REPORT An investigation of air pollution on the decks of 4 cruise ships Authored by: Ryan David Kennedy, PhD, MAES Department of Health, Behavior & Society Johns Hopkins University, Bloomberg School of Public Health rdkennedy@jhu.edu 1 Suggested citation: Kennedy, RD. An investigation of air pollution on the decks of 4 cruise ships. A report for Stand.earth, January 24, 2019. EXECUTIVE SUMMARY are no internationally agreed-upon standards for what is considered safe human health exposure for ultrafine particulates, it is generally accepted that lower counts are better for health. Particulate matter (PM) pollution consists of small solids or liquid droplets suspended in the air. When inhaled, some PM can damage the heart and lungs. Particles less than 10 micrometers (μm) in diameter are a concern for public health because they can be inhaled deep into the lungs. Very small particles, classified as ultrafine particles or UFPs (100 nanometers or 0.1 μm), and even smaller nanoparticles (less than 100 nanometers) can cause airway inflammation and immunological reactions in the lungs and can travel through the bloodstream to affect other organs. The PM emitted from ship engines burning diesel fuel or heavy fuel oil is primarily composed of particles spanning from a few nanometers to less than one micron (0.001 μm–1.0 μm). With those concerns in mind, this study measured the concentration of PM (particles with sizes between 0.021.0 µm) on the deck of two Carnival Cruise Line ships in the Caribbean during October 2017 (Carnival Liberty) and April–May 2018 (Carnival Freedom); on the deck of a Holland America ship (MS Amsterdam) that traveled along the west coast of Canada and the U.S. during October 2018, and the deck of a Princess ship (Emerald) that traveled along the west coast of the U.S. and Mexico during November 2018. Measurements were taken inconspicuously in the bow area (fore of the smokestacks), as well as in two areas in the stern, aft of the ship’s smokestack including areas designated for exercise (running tracks, basketball court). Ship exhaust contains harmful constituents, including metals and polycyclic aromatic hydrocarbons (PAHs), many of which have toxic, mutagenic and/ or carcinogenic properties. There is strong evidence that ship engine exhaust impacts air quality in port and coastal cities, but little is known about how ship engine exhaust may impact the air quality on the deck of a ship. This is of concern because tens of millions of passengers take cruises every year, and tens of thousands of people are employed on cruise ships. The average PM readings in the stern areas of each ship were significantly higher than the average readings measured fore of the smokestacks (towards the bow) (paired t-tests, alpha 0.05). The findings of this study demonstrate that a source of PM—likely, in part from the ship’s exhaust system—is contributing to poorer air quality in the stern areas of these cruise ships. Concentrations of PM on the decks of these ships are comparable to concentrations measured in polluted cities, including Beijing and Santiago. Despite being on the open water and in open air, vacationers and cruise ship staff may be exposed to elevated concentrations of PM. Studies have demonstrated UFPs have detrimental effects to the cardiovascular and respiratory systems, including a higher incidence of atherosclerosis and increased disease severity of asthma. Even short-term exposure to UFPs from traffic exhaust is associated with adverse cardiovascular outcomes. Although there “Thick Black Smoke From The Sun Princess (2012)” (cover) by Jason Thien, “trieste (2016)” (above) by Antonio Marano are licensed under CC BY 2.0. 2 THE PROBLEM biological effects on human lung cells, and public health studies indicate that PM emitted from ship engine exhaust may be to blame for tens of thousands of annual deaths from lung and cardiovascular disease.7 Documented cellular responses to exposure to particulate matter from diesel exhaust include genotoxicity, oxidative stress and inflammatory signaling.10 The global shipping sector contributes to a range of environmental issues, including poor air quality and climate change.1, 2 It is speculated that cruise ships burn heavy fuel oil (HFO)3 wherever they are permitted to do so. HFO is a residual fuel generated during the distillation of crude oil. Cruise line companies represent a growing segment of the shipping sector. In 2017, a record 26.7 million passengers took a cruise, continuing a growth trend showing a 21% increase in passengers from 2011–2016.4 Globally, three companies dominate the cruise ship sector with the following market shares: Carnival (42%), Royal Caribbean International (22%) and Norwegian Cruise lines (8%).5 The number of crew members employed worldwide on cruise ships is estimated to be greater than 223,000.6 PM in the ultrafine and nanoparticle size range are relevant to public health because of their increased toxicity. UFP can have thousands of times more surface area than fine