 SECA regulations on the Baltic Sea shipping emissions, leading to a decreased burden of disease due to outdoor PM2.5. Similar estimations may be made in the coming years to observe the effects of global sulfur fuel restrictions implemented in 2020 and other legislative changes.

Supplementary Materials

The following are available online at https://www.mdpi.com/1660-4601/17/21/7963/s1, Table S1: Summary of Baltic Sea ship and for the smaller domain around Gothenburg area emissions during 2014 and 2016. Figure S1: Emissions of SOx from Baltic Sea shipping in 2016. The unit of emissions is kg per grid cell area. Total SOx emissions during 2016 were 9727 tons, down from 75,148 tons during 2014. This reduction was because of 0.1%S required in the Baltic Sea SECA. Figure S2: Difference plot of SOx emissions in 2014–2016. Note, that reduction in emissions because of policy changes happens outside the Gothenburg harbor area. In the harbor area, no major difference is found, because the requirement to use 0.1%S fuel in harbor areas has been in place since 2010. The domain indicated by this image refers to Gothenburg area referred in the above table. Image courtesy of Landsat-8 and US Geological Survey. Figure S3: Comparison of differences in the contributions from Baltic Sea shipping between the high- and low-sulfur scenarios on PM2.5 (µg/m3) to Gothenburg; (a) regional model, (b) city-scale model. Figure S4: Population density in Sweden per km2.

Author Contributions

Conceptualization: L.S. and P.M.; methodology: P.M., N.S.M., and A.E.; formal analysis: N.S.M., P.M., J.E.J.; writing—original draft preparation: N.M.; writing—review and editing, N.S.M., J.E.J., E.S., A.E., J.-P.J., J.W., P.M. and. L.S.; supervision: L.S.; funding acquisition: L.S. and P.M. All authors have read and agreed to the published version of the manuscript.

Funding

This research is part of the EnviSuM project funded by the European Union via the Interreg Baltic Sea Region Program, grant number R027. It was also funded by the Sahlgrenska University Hospital under the ALF agreement, grant number ALFGBG-872511.

Acknowledgments

A special thanks to Lars Barregard for consultation on the paper and guidance throughout the process.

Conflicts of Interest

The authors declare no conflict of interest. The funders had no role in the design of the study, the collection, analysis, or interpretation of data, the writing of the manuscript, or the decision to publish the results.

References

  1. Health Effects Institute (HEI). State of Global Air 2019: A Special Report on Global Exposure to Air Pollution and Its Disease Burden; HEI: Boston, MA, USA, 2019. [Google Scholar]
  2. Brook, R.D.; Rajagopalan, S.; Pope, C.A.; Brook, J.R.; Bhatnagar, A.; Diez-Roux, A.V.; Holguin, F.; Hong, Y.; Luepker, R.V.; Mittleman, M.A.; et al. Particulate matter, air pollution and cardiovascular disease: An update to the Scientific Statement the American Heart Association. Circulation 2010, 121, 2331–2378. [Google Scholar] [CrossRef] [PubMed]
  3. Qian, Y.; Zhu, M.; Cai, B.; Yang, Q.; Kan, H.; Song, G.; Jin, W.; Han, M.; Wang, C. Epidemiological evidence on association between ambient air pollution and stroke mortality. J. Epidemiol. Commun. Health 2013, 67, 635–640. [Google Scholar] [CrossRef]
  4. World Health Organization. Metrics: Disability-Adjusted Life Year (DALY). Available online: https://www.who.int/healthinfo/global_burden_disease/metrics_daly/en/2020 (accessed on 14 April 2020).
