Survey results
The survey was administered to fifty workers with a 100% completion rate. Survey results are provided in Online Resource 4. Seventy six percent (76%) of the surveyed workers reported starting their work shift between 7 am and 10 am and 64% ending it between 7 pm and 9 pm. People surveyed were between 15 and 65 years old (median of 36 years). Their work-days were between 7 and 14 hours long (median of 10 hours) and the time they had been working at that job ranged from 1 month to 50 years (median of 7 years).
Spatial and temporal variability of PM2.5
Table 1 shows descriptive statistics for PM2.5 for the sites and seasons included in this study. Data from the US reference site (M1) and the MX reference site (M2) were compared to the SYPOE (M3). The median PM2.5 concentrations measured at M3 were twice those of M1 in autumn and winter, but not different in summer (p < 0.05).
The concentrations of PM2.5 at M3 are greater than M2 by 30% and 10% in winter and autumn, respectively, but they are not statistically different. Concentrations at M2 are likely the result of contributions from local sources. In Tijuana, about 80% of the circulating private vehicles are imported from the United States and around 65% are models from the year 2000 and older (INECC and SEMARNAT 2011). Older vehicles tend to increase PM2.5 emission estimates (Zavala et al 2013) and the local residents are likely exposed to these increased emissions.
Table 1
Descriptive statistics of 24-h average PM2.5 concentrations in sites M1, M2, and M3 during different seasons
Season | Parameter (PM2.5 µg m− 3) | Site | |
M1 (Ref US) | M2 (Ref MX) | M3 (SYPOE) |
Summer 2018 | Mean | 10.8 | 14.7 | 13.6 |
Median | 9.8 | 14.4 | 12.6 |
Q1 | 8.0 | 10.2 | 11.6 |
Q3 | 12.6 | 17.3 | 14.2 |
Range | 5.4–17.7 | 6.0–24.0 | 8.9–27.2 |
N | 14 | 12 | 14 |
Autumn 2018 | Mean | 11.8 | 21.5 | 28.7 |
Median | 10.7 | 18.5* | 27.2* |
Q1 | 8.6 | 17.6 | 22.9 |
Q3 | 13.4 | 24.5 | 36.6 |
Range | 4.7–23.1 | 7.5–45.8 | 13.5–41.5 |
N | 14 | 14 | 14 |
Winter 2018 | Mean | 13.2 | 21.9 | 28.2 |
Median | 10.3 | 21.0* | 23.4* |
Q1 | 8.1 | 18.7 | 17.7 |
Q3 | 20.0 | 24.6 | 30.1 |
Range | 4.9–23.8 | 8.3–41.9 | 14.4–77.5 |
N | 14 | 13 | 14 |
Entire campaign | Mean | 11.9 | 19.5 | 23.5 |
Median | 10.3** | 18.2** | 19.5** |
Q1 | 8.2 | 13.0 | 14.1 |
Q3 | 15.0 | 24.2 | 28.9 |
Range | 4.7–23.8 | 6.0-45.8 | 8.9–77.5 |
N | 42 | 39 | 42 |
*data from sites M2 and M3 significantly higher than M1 (p < 0.05) |
**Kruskal-Wallis M3 = M2 > M1 |
The median PM2.5 daily concentration at M3 for the entire campaign was 19.4 µg m− 3 and is comparable to the value of 15 µg m− 3 determined by Galaviz et al. (2014) near the SYPOE in 2010. These data are comparable to those determined by a study at the border cities of Nogales, Mexico, and Nogales, Arizona (Smith et al. 2001). In that study, the median PM2.5 daily concentrations of two sites on the Mexican side were 17.92 and 11.67 µg m− 3 and for the two sites on the US side were 7.23 and 12.05 µg m− 3.
Figure 2 shows the time series of 24-h PM2.5 concentrations measured near SYPOE (M3). These concentrations exceeded the 24-h guideline (15 µg m− 3) established by the World Health Organization (WHO, 2021): 21% (3/14) in the summer, 86% (12/14) in the autumn, and 86% (12/14) in the winter. Similarly, the 24-h PM2.5 National Ambient Air Quality Standard (NAAQS) of 35 µg m− 3 (US EPA, 2016) was exceeded 12% (5/14) in the autumn and 7% (3/14) in the winter. In contrast, the 24-h standard (45 µg m− 3) established by the NOM-025-SSA1-2014 (Secretaría de Salud 2014) was exceeded only 5% (2/14) during the winter.