particles, which can affect the relative toxicity to the respiratory system, including greater deposition efficiency deep in the lungs.11 Further, ultrafine and nanoparticles are small enough to experience translocation—meaning that solid ultrafine or nanoparticles can be inhaled into the lungs and move directly into the blood and relocate to other parts of the body.12 Many in the environmental health research community have suggested that these smallest particles may be the most dangerous to human health.13-14 Shipping exhaust generated from burning HFO contains high levels of sulfur, as well as metals and polycyclic aromatic hydrocarbons (PAHs), many of which have toxic, mutagenic and/or carcinogenic properties.7 The impacts of cruise ship exhaust on the air quality of port and coastal communities have been documented,8-9 including the impact of particulate matter (PM). The size of particles in ship exhaust spans from a few nanometers to less than one micron (1μm).15 Little work has been done to quantify the contributions of ship exhaust to air quality on the decks on cruise ships, where guests and cruise ship staff commonly spend time. Journalists in Europe have conducted studies of particulate matter concentrations on passenger decks of a cruise ship using a hand-held PM monitoring device (P-TRAK 8525), which utilizes condensation particle-counting technology. The journalists measured PM concentrations and noted high concentrations of PM—in some cases, particle concentrations were 200 times higher than air measured in a park away from traffic and other sources of air pollution.16 PM is characterized by the size of particle. The following categories of PM are based on their aerodynamic diameter including coarse particles (PM10—particles ≤ 10 µm), fine particles (PM2.5— particles ≤ 2.5 µm), ultrafine particles or UFP (particles ≤ 0.1 µm or 100 nm), and nanoparticles (particles < 100 nm). Coarse and fine particles have been monitored and reported for decades, and epidemiological studies provide compelling evidence that PM pollution derived from fossil fuel combustion is an important cause of disease and premature death. PM generated from shipping exhaust has been found to show strong CURRENT STUDY of the smokestacks (see Figures 6–8). Each ship was subject to local winds that further influenced exhaust distribution and PM concentrations present in ambient air. Measurements were also taken in those same three environments while the ship was docked in port and exhaust plumes were subject only to local winds. The sampling techniques for this study were designed to inconspicuously measure PM in three environments on each of four cruise ships to understand how each ship’s exhaust may impact the air quality on each ship’s deck. Measurements were taken while each ship was moving at sea, whereby the forward propulsion of the ship would create spaces on the deck that were upwind and downwind of the ship’s smokestacks. The upwind environment was in the bow of the ships; the downwind environments included two areas aft The quantity and characterization of shipping exhaust differs based on a variety of factors, including what fuel is being burned and the efficiency of the engine. The 3 products. If a person walked by who was smoking/ vaping while a measurement was being taken, the test was not used, and it was re-done when appropriate. Particles generated from cooking or laundry would fluctuate throughout the day depending on the food being prepared and when laundry is dried. In the current study, PM measurements were conducted at different times of day, including early mornings and late evenings when cooking and laundry activities may have been reduced or were not taking place. Salt, water, and dust from the land represent ambient or background levels and likely contribute similar particulate matter concentrations at different locations on the deck of the ship if measured at a similar time. Tests in each of the three environments were always done as close together as possible. cruise ships observed in this study were continuously emitting exhaust through their smokestacks, because in addition to providing propulsion, the ship generates electricity to power the cabins, air conditioning, and general operation of the ship. PM measured on the deck of a cruise ship could originate from a variety of possible sources, including exhaust from the ship’s engines, the ship’s generators, salt and water particles from the sea, wind-borne dust particles from land-based sources, particles emitted from laundry activities, combustion particles from cooking or tobacco products and aerosols from electronic cigarettes/vaporizers. Effort was made to not measure particle concentrations if any passengers of staff were smoking or using vaping SHIPS STUDIED Carnival Liberty Holland America MS Amsterdam Figure 1. Carnival Liberty Source: James Willamor Figure 3. Holland America MS Amsterdam Source: Sabung.hamster Traveled to: Bahamas Ports visited: P ort Canaveral, Freeport, Nassau, Port Canaveral