Chapter 4. Results The results are presented in three parts. The statistical analysis of the door-to-door journeys on each of the four modes of transport is presented first, followed by the spatial variations observed within each transport mode and the results and analysis of the dosage calculations. Concentration spikes and calculation of geometric means The s and 10 s logging intervals managed to capture the numerous short-lived concentration spikes that occurred throughout the sampling period. These short-lived spikes resulted in distributions of exposure concentration that were skewed towards larger values. To illustrate the frequency of these spikes, time-series of the measurements taken on 20 May 2013 are shown in Figure 4-1. Although some of these sharp peaks may be classified as extreme outliers (defined as > standard deviations [SD] above the mean), they are a common occurrence and a real component of the total pollutant concentration, thus they have not been removed. To account for these values without giving them too much weight, the geometric mean (GM) is used as the main descriptive statistic for each trip. The GM is essentially the logarithmic mean of the data, obtained by applying a logarithmic transformation on the data, taking the arithmetic mean, then back-transforming the mean values by taking the antilog. Other studies have also used the GM instead of the arithmetic mean to better describe the data (e.g. Gulliver and Briggs, 2004, de Nazelle et al., 2012). 57 200 -3 PM2.5 (µg m ) -3 PM1 (µg m ) 200 150 100 50 50 -3 PN (10 # cm ) 150 300 100 200 100 50 800 600 -3 pPAHs (ngm ) 600 400 200 400 200 50 2.0 40 CO (ppm) -3 BC (µgm ) 100 -3 ASA (mm m ) -3 PM10 (µg m ) 400 150 30 20 10 1.5 1.0 0.5 0.0 16:30 17:00 17:30 18:00 16:30 Local Time 17:00 17:30 18:00 Local Time Figure 4-1: Post-processed data measured on 20 May 2013. Time-series shown include one day of measurements on all transport modes. High variability and presence of spikes are evident in all measured parameters. In the rest of this thesis, the GM represents the average pollutant concentration measured during one door-to-door trip for each mode of transport or within a particular section of a trip. The arithmetic mean was only used to average HR and VE. For MRT mode, data from both trips on each day of measurement were combined for a total of 23 sets for analysis. However, for calculation of dosage, only data from the first MRT mode trip of each day were used. In the following description of results, ‘mean’ refers to the arithmetic mean of the GMs for each trip 58 or section. The boxplots presented in this chapter are also based on the GM from each trip or section. 4.1 Commuter exposure on door-to-door trips Mean values and SDs of all measurements are summarised in Table 4-1. More detailed descriptive statistics including maximum and minimum values can be found in Appendix D.1. The ratios of transport mode to background site (BG) measurements in Table 4-2 provide a measure of the difference between concentrations in the different transport microenvironments and at the ambient level. Generally, all three PM size-fractions and PN concentrations are higher for the four transport modes than at the background site, with the exception of the MRT mode. This agrees with existing research on commuter exposure, which finds that measurements in the transport microenvironment are elevated compared to