  5. European Environment Agency (EEA). Air Quality in Europe—2018 Report; EEA: Copenhagen, Denmark, 2018. [Google Scholar]
  6. Segersson, D.; Eneroth, K.; Gidhagen, L.; Johansson, C.; Omstedt, G.; Engström Nylén, A.; Forsberg, B.; Ryan, L.; Anderson, C. Health impact of PM10, PM2.5 and black carbon exposure due to different source sectors in Stockholm, Gothenburg and Umea, Sweden. Int. J. Env. Res. Public Health 2017, 14, 742. [Google Scholar] [CrossRef] [PubMed]
  7. Burnett, R.; Pope, C.; Ezzati, M.; Lim, S.; Mehta, S.; Shin, H.; Singh, G.; Hubbell, B.; Brauer, M.; Anderson, H.R.; et al. An integrated risk function for estimating the global burden of disease attributable to ambient fine particulate matter exposure. Environ. Health Perspect. 2014, 122, 397. [Google Scholar] [CrossRef]
  8. Crouse, D.L.; Peters, P.A.; Van Donkelaar, A.; Goldberg, M.S.; Villeneuve, P.J.; Brion, O.; Khan, S.; Atari, D.O.; Jerrett, M.; Pope, C.A.; et al. Risk of nonaccidental and cardiovascular mortality in relation to long-term exposure to low concentrations of fine particulate matter: A Canadian national-level cohort study. Environ. Health Perspect. 2012, 120, 708–714. [Google Scholar] [CrossRef] [PubMed]
  9. Pope, A.C.; Burnett, T.R.; Krewski, E.D.; Jerrett, J.M.; Shi, J.Y.; Calle, J.E.; Thun, J.M. Cardiovascular mortality and exposure to airborne fine particulate matter and cigarette smoke: Shape of the exposure-response relationship. Circulation 2009, 120, 941–948. [Google Scholar] [CrossRef] [PubMed]
  10. Beelen, R.; Hoek, G.; Raaschou-Nielsen, O.; Stafoggia, M.; Andersen, Z.; Weinmayr, G.; Hoffmann, B.; Wolf, K.; Samoli, E.; Fischer, P.H.; et al. Natural-cause mortality and long-term exposure to particle components: An analysis of 19 European cohorts within the Multi-Center ESCAPE Project. Environ. Health Perspect. 2015, 123, 525–533. [Google Scholar] [CrossRef] [PubMed]
  11. Langrish, J.P.; Bosson, J.; Unosson, J.; Muala, A.; Newby, D.E.; Mills, N.L.; Blomberg, A.; Sandström, T. Cardiovascular effects of particulate air pollution exposure: Time course and underlying mechanisms. J. Intern. Med. 2012, 272, 224–239. [Google Scholar] [CrossRef]
  12. Beelen, R.; Raaschou-Nielsen, O.; Stafoggia, M.; Andersen, Z.J.; Weinmayr, G.; Hoffmann, B.; Wolf, K.; Samoli, E.; Fischer, P.; Nieuwenhuijsen, M.; et al. Effects of long-term exposure to air pollution on natural-cause mortality: An analysis of 22 European cohorts within the multicentre ESCAPE project. Lancet 2014, 383, 785–795. [Google Scholar] [CrossRef]
  13. Ledoux, F.K.; Foury, A.; Delmaire, G.; Roussel, G.; El Zein, A.; Courcot, D. Contributions of local and regional anthropogenic sources of metals in PM2.5 at an urban site in northern France. Chemosphere 2017, 181, 713–724. [Google Scholar] [CrossRef]
  14. Sofiev, M.; Winebrake, J.; Johansson, L.; Carr, E.; Prank, M.; Soares, J.; Vira, J.; Kouznetsov, R.; Jalkanen, J.; Corbett, J. Cleaner fuels for ships provide public health benefits with climate tradeoffs. Nat. Commun. 2018, 9, 406. [Google Scholar] [CrossRef] [PubMed]
  15. Lang, J.; Zhou, Y.; Chen, D.; Xing, X.; Wei, L.; Wang, X.; Zhao, N.; Zhang, Y.; Guo, X.; Han, L.; et al. Investigating the contribution of shipping emissions to atmospheric PM2.5 using a combined source apportionment approach. Environ. Pollut. 2017, 229, 557–566. [Google Scholar] [CrossRef] [PubMed]
  16. Finlayson-Pitts, B.J.; Pitts, J.N. Chemistry of the Upper and Lower Atmosphere; Academic Press: San Diego, CA, USA, 2000; pp. 1–969. [Google Scholar]
  17. European Envirnomental Agency. Air Quality in Europe—2019 Report; EEA Report No. 10/2019 2019; EEA: Copenhagen, Denmark, 2019. [Google Scholar]
  18. International Maritime Organization. Emission Control Areas (ECAs). 2020. Available online: https://www.atobviaconline.com/helpFiles/WebService/index.html?seca_and_eca_areas.htm (accessed on 14 April 2020).