Sailing date: October, 2017 Guest Capacity: 2,974 Traveled along North American West Coast Ports visited: V ancouver, Astoria, San Francisco, Los Angeles Sailing date: October, 2018 Guest Capacity: 1,380 Carnival Freedom Emerald Princess Figure 2. Carnival Freedom Source: Jonathan Palombo Figure 4. Emerald Princess Source: Bahnfrend Traveled to: Western Caribbean Ports visited: G alveston, Montego Bay, Grand Cayman, Cozumel, Galveston Sailing date: April–May, 2018 Guest Capacity: 2,980 Traveled to: Mexico Ports visited: L os Angeles, Santa Barbara, Ensenada, Los Angeles Sailing date: November, 2018 Guest Capacity: 3,080 Carnival Liberty, Carnival Freedom, and Emerald Princess have very similar guest capacities. Holland America’s Amsterdam is a smaller ship with less than half the guest capacity compared to the Carnival ships. 4 EQUIPMENT isopropyl alcohol (99.5% pure) and with zero-check procedures conducted daily. The device used for each study had been factory calibrated within the last 12 months, as recommended by the manufacturer. The study measured particulate matter concentrations using a P-TRAK Ultrafine Particle Counter 8525 (the same device used in similar work in Europe). This portable instrument detects and counts particles within the size range of 0.02- 1 micrometer (0.02-1 µm or 20–1000 nm).17 This range of particle size aligns closely with the particle size distribution found in HFO, marine gas oil, and diesel exhaust.15, 18–19 Particle concentrations, reported as particles per cubic centimeter (Pt/cc), were recorded at a log interval of one minute. This one-minute concentration was calculated by averaging the PM concentrations measured each second over sixty-seconds. The P-Trak was operated as recommended, using high-purity Figure 5. P-Trak Ultrafine Particle Counter 8525, TSI Incorporated. (Source: Google Images) MONITORING areas were aft of the smokestacks in the stern. Measurements were also taken in the bow, fore of the smokestacks. Monitoring was done as a series of the three locations, with the order rotated in each series. Measurements in each series were done consecutively, resulting in comparable times for each location data set. Measurements were taken while the ship was in motion traveling at sea and while docked at port. Measurements of PM concentrations were recorded over several minutes in each location (range, 3–38 minutes). Environmental monitoring does not involve human subjects, and therefore this study was exempt from institutional review. Staff and patrons on the cruise ship were not made aware that measurements of particulate matter were being conducted. No staff or patron asked the researcher anything about the monitoring activities throughout the duration of the cruises. Every effort was made to not impact any guest’s experience on the cruise. The P-TRAK was fitted into a cloth bag with an intake tube used to draw air into the machine. PM measurements were taken in three open-air areas of the ship (see Figure 7–9). Two of the measurement Figure 6. Sampling areas of the Carnival Freedom and Carnival Liberty Figure 7. Sampling areas in the Holland America ship MS Amsterdam Figure 8. Sampling areas in the Princess ship Emerald 5 ANALYSIS reported for each ship in tables 1–4 below. Further, paired sample t-tests were conducted using an alpha of 0.05 to compare the mean particle concentrations in the bow to those in the stern and the track area or upper stern area. Mean particle concentrations (Pt/cc) were calculated based on duration of sampling, by location on ship deck while at sea and while in port. Maximum and minimum one-minute concentrations were identified for each environment under each condition. These are RESULTS The average (mean) particles per cubic centimeter (Pt/cc), and maximum 1-minute Pt/cc observed for each location, both in port and at sea, are reported in the Tables 1–4 below. Table 1. A verage and maximum particulate matter concentrations measured in different environments on the deck of the ship Carnival Liberty IN PORT AT SEA Average particle count Pt/cc Minutes of monitoring Maximum 1-minute concentration Pt/cc Average particle concentration Pt/cc Minutes of monitoring Maximum 1-minute concentration Pt/cc Stern 25,634 67 85,440 13,364 63 41,560 Track 33,514 31 46,018 13,150 54 45,063 Bow 6,126 27 19,178 4,171 48 9,668 Table 2. A verage and maximum particulate matter concentrations measured in different environments on the deck of the ship Carnival Freedom IN PORT AT SEA Average particle count Pt/cc Minutes of monitoring Maximum 1-minute concentration Pt/cc Average particle concentration Pt/cc Minutes of monitoring Maximum 1-minute concentration Pt/cc Stern 5,740 89 31,367 9,702 512 47,823 Track 11,880 73 56,091 12,747 512 73,621 Bow 15,604 100 119,983 1,540 523 14,533 Table 3. A verage and maximum particulate matter concentrations measured in different environments on the deck of the ship Holland America MS Amsterdam IN PORT AT SEA Average particle count Pt/cc Minutes of monitoring Maximum 1-minute