In summer 2018, the maximum value of PM2.5 was 27.2 µg m− 3 registered on July 5, 2018, following US Independence Day celebrations. In their study, Seidel and Birnbaum (2015) reported increases of 5 µg m− 3 of PM2.5 associated with Independence Day celebrations in several cities of the US on July 4 (night) and July 5 (morning). In another incident at the border, maxima of 41.4 µg m− 3 and 41.5 µg m− 3 were measured on November 26 and 27, 2018, respectively, which coincided with the arrival of a migrant caravan coming from Honduras. During this event, there were increased emissions from vehicles and helicopters, patrolling the area. Similarly, the maximum PM2.5 concentration measured in winter was 77.5 µg m− 3, registered on December 25, 2018. This maximum is likely associated with Christmas celebrations, which usually include fireworks. Lin et al. (2016) reported how the use of fireworks increase ambient PM2.5.
Jansen et al. (2015) showed that PM2.5 concentrations > 20 µg m− 3 exacerbates the symptoms of subjects with asthma and increase the risk associated with lung cancer, mortality, and cardiovascular disease by 4, 6, and 8%, respectively. Other findings suggest that asthmatic patients experience greater oxidation of plasmatic fluids due to PM2.5 exposure, and increased ROS generated by neutrophils (Sierra-Vargas 2009). In this study, 59% of the samples (25/42) in the entire sampling campaign were above 20 µg m− 3; therefore, the concentrations reported in this study could exacerbate the health problems of vulnerable populations.
Effect of meteorological conditions on PM2.5 concentrations
In this study, a predominantly southwest to northeast wind direction was registered in summer (Online Resource 4). Thermal inversions are only common in autumn and winter (Gobierno del Estado de Baja California 2011). The predominat wind direction for autumn was from the west (Gobierno del Estado de Baja California 2011), while the direction in winter was from the northeast.
A negative correlation (p < 0.05) was determined between temperature and PM2.5 concentrations for autumn (𝛒 = -0.61), and summer (𝛒 = -0.69), as shown in Online Resource 5; no correlation was found for the winter season. In addition, the mean wind speeds in summer were 11–37% higher than those of autumn and winter, respectively, further favoring the dispersion of pollutants.
A negative correlation (p < 0.05) between the daily concentration of PM2.5 and wind speed (𝛒 = -0.55) was also found in Autumn 2018. In a study conducted in 2010, Quintana et al. (2014) reported median PM2.5 concentrations near the SYPOE of 30.2 (February-March), 19.2 (April-June), and 4.2 µg m−3 (November). That study determined that pollution concentrations were higher during lower wind speeds or when the wind was blowing from the SYPOE toward their sampling site.
Temporal Variability of Black Carbon
Table 2 shows descriptive statistics per season for BC. The results show that autumn and winter 2018 registered higher concentrations than spring and summer 2018. Similarly, BC concentrations in autumn and winter 2017 were 1.2 and 2 times those of 2018, respectively. Lower precipitations registered in autumn/winter 2017 (0.04/0.02 mm) compared to 2018 (57.3/43.8 mm) may explain these differences.
BC concentrations on weekdays and weekends were also compared, and the results are shown in Table 2. The highest BC weekday concentrations were significantly higher (p < 0.05) than those of weekends in autumn/winter 2017. These results highlight the influence of activities related to work and school.
Table 2
Descriptive statistics of 1-h average BC concentrations in the entire campaign, during weekdays and weekends for each season studied
Period | BC (µg m− 3) | Autumn 2017 | Winter 2017 | Spring 2018 | Summer 2018 | Autumn 2018 | Winter 2018 |
Entire campaign | Average | 3.7 | 5.7 | 0.6 | 0.7 | 5.6 | 5.3 |
Median | 2.1* | 3.8* | 0.4 | 0.4 | 1.3* | 1.9* |
Q1 | 1.1 | 1.8 | 0.2 | 0.3 | 0.1 | 0.8 |
Q3 | 4.1 | 7.4 | 0.6 | 0.7 | 3.9 | 4.4 |
Maximum | 77.7 | 42.0 | 6.2 | 7.5 | 82.4 | 148.9 |
N | 409 | 362 | 336 | 365 | 337 | 361 |
Weekdays | Average | 2.2 | 6.2 | 0.4 | 0.7 | 5.9 | 6.1 |
Median | 2.3** | 3.9** | 0.3** | 0.4** | 1.5 | 1.8 |
Q1 | 1.1 | 2.0 | 0.2 | 0.3 | 0.4 | 0.8 |
Q3 | 4.3 | 7.7 | 0.5 | 0.8 | 4.8 | 4.4 |
Maximum | 77.7 | 42 | 4.0 | 7.5 | 82.4 | 148.9 |
N | 313 | 250 | 240 | 271 | 217 | 281 |
Weekends | Average | 1.8 | 4.7 | 0.5 | 0.5 | 6.2 | 2.8 |
Median | 1.6 | 3.5 | 0.6 | 0.4 | 1.4 | 1.9 |
Q1 | 1.1 | 1.3 | 0.3 | 0.3 | 0.3 | 0.8 |
Q3 | 3.1 | 6.2 | 0.8 | 0.6 | 4.2 | 3.7 |
Maximum | 15.5 | 39.2 | 6.2 | 1.6 | 67.3 | 16.3 |
N | 159 | 112 | 96 | 95 | 96 | 80 |
* Significantly higher than summer and spring 2018 p < 0.05.