  19. Department of Transportation Office & Ministry of Transport of the People’s Republic of China. ICP; ICP Report No. 05046837; Ministry of Transport of the People’s Republic of China: Beijing, China, 2015.
  20. European Commission (EC). Directive 2012/33/EU of the European Parliament and of the Council of 21 November 2012 Amending Council Directive 1999/32/EC as Regards the Sulphur Content of Marine Fuels; EC: Brussels, Belgium, 2012. [Google Scholar]
  21. Ledoux, F.; Roche, C.; Cazier, F.; Beaugard, C.; Courcot, D. Influence of ship emissions on NOx, SO2, O3 and PM concentrations in a North-Sea harbor in France. Int. J. Environ. Sci. 2018, 71, 56–66. [Google Scholar] [CrossRef] [PubMed]
  22. Thirteenth Meeting of the Focal Points of the Regional Marine Pollution Emergency Response Centre for the Mediterranean Sea (REMPEC). 2019. Available online: https://www.rempec.org/en/knowledge-centre/online-catalogue/e-fpm-2019-wg-45-inf-9-med-eca-study.pdf (accessed on 24 August 2020).
  23. Jonson, J.E.; Jalkanen, J.P.; Johansson, L.; Gauss, M.; Denier Van Der Gon, H.A.C. Model calculations of the effects of present and future emissions of air pollutants from shipping in the Baltic Sea and the North Sea. Atmos. Chem. Phys. 2015, 15, 783–798. [Google Scholar] [CrossRef]
  24. European Environmental Agency (EEA). The Impact of International Shipping on European Air Quality and Climate Forcing; EEA Technical Report No. 2013 4/2013; EEA: Copenhagen, Denmark, 2013. [Google Scholar]
  25. Tang, L.; Ramacher, M.; Moldanová, J.; Matthias, V.; Karl, M.; Johansson, L.; Jalkanen, J.; Yaramenka, K.; Aulinger, A.; Gustafsson, M. The impact of ship emissions on air quality and human health in the Gothenburg area—Part I: 2012 emissions. Atmos. Chem. Phys. Discuss. 2020, 20, 1–36. [Google Scholar] [CrossRef]
  26. Olstrup, H.; Johansson, C.; Forsberg, B.; Orru, H.; Nguyen, H.; Molnar, P. Trends in air pollutants and health impacts in three Swedish cities over the past three decades. Atmos. Chem. Phys. 2018, 18, 15705–15723. [Google Scholar] [CrossRef]
  27. Brandt, J.; Silver, J.D.; Christensen, J.H.; Andersen, M.S.; Bønløkke, J.H.; Sigsgaard, T.; Geels, C.; Gross, A.; Hansen, A.B.; Hansen, K.M.; et al. Assessment of past, present and future health-cost externalities of air pollution in Europe and the contribution from international ship traffic using the EVA model system. Atmos. Chem. Phys. 2013, 13, 7747–7764. [Google Scholar] [CrossRef]
  28. Barregard, L.; Molnár, P.; Jonson, J.; Stockfelt, L. Impact on population health of Baltic shipping emissions. Int. J. Env. Res. Public Health 2019, 16, 1954. [Google Scholar] [CrossRef]
  29. Viana, M.; Rizza, V.; Tobías, A.; Carr, E.; Corbett, J.; Sofiev, M.; Karanasioua, A.; Buonanno, G.; Fann, N. Estimated health impacts from maritime transport in the Mediterranean region and benefits from the use of cleaner fuels. Environ. Int. 2020, 138, 105670. [Google Scholar] [CrossRef] [PubMed]
  30. Jalkanen, J.P.; Brink, A.; Kalli, J.; Pettersson, H.; Kukkonen, J.; Stipa, T. A modelling system for the exhaust emissions of marine traffic and its application in the Baltic Sea area. Atmos. Chem. Phys. 2009, 9, 9209–9223. [Google Scholar] [CrossRef]
  31. Jalkanen, J.P.; Johansson, L.; Kukkonen, J.; Brink, A.; Kalli, J.; Stipa, T. Extension of an assessment model of ship traffic exhaust emissions for particulate matter and carbon monoxide. Atmos. Chem. Phys. 2012, 12, 2641–2659. [Google Scholar] [CrossRef]
  32. Johansson, L.; Jalkanen, J.-P.; Kukkonen, J. Global assessment of shipping emissions in 2015 on a high spatial and temporal resolution. Atmos. Environ. 2017, 167, 403–415. [Google Scholar] [CrossRef]
  33. IHS Markit. 2020. Available online: https://ihsmarkit.com/products/maritime-data-index.html (accessed on 7 October 2020).