concentration Pt/cc Average particle concentration Pt/cc Minutes of monitoring Maximum 1-minute concentration Pt/cc Stern 2,284 89 38,333 16,831 375 76,780 Upper Stern 4,217 88 13,711 14,508 362 43,486 Bow 5,904 95 25,290 2,284 366 12,781 6 Table 4. A verage and maximum particulate matter concentrations measured in different environments on the deck of the ship Emerald Princess IN PORT AT SEA Average particle count Pt/cc Minutes of monitoring Maximum 1-minute concentration Pt/cc Average particle concentration Pt/cc Minutes of monitoring Maximum 1-minute concentration Pt/cc Upper Stern 6,502 42 15,416 30,647 269 144,500 Lower Stern 8,234 42 17,140 32,628 268 157,716 33,408 43 126,786 5,167 257 24,696 Bow While in port, particle counts on the Carnival Liberty were generally higher than at sea, with average particle counts four to five times higher aft of the smokestacks than in the bow. On the Carnival Freedom, particle counts were lowest on the stern, but on the bow still measured approximately four times higher than particle counts while at sea. The Holland America ship, the MS Amsterdam, had much lower particle counts in the stern in port as compared to when the ship was at sea. On the Princess’ ship Emerald, port readings in the bow were comparable to readings observed in the stern while the ship was moving. While all four ships were traveling at sea, average particle counts (Pt/cc) were significantly higher in the areas aft of the smokestacks towards the stern, compared to areas towards the bow (forward of the smokestacks) (p<0.05). On the Carnival Liberty, particle counts in the stern or running track area were approximately three times higher than the bow. On the Carnival Freedom, particle counts measured on the stern were approximately six times higher, and those on the running track were eight times higher, than concentrations measured near the bow. On the Holland America MS Amsterdam, average particle counts were approximately eight times higher in the stern areas compared to the concentrations measured near the bow. On the Princess ship Emerald, particle counts in the stern area were approximately 6 times higher in the stern areas compared to the bow. For detailed findings from each test conducted on each ship, refer to the tables presented in Appendix A–D. Using paired t-tests (compare means), it was found that the difference in means between stern areas and the bow areas were found to be statistically significant (alpha 0.05). See Appendix E for each paired t-test result. Test-by-test readings, including maximum and minimum one-minute particle counts measured on each ship, are detailed at the end of the report in the appendices. 7 DISCUSSION The results of this study demonstrate that while each cruise ship was at sea, concentrations of particulate matter were significantly higher in the areas aft of the smokestacks compared to the bow area. This suggests that a source between the bow and the aft of the ship contributed particulate matter. The size of particulate matter measured in this study aligns closely with the size of particles known to be generated by ship engines, and the ship’s exhaust system is located between the environments with disparate PM measurements, suggesting the source of the particulate matter is likely, in part, the ships’ engine exhaust. P-TRAK device, it is possible to compare the findings of this study to other environments. While at sea, average particulate matter concentrations in the areas aft of the smokestacks of the ships ranged between 9,702–32,628 Pt/cc, with a maximum one-minute particulate matter concentration of 157,716 Pt/cc. There is not universal agreement on how to measure or report particulate matter from the UFP or nanoparticle size range. However, by reviewing studies that have similarly measured outdoor particulate matter with a There remain some unknowns with this study, including which fuel types were being used by the ships throughout the voyage and how efficiently the engines were operating. Differences in particulate matter concentrations observed in port and at sea may be explained by local winds distributing ship exhaust differently, and therefore contributing to higher concentrations in the bow area, compared to when the ship is moving (see appendices). “Ship engine emissions are important with regard to lung and cardiovascular diseases especially in coastal regions worldwide … Epidemiological studies attribute up to 60,000 annual deaths from lung and cardiovascular disease to ship engine [particulate matter].” Oeder S, Kanashova T, Sippula O,Sapcariu SC, Streibel T, Arteaga-Salas JM, et al. (2015) Particulate Matter from Both Heavy Fuel Oil and Diesel Fuel Shipping Emissions Show Strong Biological Effects on Human Lung Cells at Realistic and Comparable In Vitro Exposure Conditions. PLoS ONE 10(6): e0126536. doi:10.1371/journal. pone.0126536 Other studies from around the world measuring particulate matter in outdoor settings with a P-Trak are presented here: Based on the findings of these