ambient readings (Kaur et al., 2005b; Gulliver and Briggs, 2007; de Nazelle et al., 2012). The Walk (MRT) mode exhibits the highest (lowest) mean pollutant concentrations, while CO values in the transport environment and at BG are similar. The results of the individual pollutant metrics are described in more detail in the following sections. Table 4-1: Mean (SD) of pollutant metrics from all trips for transport modes and measured at the background site. (N = 23 for Bus, Taxi, and MRT, N = 22 for Walk) PM1 (µg m-3) PM2.5 (µg m-3) PM10 (µg m-3) PN (# cm-3) Bus MRT 28 (6) 27 (4) 29 (6) 27 (4) 32 (6) 30 (4) 28,916 (6,768) 14,418 (2,625) Taxi 27 (7) 28 (7) 30 (7) 30,882 (13,799)a Walk 37 (9) 37 (9) 42 (9) 44,490 (6,142) BG 25 (8) 25 (8) 27 (8) 21,893 (5,369) ASA (mm2 m-3) pPAHs (ng m-3) BC (µg m-3) CO (ppm) Bus 133 (25) 79 (27) 4.416 (2.318) 0.8 (0.7) MRT 114 (16) 30 (6) 2.643 (1.345) 0.5 (0.7) Taxi 140 (36) 86 (48) 4.368 (3.428) 1.0 (0.9) Walk BG 137 (23) - 99 (17) - 7.590 (3.975) - 0.6 (0.8) 0.9 (1.1) Transport mode 59 Table 4-2: Transport mode to BG ratios from all trips for pollutant metrics measured in both environments. Transport mode PM1 PM2.5 PM10 PN CO Bus 1.15 1.16 1.20 1.32 0.88 MRT 1.07 1.08 1.10 0.66 0.53 Taxi Walk 1.10 1.47 1.10 1.49 1.09 1.54 1.41 2.03 1.09 0.69 Statistical tests were also performed on the GM data to identify if concentrations measured on each mode of transport were significantly different from each other and BG. The results of the Kruskal-Wallis tests show a statistically significant difference across all measured metrics at the 95% confidence level (pvalue < 0.05) (Table 4-3). Post-hoc multiple comparisons were carried out using the Mann-Whitney U Test with Bonferroni correction to further discern which specific modes contributed to the result. The results of these multiple-comparison tests are presented in the following sections. Table 4-3: Results from the Kruskal-Wallis test validating that concentrations measured on each mode of transport were significantly different from each other and the background site. H = test statistic, df = degrees of freedom. Metric H df p-value PM1 25.198 0.00005* PM2.5 25.4264 0.00004* PM10 29.6262 [...]... =0.418 20 13/05/17 20 13/05 /20 2 2 r =0.181 r =0.585 20 13/05 /22 2 r =0.541 750 500 25 0 0 20 13/05 /23 1000 2 r =1 20 13/05 /27 2 r =0 .23 9 20 13/05 /28 20 13/05/30 2 2 r =0.0306 r =0.3 52 2013/06/03 2 r =0.335 750 500 25 0 0 20 13/06/10 1000 2 r =0 20 13/06/11 2 r =1 20 13/06/ 12 2 r =0.8 92 750 500 25 0 0 0 20 0 400 600 0 20 0 400 600 0 20 0 400 600 2 -3 ASA (mm m ) Figure 4-13:... UFPs in the nuclei mode 84 20 13/04/10 1000 2 r =0.49 20 13/04/17 2 r =1 20 13/04/18 20 13/04 /24 2 2 r =0.776 r =1 20 13/04 /25 2 r =0 .26 750 500 25 0 0 20 13/04 /26 1000 750 2 r =0.03 42 2013/04/30 2 r =0.594 20 13/05/03 20 13/05/06 2 2 r =0.137 r =0 20 13/05/14 2 r =0.959 500 25 0 0 -3 pPAHs (ng m ) 20 13/05/15 1000 2 r =0.636 20 13/05/16 2 r =0.418 20 13/05/17 20 13/05 /20 ... Bus-stop sections plotted against each other for each day of sampling Linear regression lines and the r2 of the relationship are also plotted 85 20 13/04/10 1000 2 r =0.0378 20 13/04/17 2 r =0.0383 20 13/04/18 20 13/04 /24 2 2 r =0 .25 6 20 13/04 /25 2 r =0.459 r =0 .23 20 13/05/06 20 13/05/14 750 500 25 0 0 20 13/04 /26 1000 2 2013/04/30 2 r =0.53 r =0.191 20 13/05/15 750 20 13/05/03 2 2013/05/16... 