** Significantly higher than weekends p < 0.05.
Online Resource 7 shows the 1-h average BC concentrations time series during the entire campaign. In autumn 2017 (November 24, 2017), the maximum value of 77.7 µg m− 3 coincided with ¨Black Friday¨, a popular sale day in the US, which led to increased traffic from Mexico to San Ysidro. Other common sources of BC are clandestine open fires on the Mexican side, which can raise local concentrations during autumn and winter seasons. The maximum BC concentrations in winter 2018 were 148.9 µg m− 3 (December 24, 2018) and 54.3 µg m− 3 (January 1, 2019), respectively. Both dates coincide with important holidays (i.e., Christmas and New Year) and are likely the combined results of fireworks, bonfires, and clandestine open fires. Takahama et al. (2014) in a study performed in US-Mexico, found peaks of high concentrations of BC suggestive of clandestine burning activities.
Effect of meteorological conditions on BC concentrations
Spearman correlations between 1-h average BC concentrations and meteorological conditions (temperature, RH, barometric pressure, precipitation, and wind speed) were calculated and are included in Online Resource 8. A moderate negative correlation (p < 0.05) was determined between BC concentrations and temperature in winter 2017 (𝛒 = -0.65) and 2018 (𝛒 = -0.51). Studies have a negative relationship between temperature and BC concentrations.
A moderate negative correlation (p < 0.05) was determined for wind speeds and BC concentration during autumn 2017 (𝛒 = -0.51), winter 2017 (𝛒 = -0.54), autumn 2018 (𝛒 = -0.56) and winter 2018 (𝛒 = -0.59). Table 3 shows significantly higher BC concentrations during low wind speed (< 5 m s− 1) periods compared to other wind speeds (> 5 m s− 1) for autumn 2017, summer 2018, autumn 2018, and winter 2018. Lower wind speeds decrease dispersion leading to increased BC concentrations. Saha and Despiau (2009) as well as Quintana et al. (2014) previously reported that 1-h average BC concentrations were higher during periods of lower wind speeds compared to higher wind speeds (> 0.5 m s− 1).
Table 3
Descriptive statistics of 1-h average BC concentrations during low wind and other wind speeds in the different seasons studied
BC conc. (µg m− 3) | Autumn 2017 | Winter 2017 | Spring 2018 | Summer 2018 | Autumn 2018 | Winter 2018 |
low wind speed | Average | 5.0 | 5.3 | 0.3 | 1.1 | 8.6 | 5.8 |
Median | 3.2* | 4.0 | 0.4 | 0.8* | 3.2* | 2.6* |
Q1 | 1.8 | 2.8 | 0.3 | 0.5 | 1.8 | 1.4 |
Q3 | 6.0 | 6.4 | 0.9 | 1.5 | 12.6 | 5.0 |
Maximum | 77.7 | 27.4 | 4.0 | 3.6 | 45.6 | 94.8 |
N | 141 | 81 | 33 | 47 | 38 | 60 |
other wind speed | Average | 3.0 | 5.8 | 0.6 | 0.6 | 5.5 | 5.2 |
Median | 1.5 | 3.6 | 0.4 | 0.4 | 1.4 | 1.7 |
Q1 | 0.9 | 1.6 | 0.2 | 0.3 | 0.3 | 0.7 |
Q3 | 3.2 | 7.8 | 0.6 | 0.6 | 4.0 | 4.0 |
Maximum | 21.1 | 42.0 | 6.2 | 7.5 | 82.4 | 148.9 |
N | 268 | 281 | 303 | 318 | 299 | 301 |
* Significantly higher than other wind speeds (> 0.5 m s− 1) p < 0.05.