  34. Jonson, J.; Gauss, M.; Jalkanen, J.P.; Johansson, L. Effects of strengthening the Baltic Sea ECA regulations. Atmos. Chem. Phys. 2019, 19, 13469–13487. [Google Scholar] [CrossRef]
  35. Simpson, D.; Benedictow, A.; Bergstrom, R.; Emberson, L.D.; Fagerli, H.; Flechard, C.R.; Hayman, G.D.; Gauss, M.; Jonson, J.E.; Jenkin, M.E.; et al. The EMEP MSC-W chemical transport model—Technical description. Atmos Environ. 2012, 12, 7825–7865. [Google Scholar] [CrossRef]
  36. AIRVIRO. 2020. Available online: https://www.airviro.com/airviro/extras/pdffiles/UserRef_Volume2_Dispersion_v4.00.pdf (accessed on 7 October 2020).
  37. Segersson, D. A Dynamic Model for Shipping Emissions Adaptation of Airviro and Application in the Baltic Sea; Meteorology No. 153; SMHI: Norrkoping, Sweden, 2013. [Google Scholar]
  38. Kukkonen, J.; Lopez-Aparicio, S.; Segersson, D.; Geels, C.; Kangas, L.; Kauhaniemi, M.; Maragkidou, A.; Jensen, A.; Assmuth, T.; Karppinen, A.; et al. The influence of residential wood combustion on the concentrations of PM2.5 in four Nordic cities. Atmos. Chem. Phys. 2020, 20, 4333–4365. [Google Scholar] [CrossRef]
  39. Olsson, D.; Johansson, C.; Forsberg, B. Associations between Vehicle Exhaust Particles and Ozone at Home Address and Birth Weight. Int. J Environ. Res. Public Health 2020, 17, 3836. [Google Scholar] [CrossRef]
  40. Environmental office, city of Gothenburg. Luftkvaliteten i Göteborgsområdet Årsrapport 2016, R 2017:06. [Air quality in the Gothenburg area Annual report 2016]. 2016. Available online: https://goteborg.se/wps/wcm/connect/d8f77f58-e8ba-4381-a4ec-fa49f732f273/N800_R_2017_6.pdf?MOD=AJPERES (accessed on 7 October 2020).
  41. European Commission. Eurostat: Your key to European statistics. 2015. Available online: https://ec.europa.eu/eurostat/web/gisco/geodata/reference-data/population-distribution-demography/geostat (accessed on 14 April 2020).
  42. Center for International Earth Science Information Network; Columbia University; Centro Internacional DedAgricultura Tropical. Gridded Population of the World; Version 3 (GPWv3): Centroids; NASA Socioeconomic Data and Applications Center: Palisades, NY, USA, 2005.
  43. Socialstyrelsen (National Board of Health and Welfare). Statistik. 2015. Available online: https://www.socialstyrelsen.se/statistik-och-data/statistik/ (accessed on 14 April 2020).