20 13/05/16 2 r =0.00067 r =0.0596 2 r =0.145 500 25 0 -3 pPAHs (ng m ) 0 1000 2 2 r =0.503 r =0.101 20 13/05 /23 20 13/05/17 20 13/05 /20 2 2013/05 /27 2 r =0.00801 r =0.0686 20 13/05 /22 2 r =0 .26 7 750 500 25 0 0 1000 2 r =0.413 2 r = 2. 07e-07 20 13/05 /28 20 13/05/30 2 2 r =0.118 r =0.105 20 13/06/03 2 r =0.159 750 500 25 0 0 20 13/06/10 1000 2 r =0.0381 20 13/06/11 2 r... of exhaust emissions in the indoor spaces Table 4 -20 : Mean (SD) of measured pollutant metrics in different sections of MRT mode journeys Section PM1 (µg m-3) PM2.5 (µg m-3) PM10 (µg m-3) PN (# cm-3) In- vehicle Train Indoor 34 (3) 34 (3) 37 (4) 12, 789 (2, 577) Mall 27 (5) 27 (5) 32 (6) 20 ,835 (5 ,22 2) Platform 26 (4) 27 (4) 31 (5) 9,570 (1,930) 21 (4) 22 (4) 24 (4) 13,833 (2, 6 72) 28 (7) 28 (7) 30 (8) 33,034... 33,034 (9,315) Station Outdoor Sidewalk 2 -3 -3 -3 ASA (mm m ) pPAHs (ng m ) BC (µg m ) CO (ppm) Train Indoor 85 (21 ) 32 (6) 2. 529 (2. 248) 0.5 (0.7) Mall 167 (48) 43 (14) 3.355 (1.749) 0.7 (0.7) 97 (30) 24 (7) 2. 204 (0.955) 0.4 (0.7) 1 42 (28 ) 27 (7) 2. 473 (1 .20 4) 0.5 (0.8) 148 ( 42) 44 (16) 5.996 (3. 421 ) 0.5 (0.7) In- vehicle Platform Station Outdoor Sidewalk 90 Table 4 -21 : Mean PM1/PM2.5, PM2.5/PM10, PC/DC... pPAHs in the transport microenvironment Velasco et al (20 04) measured pPAHs and ASA concentrations in different outdoor, indoor, and street environments of Mexico City, Mexico, and found mean ASA and pPAHs concentrations ranging from 122 – 7 02 mm2 m-3 and 17 – 5 82 ng m-3, respectively in the outdoor environments The mean ASA concentrations found in the present study are similar to those found in Mexico... =0.1 72 2013/06/ 12 2 r =0.000864 750 500 25 0 0 0 20 0 400 6000 20 0 400 6000 20 0 400 600 2 -3 ASA (mm m ) Figure 4-14: pPAHs and ASA data collected at Sidewalk sections during Bus mode trips plotted against each other for each day of sampling Linear regression lines and the r2 of the relationship are also plotted 86 4 .2. 2 MRT The MRT mode trips were split into five sections as well (Figure 4-15) Since... on each mode of transport were 14.6 (Bus), 14.3 (MRT), 18 (Taxi), and 25 .8 (Walk) minutes, respectively This, combined with variations in pollutant concentrations measured in each section, most likely contributed to the trends in exposure described in Section 4.1 76 The following sections describe the spatial variation of pollutant concentrations observed on Bus, MRT, Taxi, and Walk mode journeys in. .. PM2.5 (top) and PN (bottom) concentrations during the Bus mode trip on 10 June 20 13 79 Mean PM1, PM2.5, PM10, and PN concentrations measured in the five sections of Bus mode trips were slightly higher than at the background site, with outdoor sections (Bus-stop and Sidewalk) exhibiting much higher mean concentrations (Figure 4- 12) Except for concentrations inside the Bus, CO concentrations in the other . al., 20 02) . Along with the plots in Figure 4-6, the clustering of points close to the origin in Figure 4-7 indicates the presence of a nuclei mode in the observations, suggesting the strong influence. combined with variations in pollutant concentrations measured in each section, most likely contributed to the trends in exposure described in Section 4.1. 0 1 2 3 4 BG Bus MRT Taxi Walk Transport. proportion of fine to coarse-particles (Table 4-4). The ratios are very high, demonstrating that fine particles are the dominant component of PM in the Singapore transport microenvironment.