other studies, particulate matter concentrations observed in different parts of the cruise ships’ decks were comparable to some polluted environments, including urban settings in Asia, South America, and Europe. • UFP levels in train stations in Taipei, measured using a P-Trak in 2009, averaged 15,500 Pt/cc.20 The results of this study suggest that patrons and staff who are in the aft areas of cruise ships, like those in this study, are likely exposed to elevated levels of particulate matter, and that some of the particulate matter observed in this study was likely generated by shipping exhaust. This raises health concerns, given that ship exhaust is known to contain such dangerous constituents as metals and PAHs.7 • An air quality study measured UFP with a P-Trak in different locations in Beijing, China in 2009. Average UFP concentration was 30,000 Pt/cc on a busy street.21 • A study in industrialized southeast China measured concentrations of ultrafine particles in a high-traffic area using a P-TRAK; median value was 45,805.22 • UFP concentrations measured with a P-Trak in Santiago, Chile found ranges of 8,000–30,100 Pt/ cc.23 Exposure to UFP is not advisable to any population; however, the US EPA and the World Health Organization warn that specific groups within the general population, including children, the elderly and people with respiratory and cardiovascular disease,25-26 may have a greater risk of pollution effects. • Outdoor median concentrations of UFPs were measured with a P-Trak in Antwerp, Belgium 15,600 Pt/cc.24 8 APPENDIX A Particulate matter concentrations (Pt/cc) measured on the Carnival Liberty (Pilot study) Test Date Start time Location 1 Stern 2 Track 3 Bow 1 Oct 22 22:04 1 Oct 22 1 Particle concentration (Pt/cc) Average 1-minute Max 1-minute Min # of data points (# of minutes) 3 464 476 456 4 1 22:13 2 6,334 7,662 4,666 4 1 Oct 22 22:25 1 7,420 10,063 6,115 4 1 2 Oct 23 6:49 1 50,323 64,935 42,467 7 2 2 Oct 23 7:00 3 19,178 24,240 13,265 4 2 2 Oct 23 7:08 1 28,744 41,662 22,494 4 2 2 Oct 23 14:51 1 14,104 30,695 6,985 13 2 1 At sea 2 In port 3 Oct 23 15:06 2 4,837 7,184 3,602 7 2 3 Oct 23 15:19 3 2,816 6,048 692 8 2 4 Oct 23 16:39 1 28,608 32,351 22,578 3 1 4 Oct 23 16:43 2 19,023 45,063 4,687 18 1 4 Oct 23 17:07 3 911 1,131 561 5 1 5 Oct 24 7:10 3 6,686 14,744 2,509 8 2 5 Oct 24 7:21 1 20,001 37,365 6,474 38 2 6 Oct 24 14:34 1 61,366 85,440 34,891 5 2 6 Oct 24 15:04 2 43,018 62,650 23,383 7 2 6 Oct 24 15:17 3 1,812 5,040 542 7 2 6 Oct 24 15:33 2 41,408 67,163 16,740 17 2 7 Oct 24 21:50 1 22,456 41,560 502 10 1 7 Oct 24 22:05 3 9,668 18,371 3,441 8 1 7 Oct 24 22:16 2 8,884 15,450 4,132 6 1 8 Oct 25 13:40 1 13,942 17,013 10,444 13 1 8 Oct 25 13:56 2 9,551 11,288 8,710 6 1 8 Oct 25 14:17 3 2,978 3,597 2,303 10 1 9 Oct 25 18:34 1 12,744 16,382 10,632 17 1 9 Oct 25 18:52 2 16,440 34,361 6,673 9 1 9 Oct 25 19:04 3 3,970 4,114 3,843 12 1 10 Oct 25 22:04 1 8,046 9,487 6,525 5 1 10 Oct 25 22:10 2 8,052 14,543 5,871 8 1 10 Oct 25 22:21 3 4,185 4,839 3,547 4 1 11 Oct 26 5:43 3 4,456 7,079 3,769 5 1 11 Oct 26 5:51 2 6,389 8,087 3,374 7 1 11 Oct 26 5:59 1 5,483 9,995 4,277 7 1 9 APPENDIX B Particulate matter concentrations (Pt/cc) measured on the Carnival Freedom Test group Date Start time Location 1 Stern 2 Track 3 Bow 1 April 28 16:06 1 April 28 1 Particle concentration (Pt/cc) Average 1-minute Max 1-minute Min # of data points (# of minutes) 2 13,762 53,539 1982 14 1 16:23 1 6,933 10,069 4,522 11 1 April 28 16:38 3 1,871 2,101 1,681 7 1 2 April 28 18:30 1 8,701 11,414 6,596 12 1 2 April 28 18:44 2 8,734 9,955 3,769 11 1 2 April 28 18:58 3 1,810 1,922 1,690 16 1 3 April 28 21:35 2 6,230 9,898 1,644 16 1 1 At sea 2 In port 3 April 28 2153 1 2,901 3,944 1,653 9 1 3 April 28 2207 3 1,173 1,290 1,047 17 1 4 April 29 9:12 3 783 840 748 10 1 4 April 29 9:25 2 8,582 34,779 2,811 12 1 4 April 29 9:41 1 4,925 7,044 2,809 11 1 5 April 29 12:55 1 9,756 11,896 7,507 13 1 5 April 29 13:10 2 6,197 15,456 1,549 11 1 5 April 29 13:25 3 1,257 1,373 1,075 14 1 6 April 29 16:55 2 12,581 43,750 1,039 14 1 6 April 29 17:12 1 3,095 4,121 2,629 9 1 6 April 29 17:25 1 3,689 5,048 2,759 12 1 6 April 29 17:40 3 918 985 828 13 1 7 April 29 21:20 1 6,304 7,114 4,988 17 1 7 April 29 21:39 2 6,667 14,445 686 11 1 7 April 29 21:54 3 737 796 701 16 1 8 April 30 5:22 3 361 501 291 13 1 8 April 30 5:38 2 1,535 3,670 292 11 1 8 April 30 5:50 1 3,226 4,308 540 12 1 9 April 30 8:40 1 9,177 12,290 3,983 20 1 9 April 30 9:02 2 19,583 32,155 7,032 30 1 9 April 30 9:36 3 1,158 1,286 1,080 23 1 10 April 30 12:44 2 12,481 22,670 1,277 12 1 10 April 30 12:57 1 12,501 15,699 11,082 11 1 10 April 30 13:12 3 758 830 716 20 1 11 April 30 16:23 1 10,685 12,916 8,915 17 1 11 April 30 16:42 2 16,661 21,756 12,137 17 1 11 April 30 17:02 3 450 504 414 22 1 12 April 30 19:30 1 5,310 7,199 3,632 21 1 12 April 30 19:53 2 12,149 17,125 7,153 34 1 12 April 30 20:30 3 415 442 396 13 1 13 May 1 4:55 3 499 599 383 30 1 13 May 1 5:29 2 12,210 16,963 5,896 24 1 13 May 1 5:55 1 4,416 6,550 2,960 29 1 10 