The impact of wind direction on BC concentrations was also investigated. The pollution roses (Online Resource 9) show pollutant transport from the southwest (Tijuana) to northeast (San Ysidro) in spring and summer. This behavior was previously reported by Shores et al. (2013) and Bei et al. (2013). The pollutant transport in autumn (2017 and 2018) and winter (2017 and 2018), was from northeast to southwest. Similarly, contributions from the west indicate impacts from both the urban area and the SYPOE. The influence of the northeast to southwest direction indicates the transport of pollutants from the SYPOE towards Tijuana. Therefore, any strategy to reduce pollutants at the border crossing must consider the associated transport of pollutants.
Diurnal behavior
The diurnal behavior of BC concentration is presented in Fig. 3. Results show increased BC concentrations in the morning (4:00 am to 9:00 am), which reflect the start of anthropogenic activities, like morning commute. An additional increase was observed around 6:00 pm, coinciding with the evening commute. Similar daily diurnal patterns of high concentrations from 7:00 am to 9:00 am have been reported for different urban areas (Eidels-Dubovoi 2002; Park et al. 2002; Saha and Despiau 2009; Limon-Sanchez et al. 2011; Liu et al. 2019; Liñán Abanto et al. 2020). In the present study, these activities begin at earlier times in the morning, starting at 4:00 am because people working or studying in the US and living in Tijuana must also cross the border, adding to their daily commute (Rocha and Orraca 2018).
BC and PM2.5 correlation and ratio
BC and PM2.5 were found to be moderately positively correlated (p < 0.05) for summer (𝛒 =0.69), autumn (𝛒 =0.52), and winter (𝛒 =0.59) indicating that they are likely from similar sources (US EPA 2012). This result agrees with reports from other urban sites (Tiwari et al 2013; Peralta et al 2019). In addition, the ratio BC/PM2.5 has been shown to vary significantly (0.03–0.77) among Mobile Sources (US EPA 2012) and has been used to identify specific sources. Daily BC/PM2.5 ratios were calculated and presented in Online Resource 10. The mean BC/PM2.5 for winter was 0.1, which is associated with non-road gasoline sources. The highest BC/PM2.5 ratio for winter was 0.7, which is associated with non-road diesel sources. Ratio variations may be due to the speed, weight load, and driving conditions. Gaitan et al. (2016) reported BC/PM2.5 ratios ranging from 0.02–0.10 associated with gasoline vehicle emissions in Monterrey, Mexico. Meanwhile, Liu et al. (2016) reported BC/PM2.5 ratios ranging from 0.02 to 0.27 in Beijing, China, linked to traffic emissions. In this study, the emissions include those from diesel sources, unlike those other studies.
Oxidative Potential
The oxidative potential associated with PM2.5 measured during the entire sampling campaign is shown in Fig. 4. The median values of OPDTTm (and OPDTTv) in summer, autumn, and winter were 12.7 (0.2), 11.7 (0.3), and 18.5 (0.6) pmol min− 1 µg− 1 (nmol min− 1 m− 3), respectively. Verma et al. (2014) and Gao et al. (2020) also reported highest OPDTTm and OPDTTv in the colder seasons. Shirmohammadi et al. (2016) reported OPDTTm increases of 40–90% (OPDTTv of 20–40%) in colder months compared to the warm months in samples collected in Los Angeles, US.
Abrams et al. (2017) proposed OPDTTv as an indicator of air pollution toxicity, reporting a relation between OPDTTv and cardiorespiratory emergency department visits. Delfino et al. (2013) found increases of 8.7–9.9% in exhaled nitric oxide (a biomarker of airway inflammation) associated with an interquartile range of OPDTTv of 0.43 nmol min− 1 m− 3, in children from 9 to 18 years. In this study, OPDTT values > 0.43 nmol min− 1 m− 3 were found on 77% of the winter sampling days. Which suggests that the probability of respiratory disease complications is higher during this season.
On the other hand, the OPDTT for summer and autumn were not significantly different (p > 0.05). OPDTT in autumn was positively correlated to the waiting time of vehicles crossing the SYPOE (rho = 0.673, p < 0.05). M3 was located 200 m away from SYPOE and Cho et al (2005) reported higher OPDTT in areas near traffic; therefore, the influence of vehicles waiting and the association with population exposure should be considered.