  44. Héroux, M.E.; Anderson, H.; Atkinson, R.; Brunekreef, B.; Cohen, A.; Forastiere, F.; Hurley, F.; Katsouyanni, K.; Krewski, D.; Krzyzanowski, M.; et al. Quantifying the health impacts of ambient air pollutants: Recommendations of a WHO/Europe project. Int. J. Public Health 2015, 60, 619–627. [Google Scholar] [CrossRef] [PubMed]
  45. Hoek, G.; Krishnan, R.M.; Beelen, R.; Peters, A.; Ostro, B.; Brunekreef, B.; Kaufman, J.D. Long-term air pollution exposure and cardio-respiratory mortality: A review. Environ. Health. 2013, 12, 43. [Google Scholar] [CrossRef]
  46. Stanaway, J.D.; Afshin, A.; Gakidou, E.; Lim, S.S.; Abate, D.; Abate, K.H.; Abbafati, C.; Abbasi, N.; Abbastabaret, H.; Abd-Allah, F.; et al. GBD Collaborators. Global, regional, and national comparative risk assessment of 84 behavioral, environmental and occupational, and metabolic risks or clusters of risks for 195 countries and territories, 1990-2017: A systematic analysis for the Global Burden of Disease study 2017. Lancet 2018, 392, 1923–1994. [Google Scholar]
  47. Cesaroni, G.; Forastiere, F.; Stafoggia, M.; Andersen, Z.J.; Badaloni, C.; Beelen, R.; Caracciolo, B.; de Faire, U.; Erbel, R.; Eriksen, K.T.; et al. Long term exposure to ambient air pollution and incidence of acute coronary events: Prospective cohort study and meta-analysis in 11 European cohorts from the ESCAPE project. BMJ 2014, 348, f7412. [Google Scholar] [CrossRef]
  48. Stafoggia, M.; Cesaroni, G.; Peters, A.; Andersen, Z.J.; Badaloni, C.; Beelen, R.; Caracciolo, B.; Cyrys, J.; de Faire, U.; de Hoogh, K.; et al. Long-term exposure to ambient air pollution and incidence of cerebrovascular events: Results from 11 European cohorts within the ESCAPE project. Environ. Health Perspect. 2014, 122, 919–925. [Google Scholar] [CrossRef] [PubMed]
  49. De Ridder, K.; Viaene, P.; Van De Vel, K.; Brasseur, O.; Cheymol, A.; Fierens, F. The impact of model resolution on simulated ambient air quality and associated human exposure. Atmoósfera 2014, 27, 403–410. [Google Scholar] [CrossRef]
  50. Liu, X.-H.; Zhang, Y. Understanding of the formation mechanisms of ozone and particulate matter at a fine scale over the southeastern U.S.: Process analyses and responses to future-year emissions. Atmos. Environ. 2013, 74, 259–276. [Google Scholar] [CrossRef]
  51. Queen, A.; Zhang, Y. Examining the sensitivity of MM5-CMAQ predictions to explicit microphysics schemes and horizontal grid resolutions, part iii—The impact of horizontal grid resolution. Atmos. Environ. 2008, 42, 3869–3881. [Google Scholar] [CrossRef]
  52. Schaap, M.; Cuvelier, C.; Hendriks, C.; Bessagnet, B.; Baldasano, J.M.; Colette, A.; Thunis, P.; Karam, D.; Fagerli, H.; Graff, A.; et al. Performance of European chemistry transport models as function of horizontal resolution. Atmos. Environ. 2015, 112, 90–105. [Google Scholar] [CrossRef]
  53. Tan, J.; Zhang, Y.; Ma, W.; Yu, Q.; Wang, J.; Chen, L. Impact of spatial resolution on air quality simulation: A case study in a highly industrialized area in Shanghai, China. Atmos. Pollut. Res. 2015, 6, 322–333. [Google Scholar] [CrossRef]
  54. Jaganathan, S.; Jaacks, L.M.; Magsumbol, M.; Walia, G.K.; Sieber, N.L.; Shivasankar, R.; Dhillon, P.K.; Hameed, S.S.; Schwartz, J.; Prabhakaran, D. Association of long-term exposure to fine particulate matter and cardio-metabolic diseases in low- and middle-income countries: A systematic review. Int. J. Env. Res. Public Health 2019, 16, 2541. [Google Scholar] [CrossRef]
  55. Ghatnekar, O.; Persson, U.; Asplund, K.; Glader, E.L. Costs for stroke in Sweden 2009 and developments since 1997. Int. J. Technol. Assess. 2014, 30, 203–209. [Google Scholar] [CrossRef] [PubMed]