14 May 1 8:05 3 5,668 6,513 4,934 11 1 14 May 1 8:20 1 9,762 18,563 4,248 11 1 14 May 1 8:32 2 12,066 37,621 6,704 15 1 15 May 1 15:38 1 3,652 6,147 2,086 9 2 15 May 1 15:49 2 2,886 4,506 2,225 12 2 15 May 1 16:05 3 781 962 580 13 2 16 May 1 18:02 2 2,967 3,650 2,361 11 1 16 May 1 18:15 1 4,035 5,895 2,665 10 1 16 May 1 18:29 3 645 1,057 356 15 1 17 May 1 20:48 2 5,579 9,187 1,001 24 1 17 May 1 21:14 1 4,118 6,693 2,561 23 1 17 May 1 21:44 3 436 527 382 27 1 18 May 2 6:23 1 8,326 17,038 4,671 10 1 18 May 2 6:35 2 8,913 15,035 2,167 11 1 18 May 2 6:49 3 554 630 494 12 1 19 May 2 14:29 1 1,251 4,492 525 11 2 19 May 2 14:42 2 1,357 2,454 568 9 2 19 May 2 14:55 3 21,246 44,513 1,057 35 2 20 May 2 15:52 3 1,700 2,678 1,367 11 2 20 May 2 16:08 1 18,151 31,367 10,820 16 2 20 May 2 16:25 2 24,607 56,091 2,460 25 2 21 May 2 19:41 1 16,428 18,819 14,749 13 1 21 May 2 20:12 2 7,292 9,617 5,963 15 1 21 May 2 20:31 3 1,601 1,735 1,516 15 1 22 May 2 21:49 3 1,844 2,315 1,323 20 1 22 May 2 22:13 2 6,008 10,796 2,262 21 1 22 May 2 22:40 1 11,756 14,306 7,776 21 1 23 May 3 4:46 1 20,785 40,308 6,558 20 1 23 May 3 5:09 2 25,007 36,790 2,942 30 1 23 May 3 5:46 3 3,114 3,260 2,911 29 1 24 May 3 8:42 1 10,745 20,665 5,442 21 1 24 May 3 9:05 2 40,756 73,621 8,795 21 1 24 May 3 9:33 3 18,590 119,983 3,567 25 2 25 May 3 11:45 1 1,950 3,247 1,627 30 2 25 May 3 12:17 2 5,855 12,523 3,455 21 2 25 May 3 12:43 3 18,479 52,921 4,187 30 2 26 May 3 16:40 1 2,402 3,741 1,835 15 2 26 May 3 17:00 3 18,856 56,108 4,218 17 2 26 May 3 17:28 3 18,008 40,956 2,512 19 2 27 May 3 18:09 1 13,716 47,823 2,733 30 1 27 May 3 18:42 3 1,200 1,349 744 25 1 28 May 3 22:00 1 15,330 17,932 12,847 13 1 28 May 3 22:12 2 15,561 18,918 13,859 6 1 28 May 3 22:24 3 1,547 1,813 1,329 14 1 29 May 4 6:03 1 8,517 21,555 3,432 31 1 29 May 4 6:40 3 2,862 10,680 824 31 1 11 29 May 4 7:18 2 18,950 42,420 7,379 33 1 30 May 4 10:00 1 13,476 24,430 7,218 33 1 30 May 4 10:49 2 17,167 29,186 10,851 31 1 30 May 4 11:26 3 1,828 2,325 1,646 33 1 31 May 4 15:13 1 18,788 40,646 2,473 27 1 31 May 4 15:51 2 5,672 12,558 1,423 26 1 31 May 4 16:40 3 1,286 1,707 1,159 30 1 32 May 4 19:12 1 9,854 13,274 8,267 17 1 32 May 4 19:33 2 6,917 14,429 4,098 20 1 32 May 4 19:56 3 1,480 1,641 1,376 19 1 33 May 4 22:05 1 11,554 15,084 7,703 13 1 33 May 4 22:20 2 6,415 14,308 5,109 14 1 33 May 4 22:37 3 1,824 1,948 1,721 14 1 34 May 5 5:09 1 3,167 5,055 2,323 15 1 34 May 5 5:26 2 2,541 5,773 1,968 15 1 34 May 5 5:46 3 5,925 14,533 3,114 14 1 35 May 5 7:43 2 13,707 27,677 8,608 6 2 35 May 5 7:50 1 9,915 17,476 5,558 8 2 35 May 5 8:04 3 2,972 3,303 2,630 9 2 12 APPENDIX C Particulate matter concentrations (Pt/cc) measured on the Holland America MS Amsterdam Test Test group Date Start time Location 1 Stern 2 Track 3 Bow 3 1 Sep 25 15:39 4 1 Sep 25 5 1 6 Particle concentration (Pt/cc) Average 1-minute Max 1-minute Min # of data points (# of minutes) 2 7,213 13,711 3,529 22 2 16:04 3 16,056 25,290 7,489 20 2 Sep 25 16:28 1 15,321 38,333 8,284 25 2 2 Sep 25 18:38 1 21,251 45,561 4,387 25 1 8 2 Sep 25 19:39 2 4,589 19,116 218 21 1 9 2 Sep 25 20:05 3 1,498 2,415 1,002 24 1 11 3 Sep 25 23:04 3 4,003 7,385 2,221 22 1 1 At sea 2 In port 12 3 Sep 25 23:29 2 8,057 32,770 3,106 23 1 13 3 Sep 25 23:52 1 46,106 76,780 18,930 26 1 14 4 Sep 26 8:23 1 30,303 41,820 20,385 25 1 15 4 Sep 26 8:48 2 29,879 37,486 22,845 23 1 16 4 Sep 26 9:15 3 2,580 3,086 23,318 23 1 20 5 Sep 26 17:32 3 2,661 3,198 2,502 27 2 22 5 Sep 26 18:12 1 3,230 4,067 2,774 21 2 23 5 Sep 26 18:34 2 2,539 2,663 2,256 21 2 24 6 Sep 26 21:07 3 2,621 11,046 1,786 20 1 25 6 Sep 26 21:30 2 2,062 2,852 1,845 20 1 26 6 Sep 26 21:51 1 1,906 1,991 1,862 21 1 27 7 Sep 27 7:00 2 3,276 5,843 2,707 24 1 28 7 Sep 27 7:25 1 3,209 3,910 2,665 24 1 29 7 Sep 27 7:52 3 8,311 12,781 2,461 23 1 30 8 Sep 27 10:05 1 14,614 21,695 7,095 23 1 31 8 Sep 27 10:29 2 3,195 5,159 2,346 22 1 32 8 Sep 27 10:55 3 2,140 2,334 2,038 23 1 33 9 Sep 27 12:39 3 1,615 2,229 1,243 20 1 34 9 Sep 27 13:02 2 1,825 2,481 1,582 26 1 35 9 Sep 27 13:28 1 16,506 33,618 8,663 22 1 36 10 Sep 27 16:54 3 1,073 1,166 1,020 22 1 37 10 Sep 27 17:19 1 23,106 27,425 17,951 23 1 38 10 Sep 27 17:43 2 28,600 32,081 20,255 22 1 39 11 Sep 27 21:41 3 896 829 755 23 1 40 11 Sep 27 22:06 2 33,633 43,486 23,172 21 1 41 11 Sep 27 22:49 1 27,108 34,246 18,237 21 1 42 12 Sep 28 4:45 2 16,347 25,160 5,454 21 1 43 12 Sep 28 5:07 3 1,009 1,063 957 21 1 44 12 Sep 28 5:33 1 10,720 20,021 6,796 23 1 45 13 Sep 28 9:23 1 2,439 6,216 1,639 21 2 46 13 Sep 28 9:44 2 2,250 3,925 1,481 22 2 47 13 Sep 28 10:08 3 2,877 5,188 1,339 22 2 48 14 Sep 28 21:03 1 4,298 4,876 3,533 22 2 13 49 14 Sep 28 21:29 2 4,763 5,835 3,510 23 2 50 14 Sep 28 21:48 3 4,024 4,694 3,100 26 2 51 15 Sep 29 6:20 2 23,795 25,968 20,206 23 1 52 15 Sep 29 6:44 1 15,284 16,392 12,883 21 1 53 15 Sep 29 7:10 3 691 772 620 22 1 54 16 Sep 29 10:55 3 775 2,307 597 26 1 55 16 Sep 29 11:24 2 12,639 19,150 12,639 26 1 56 16 Sep 29 11:50 1 7,973 14,511 2,778 30 1 57 17 Sep 29 14:17 2 11,877 14,858 7,464 25 1 58 17 Sep 29 