The OPDTT values determined in this study are comparable to those found in Los Angeles, US (15–54 pmol min− 1 µg− 1), Fresno, US (27–61 pmol min− 1µg− 1), and Mexico City, Mexico (15–40 pmol min− 1 µg− 1) (Shirmohammadi et al. (2017, Charrier and Anastacio 2012, De Vizcaya-Ruiz et al. 2006).
Exposure and inhalation dose of BC
Table 4 shows the median 24-h average BC concentrations for all seasons included in this study. The highest daily average and median values were registered in 2017.
Table 4
Descriptive statistics of 24-h average BC concentrations during the different seasons studied
Parameter (µg m− 3) | Autumn 2017 | Winter 2017 | Spring 2018 | Summer 2018 | Autumn 2018 | Winter 2018 |
Average | 3.7 | 5.7 | 0.9 | 0.7 | 7.1 | 5.4 |
Median | 3.0 | 5.1 | 0.5 | 0.5 | 4.2 | 3.9 |
Q1 | 1.9 | 3.8 | 0.3 | 0.4 | 2.0 | 1.5 |
Q3 | 4.7 | 6.6 | 0.9 | 0.8 | 10.5 | 5.1 |
Maximum | 9.4 | 13.2 | 0.5 | 2.3 | 21.2 | 18.3 |
N | 17 | 15 | 14 | 15 | 11 | 15 |
Madrigano et al. (2010) reported that human exposure to ambient BC concentration of 1.77 µg m− 3 in a 24-h period was associated with inflammation and decreased endothelial function. Jansen et al. (2015) reported increases of 11.2% in fractional exhaled nitric oxide for asthmatic children for BC concentrations of 3 µg m− 3 in a 24 h period. Increases in fractional exhaled nitric oxide for asthmatic children also worsened the respiratory effects. The BC concentrations in this study, autumn and winter levels were > 3 µg m− 3 in a 24-h period, presenting similar concerns.
The estimated (95 percentile) pedestrian waiting time, average BC concentration, and inhalation dose of BC for each season are shown in Table 5. The highest doses were estimated for autumn 2018 and winter 2017. Notably, a pedestrian waiting time of 50 min led to a BC inhalation dose in winter 2017 that was eight times higher than in spring 2018.
Table 5
Inhalation dose (µg) of BC as a function of pedestrian and workers on a 24-hr period, by season
Season | Pedestrian waiting time | Average BC concentration (µg m− 3) | Pedestrian Inhalation dosea | Workers Inhalation doseb |
min (95 Percentile) | (µg) | (µg) |
Autumn 2017 | 58 | 5.0 | 2.5 | 12.0 |
Winter 2017 | 50 | 8.2 | 3.5 | 19.0 |
Spring 2018 | 50 | 0.9 | 0.4 | 3.9 |
Summer 2018 | 65 | 1.0 | 0.6 | 3.5 |
Autumn 2018 | 60 | 11.6 | 5.9 | 13.0 |
Winter 2018 | 60 | 5.0 | 2.6 | 10.6 |
aInhalation dose was estimated for a pedestrian waiting time in minutes on a travel day.
bInhalation dose was estimated for a day of work (average time of 10 hours).
Liu et al. (2019) reported an inhalation dose of 5.7 µg for walking in Macau, China (60 min) during traffic hours (7:30 am- 9:00 am). That dose is comparable the value of 5.9 for autumn 2018 in this study. Velasco et al., (2019) estimated an inhalation dose of BC of 6.0 µg for people walking (60 min) in Mexico City in winter. This study estimated 60-minute doses of 4.2 and 2.6 µg for the winters of 2017 and 2018, respectively, which are comparable to those reported in other urban environments.
Table 5 also shows the 24-hr inhalation dose for workers at the border, who work an average of 10 hours per day and 6 days per week. For the workers, a daily dose was also estimated based on their individual working schedule, and it ranged from 3.5 to 19 µg (Online Resource 11). These 24-hr values are 2 to 9 times higher than the estimated daily dose for pedestrians during their crossing. Further, a worker at this crossing received an annual inhalation dose of 28.2 mg of BC in 2018. A year-long exposure to BC was associated with decreased annual growth of working memory (Alvarez-Pedrerol et al. 2017).
Figure 5 shows the diurnal distribution of BC inhalation dose for Winter 2017, which registered the highest doses in this study. To reduce the daily inhalation dose, workers should avoid the periods between 7:00 to 9:00 am and 7:00 to 9:00 pm.