  56. Lindgren, P.; Glader, E.L.; Jönsson, B. Utility loss and indirect costs after stroke in Sweden. Eur. J. Cardiovasc. Prev. Rehabil. 2008, 15, 230–233. [Google Scholar] [CrossRef] [PubMed]
  57. Hänninen, O.; Rumrich, I.; Asikainen, A. Challenges in estimating health effects of indoor exposures to outdoor particles: Considerations for regional differences. Sci. Total Environ. 2017, 589, 130–135. [Google Scholar] [CrossRef]
  58. Ramacher, M.O.P.; Karl, M.; Bieser, J.; Jalkanen, J.P.; Johansson, L. Urban population exposure to NOx emissions from local shipping in three Baltic Sea harbor cities—A generic approach. Atmos. Chem. Phys. 2019, 19, 9153–9179. [Google Scholar] [CrossRef]
  59. Singh, V.; Sokhi, R.S.; Kukkonen, J. An approach to predict population exposure to ambient air PM2.5 concentrations and its dependence on population activity for the megacity London. Environ. Pollut. 2020, 257, 113623. [Google Scholar] [CrossRef] [PubMed]
  60. Han, F.; Yang, X.; Xu, D.; Wang, Q.; Xu, D. Association between outdoor PM and prevalence of COPD: A systematic review and meta-analysis. Postgrad. Med. J. 2019, 95, 612–618. [Google Scholar] [CrossRef] [PubMed]
  61. Yang, Y.; Guo, Y.; Qian, Z.; Ruan, Z.; Zheng, Y.; Woodward, A.; Ai, S.; Howard, S.W.; Vaughn, M.G.; Ma, W.; et al. Ambient fine particulate pollution associated with diabetes mellitus among the elderly aged 50 years and older in China. Environ. Pollut. 2018, 243, 815–882. [Google Scholar] [CrossRef]
  62. Harris, G.; Thompson, W.D.; Fitzgerald, E.; Wartenberg, D. The association of PM2.5 with full term low birth weight at different spatial scales. Environ Res. 2014, 134, 427–434. [Google Scholar] [CrossRef] [PubMed]
  63. Lipfert, F.W. A critical review of the escape project for estimating long-term health effects of air pollution. Environ. Int. 2017, 99, 87–96. [Google Scholar] [CrossRef]
  64. Sanguineti, P.B.; Lanzaco, B.L.; Loópez, M.L.; Achad, M.; Palancar, G.G.; Olcese, L.E.; Toselli, B.M. PM2.5 monitoring during a 10-year period: Relation between elemental concentration and meteorological conditions. Environ. Monit. Assess. 2020, 192, 313. [Google Scholar] [CrossRef]
Figure 1. Sulfur emission control areas of the North and Baltic Seas.
Figure 1. Sulfur emission control areas of the North and Baltic Seas.
Ijerph 17 07963 g001
Figure 2. Exposure to PM2.5 (µg/m3) from shipping in the Baltic Sea in Gothenburg using a city-scale model: (a) high-sulfur; (b) low-sulfur.
Figure 2. Exposure to PM2.5 (µg/m3) from shipping in the Baltic Sea in Gothenburg using a city-scale model: (a) high-sulfur; (b) low-sulfur.
Ijerph 17 07963 g002
Figure 3. Exposure to PM2.5 (µg/m3) from shipping in the Baltic Sea in Gothenburg using a regional model: (a) high-sulfur; (b) low-sulfur.
Figure 3. Exposure to PM2.5 (µg/m3) from shipping in the Baltic Sea in Gothenburg using a regional model: (a) high-sulfur; (b) low-sulfur.
Ijerph 17 07963 g003
Figure 4. Comparison of the contribution from Baltic Sea shipping to PM2.5 (µg/m3) from ship emissions in Sweden and the surrounding area with the low-resolution EMEP model (a) high-sulfur; (b) low-sulfur.
Figure 4. Comparison of the contribution from Baltic Sea shipping to PM2.5 (µg/m3) from ship emissions in Sweden and the surrounding area with the low-resolution EMEP model (a) high-sulfur; (b) low-sulfur.
Ijerph 17 07963 g004
Table 1. Sweden’s 21 counties, populations in 2015, and population-weighted PM2.5 levels of exposure due to shipping using the regional model (1 × 1 km) grids, using both high- and low-sulfur scenarios.
Table 1. Sweden’s 21 counties, populations in 2015, and population-weighted PM2.5 levels of exposure due to shipping using the regional model (1 × 1 km) grids, using both high- and low-sulfur scenarios.