14:43 1 12,521 15,505 1,325 24 1 59 17 Sep 29 15:08 3 598 686 500 27 1 60 18 Sep 29 18:27 3 945 1,756 407 26 1 61 18 Sep 29 18:58 2 17,587 22,216 12,844 23 1 62 18 Sep 29 19:23 1 12,206 16,771 6,358 22 1 63 19 Sep 29 22:06 3 419 479 368 21 1 64 19 Sep 29 22:29 2 10,087 19,311 2,750 21 1 65 19 Sep 29 22:53 1 2,271 8,089 1,166 22 1 66 20 Sep 30 5:13 1 20,924 27,980 13,195 23 1 67 20 Sep 30 5:35 2 27,491 30,623 21,533 21 1 68 20 Sep 30 5:59 3 7,706 9,607 5,602 23 1 14 APPENDIX D Particulate matter concentrations (Pt/cc) measured on the Emerald Princess Location 1 U. Stern Start time 2 L. Stern 3 Bow Average 1-minute Max 1-minute Min # of data points (# of minutes) 1 11,808 29,615 5,163 24 1 17:04 2 31,783 113,630 11,869 22 1 Nov 25 17:31 3 4,130 5,032 3,891 22 1 2 Nov 25 19:59 3 7,703 16,136 5,254 22 1 7 2 Nov 25 20:24 1 31,981 124,738 4,906 22 1 8 2 Nov 25 20:51 2 44,139 100,058 25,138 24 1 9 3 Nov 26 5:18 1 29,491 41,910 20,835 24 1 10 3 Nov 26 5:50 2 29,123 52,801 4,377 25 1 11 3 Nov 26 6:18 3 6,084 9,082 4,479 23 1 12 4 Nov 26 11:00 3 3,822 8,999 3,027 22 2 13 4 Nov 26 11:25 1 7,309 15,416 4,012 21 2 14 4 Nov 26 11:51 2 12,582 17,140 8,191 22 2 16 5 Nov 26 18:56 2 7,612 12,188 4,262 24 1 17 5 Nov 26 19:24 1 37,754 144,500 3,575 22 1 18 5 Nov 26 19:54 3 5,634 6,815 4,645 21 1 19 6 Nov 27 1:30 1 25,152 45,386 5,896 22 1 20 6 Nov 27 1:54 2 50,900 71,786 30,195 21 1 Test Test group Date 2 1 Nov 25 16:34 4 1 Nov 25 5 1 6 Particle concentration (Pt/cc) 1 At sea 2 In port 21 6 Nov 27 2:19 3 12,773 17,026 10,022 21 1 22 7 Nov 27 6:59 3 4,845 5,167 4,680 21 1 23 7 Nov 27 7:24 2 26,113 34,800 17,743 21 1 24 7 Nov 27 7:47 1 27,467 36,850 8,174 23 1 26 8 Nov 27 12:45 3 1,592 1,606 1,569 21 1 28 8 Nov 27 13:39 2 23,940 64,180 6,048 21 1 29 8 Nov 27 14:12 1 58,186 96,820 8,786 21 1 30 9 Nov 27 18:20 2 14,786 72,905 1,323 21 1 31 9 Nov 27 18:54 3 6,097 24,696 994 22 1 32 9 Nov 27 19:22 1 28,568 59,026 8,527 24 1 33 10 Nov 27 21:43 3 1,651 3,444 1,294 21 1 34 10 Nov 27 22:08 1 32,837 51,721 9,913 21 1 35 10 Nov 27 22:34 2 15,439 24,670 7,297 21 1 36 11 Nov 28 6:04 3 5,595 6,587 4,306 21 1 37 11 Nov 28 6:31 2 96,289 157,716 38,018 21 1 38 11 Nov 28 6:54 1 33,579 78,796 6,670 21 1 39 12 Nov 28 15:28 2 3,679 6,960 1,630 21 2 40 12 Nov 28 15:53 1 5,695 12,642 982 21 2 41 12 Nov 28 16:30 3 64,402 126,786 13,864 21 2 42 13 Nov 28 18:45 3 1,245 3,859 312 21 1 43 13 Nov 28 19:10 1 30,717 70,995 8,228 21 1 44 13 Nov 28 19:36 2 48,878 98,371 22,686 21 1 45 14 Nov 29 4:49 2 10,452 56,766 1,797 26 1 15 46 14 Nov 29 5:17 1 24,429 52,210 6,434 24 1 47 14 Nov 29 5:47 3 4,446 13,833 1,214 21 1 APPENDIX E – PAIRED T-TESTS Liberty t n p Sd 3.1444 7 0.019956 8663 Bow vs. track area 2.436422 7 0.030849 6465 Stern vs. track area  -0.879629 7 0.4129 6665 Bow vs. stern The difference between the average of the stern minus bow and μ0 is big enough to be statistically significant. The difference between the average of the track minus bow and μ0 is big enough to be statistically significant. The difference between the average of the track minus stern and μ0 is not big enough to be statistically significant. Freedom t n p Sd Bow vs. stern 7.932814 27 2.07E-08 4891 Bow vs. track area 7.818732 27  2.71118e-8 5800 Stern vs. track area 1.14615 27 0.262173 5710 The difference between the average of the stern minus bow and μ0 is big enough to be statistically significant. The difference between the average of the track minus bow and μ0 is big enough to be statistically significant. The difference between the average of the track minus stern and μ0 is not big enough to be statistically significant. Amsterdam t n p Sd Bow vs. stern 4.315514 16 0.00061 11474 Bow vs. upper stern 4.920617 16 0.00018 11641 Stern vs. upper stern -0.62143 16 0.54365 12499 The difference between the average of the stern minus bow and μ0 is big enough to be statistically significant. The difference between the average of the upper stern minus bow and μ0 is big enough to be statistically significant. The difference between the average of the stern minus upper stern and μ0 is not big enough to be statistically significant. Emerald t n p Sd Bow vs. upper stern 4.064211 12 0.00186975 23983 Bow vs. lower stern 7.407945 12 0.0000134638 12087 -0.290946 12 0.776505 27270 Lower stern vs. upper stern The difference between the average of the lower stern minus bow and μ0 is big enough to be statistically significant. The difference between the average of the upper stern minus bow and μ0 is big enough to be statistically significant. The difference between the average of the lower stern minus upper stern and μ0 is not big enough to be statistically significant. 16 Notes 1 CJ Kruse, LM. DeSantis, SJ. Eaton, R Billings. Marine Transportation and the Environment - Trends and Issues. TR NEWS 313 JANUARY–FEBRUARY 2018. Available from: http://www.trb.org/main/blurbs/177286.aspx 2 Streibel T, Schnelle-Kreis J, Czech H, Harndorf H, Jakobi G, Jokiniemi J, Karg E, Lintelmann J, Matuschek G, Michalke B, Müller L. Aerosol emissions of a ship diesel engine operated with diesel fuel or heavy fuel oil. Environmental Science and Pollution Research. 2017 Apr 1;24(12):10976-91. 3 Haines G. 3 July 2017. “Revealed: Why going on a cruise should come with a health warning. The Telegraph. Available from: https://www.telegraph.co.uk/travel/cruises/news/why-going-on-a-cruise-should-come-with-a-health-warning/ 4 Cruise Line International Association. 2019 CRUISE TRENDS & INDUSTRY OUTLOOK. [Internet] Available from: https://cruising.org/news-and-research/-/media/CLIA/Research/CLIA%202019%20State%20of%20the%20Industry. pdf 5 Wang K, Wang S, Zhen L, Qu X. Cruise shipping review: operations planning and research opportunities. Maritime Business Review. 2016 Jun 30;1(2):133-48. 6 Crew Center – Ship Information Station. “How many crew members work in the cruise industry?”. Accessed June 8, 2018. Available from: http://crew-center.com/find-out-how-many-crew-members-work-cruise-industry. 7 Oeder S, Kanashova T, Sippula O, Sapcariu SC, Streibel T, Arteaga-Salas JM, et al. (2015) Particulate Matter from Both Heavy Fuel Oil and Diesel Fuel Shipping Emissions Show Strong Biological Effects on Human Lung Cells at Realistic and Comparable In Vitro Exposure Conditions. PLoS ONE 10(6): e0126536. doi:10.1371/journal. 8 Murena F, Mocerino L, Quaranta F, Toscano D. Impact on air quality of cruise ship emissions in Naples, Italy. Atmospheric Environment. 2018 May 29. 9 Maragkogianni A, Papaefthimiou S. Evaluating the social cost of cruise ships air emissions in major ports of Greece. Transportation Research Part D: Transport and Environment. 2015 May 31;36:10-7. 10 Schwarze PE, Totlandsdal AI, Lag M, Refsnes M, Holme JA, Ovrevik J. Inflammation-related effects of diesel engine exhaust particles: studies on lung cells in vitro. BioMed research international. 2013;(685142):1–13. pmid:23509760 11 C V Howard. Statement of Evidence – Particulate Emissions and Health. June 2009. Available here: http://www. nottinghamshire.gov.uk/media/110338/kc3-particulate-emissions-and-health-statement-of-evidence-to-ringaskiddyinquiry.pdf 12 Baldauf RW, Devlin RB, Gehr P, Giannelli R, Hassett-Sipple B, Jung H, Martini G, McDonald J, Sacks JD, Walker K. Ultrafine particle metrics and research considerations: review of the 2015 UFP workshop. International journal of environmental research and public health. 2016 Oct 28;13(11):1054. 13 Araujo JA, Barajas B, Kleinman M, Wang X, Bennett BJ, Gong KW, Navab M, Harkema J, Sioutas C, Lusis AJ, Nel AE. Ambient particulate pollutants in the ultrafine range promote early atherosclerosis and systemic oxidative stress. Circulation research. 2008 Mar 14;102(5):589. 14 Utell MJ, FramPton MW. Acute health effects of ambient air pollution: the ultrafine particle hypothesis. J Aerosol Med. 2000;13:355–359 15 Di Natale F, Carotenuto C. Particulate matter in marine diesel engines exhausts: Emissions and control strategies. Transportation Research Part D: Transport and Environment. 2015 Oct 1;40:166-91. 16 Nabu. Extreme air pollution levels found on deck of cruise ship Cruise ships’ exhaust gases harm human health. Available from: https://en.nabu.de/news/2017/21870.html 17 TSI Inc. P-TRAK ULTRAFINE PARTICLE COUNTER 8525. Available here: http://www.tsi.com/p-trak-ultrafine-particlecounter-8525/ 18 DieselNet – “Diesel Exhaust Particle Size” – available here: https://www.dieselnet.com/tech/dpm_size.php 19 Hall, D.E., et al., 2001. “Measurement of the number and mass weighted size distributions of exhaust particles emitted from European heavy duty engines”, CONCAWE Report 01/51, January 2001, https://www.concawe.eu/wp-content/ uploads/2017/01/2002-00235-01-e.pdf 20 Cheng YH, Liu CC, Lin YL. Levels of ultrafine particles in the Taipei Rapid Transit System. Transportation Research Part D: Transport and Environment. 2009 Oct 1;14(7):479-86. 17 21 Dong S, Yao M. Exposure assessment in Beijing, China: biological agents, ultrafine particles, and lead. Environmental monitoring and assessment. 2010 Nov 1;170(1-4):331-43. 22 Jian L, Zhu YP, Zhao Y. Monitoring fine and ultrafine particles in the atmosphere of a Southeast Chinese city. Journal of Environmental Monitoring. 2011;13(9):2623-9. 23 Suárez L, Mesías S, Iglesias V, Silva C, Cáceres DD, Ruiz-Rudolph P. Personal exposure to particulate matter in commuters using different transport modes (bus, bicycle, car and subway) in an assigned route in downtown Santiago, Chile. Environmental Science: Processes & Impacts. 2014;16(6):1309-17. 24 Peters J, Theunis J, Van Poppel M, Berghmans P. Monitoring PM10 and ultrafine particles in urban environments using mobile measurements. Aerosol and Air Quality Research. 2013 Apr 1;13(2):509-22. 25 US EPA. Managing Air Quality. Available here: https://www.epa.gov/air-quality-management-process/managing-airquality-human-health-environmental-and-economic 26 World Health Organization. News Release. March 2014. “7 Million premature deaths annually linked to air pollution”. Available here: https://www.who.int/mediacentre/news/releases/2014/air-pollution/en/ 18