CountyPopulationAverage PM2.5 Population-Weighted Levels (µg/m3) High-Sulfur ScenarioAverage PM2.5 Population-Weighted Levels (µg/m3) Low-Sulfur Scenario
Sweden9,539,3930.350.23
Stockholm2,121,7820.440.25
Uppsala341,2980.230.14
Södermanland274,3910.250.15
Östergötland433,2200.250.16
Jönköping338,7940.250.16
Kronoberg185,6560.330.22
Kalmar233,3220.390.26
Gotland57,1630.470.28
Blekinge152,0540.540.38
Skåne1,259,8810.730.54
Halland304,9890.460.33
Västra Götaland1,615,6640.320.21
Värmland272,7620.120.07
Örebro282,7610.170.10
Västmanland255,7070.190.12
Dalarna276,2090.070.06
Gävleborg276,2820.120.07
Västernorrland241,7260.110.06
Jämtland126,0880.040.02
Västerbotten259,9630.090.05
Norrbotten248,3780.060.03
Table 2. Mean PM2.5 population exposure (µg/m3) from shipping emissions at city-scale and regional models using high- and low-sulfur scenarios for the Gothenburg area.
Table 2. Mean PM2.5 population exposure (µg/m3) from shipping emissions at city-scale and regional models using high- and low-sulfur scenarios for the Gothenburg area.
City-Scale ModelRegional Model
Population (Pop)High-SulfurLow-SulfurDifferenceHigh-SulfurLow-SulfurDifference
Pop centroids664,7151.601.490.110.530.340.19
Pop 1 km grid643,9641.611.500.110.540.360.19
Pop 100 m grid657,2431.601.490.110.540.350.19
Table 3. Estimated number of premature deaths due to PM2.5 emissions from Baltic shipping at the regional model in the high- and low-sulfur scenarios according to HRAPIE and ESCAPE exposure–response (ER) functions.
Table 3. Estimated number of premature deaths due to PM2.5 emissions from Baltic shipping at the regional model in the high- and low-sulfur scenarios according to HRAPIE and ESCAPE exposure–response (ER) functions.
CountyMortality at Age ≥30 in 2015Premature Deaths High-SulfurYears of Life Lost High-SulfurPremature Deaths Low-SulfurYears of Life Lost Low-SulfurReduction (%)
Sweden89,702196–4101820–4161126–2641172–268036
Stockholm15,16041–87507–115923–49283–64844
Uppsala27274–842–972–526–5939
Södermanland28844–937–843–622–5139
Östergötland41236–1358–1324–836–8337
Jönköping33525–1146–1063–730–6835
Kronoberg18234–833–752–522–5132
Kalmar27517–1450–1134–933–7534
Gotland6172–414–331–29–2039
Blekinge16686–1244–1024–831–7130
Skåne1165152–110495–113239–81366–83626
Halland27608–1675–1716–1254–12328
Västra Götaland1476929–61281–64219–40182–41735
Värmland32152–517–391–310–2438
Örebro29053–625–582–416–3637
Västmanland25723–627–612–416–3739
Dalarna31351–310–221–28–1916
Gävleborg33133–518–421–310–2444
Västernorrland29882–414–311–27–1648
Jämtland14810.3–12–60.1–0.31.12–2.556
Västerbotten26111–313–291–21–1547
Norrbotten27911–28–180.4–13–855
Table 4. Estimated number of extra annual cases of myocardial infarction and stroke due to PM2.5 emissions from Baltic shipping at the regional model in the high- and low-sulfur scenarios.
Table 4. Estimated number of extra annual cases of myocardial infarction and stroke due to PM2.5 emissions from Baltic shipping at the regional model in the high- and low-sulfur scenarios.
CountyExtra cases MI High-SulfurExtra Cases MI Low-SulfurReduction
(n)
Extra Cases Stroke High-SulfurExtra Cases Stroke Low-SulfurReduction
(n)
YLD MIYLD StrokeDALY High-SulfurDALY Low-Sulfur
Sweden18411865274177984155957244243
Stockholm36201660342712661425915
Uppsala422532046143105
Södermanland431.753204312794
Östergötland642853060193144
Jönköping532853055161124
Kronoberg43153202910480
Kalmar532753035148110
Gotland2113210104330
Blekinge64275202512796
Skåne493613765620119313261030
Halland9631073041212164
Västra Götaland271794428151253895670
Värmland2113210468570
Örebro2214210419978
Västmanland32143204110279
Dalarna210.2220.30608380
Gävleborg21142206010284
Västernorrland2112110427459
Jämtland0.30.10.10.40.20.20222825
Västerbotten2112110427157
Norrbotten10.50.6210.80.2496757
YLD: years lost due to disability, DALY: disability-adjusted life years.
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Back to TopTop