Advertisement

Particulate matter levels in a South American megacity: the metropolitan area of Lima-Callao, Peru

  • Jose Silva
  • Jhojan Rojas
  • Magdalena Norabuena
  • Carolina Molina
  • Richard A. Toro
  • Manuel A. Leiva-Guzmán
Article

Abstract

The temporal and spatial trends in the variability of PM10 and PM2.5 from 2010 to 2015 in the metropolitan area of Lima-Callao, Peru, are studied and interpreted in this work. The mean annual concentrations of PM10 and PM2.5 have ranges (averages) of 133–45 μg m−3 (84 μg m−3) and 35–16 μg m−3 (26 μg m−3) for the monitoring sites under study. In general, the highest annual concentrations are observed in the eastern part of the city, which is a result of the pattern of persistent local winds entering from the coast in a south-southwest direction. Seasonal fluctuations in the particulate matter (PM) concentrations are observed; these can be explained by subsidence thermal inversion. There is also a daytime pattern that corresponds to the peak traffic of a total of 9 million trips a day. The PM2.5 value is approximately 40% of the PM10 value. This proportion can be explained by PM10 re-suspension due to weather conditions. The long-term trends based on the Theil-Sen estimator reveal decreasing PM10 concentrations on the order of −4.3 and −5.3% year−1 at two stations. For the other stations, no significant trend is observed. The metropolitan area of Lima-Callao is ranked 12th and 16th in terms of PM10 and PM2.5, respectively, out of 39 megacities. The annual World Health Organization thresholds and national air quality standards are exceeded. A large fraction of the Lima population is exposed to PM concentrations that exceed protection thresholds. Hence, the development of pollution control and reduction measures is paramount.

Keywords

Particulate matter Air pollution assessment Long-term trend Metropolitan area of Lima-Callao, Peru Megacity 

Introduction

At present, there are a total of 3.3 million premature deaths globally from exposure to air pollution (WHO 2016a). This figure could double by 2050 if pollution continues to rise at the current rates (Lelieveld et al. 2015). The high levels of air pollution are mainly due to the excess emissions released into the atmosphere from vehicular, industrial, and residential emissions as well as power energy generation, among others. One of the air pollutants that negatively impacts public health is particulate matter (PM) (EPA-US 2009).

Particles with a diameter of less than 10 μm (PM10) have the greatest ability to access—and thus have the greatest deleterious effect on—airways. Within the PM10 fraction, smaller particles (less than 2.5 μm, PM2.5) are deposited in the alveoli, the deepest part of the respiratory system, and become trapped, possibly having severe health effects (Stanek et al. 2011; Tager 2012; Valavanidis et al. 2013). Epidemiological studies have revealed positive associations between PM contamination and serious harm to human health (EPA-US 2009). The harm caused by PM10 and PM2.5 to human health is manifested as cardiac and respiratory mortality, decreased lung capacity in children and adults with asthma, and an increased frequency of chronic obstructive pulmonary disease (EPA-US 2009).

To protect human health, organizations such as the World Health Organization (WHO) have established acceptable concentration thresholds, below which the harm to humans is sufficiently minimal (WHO 2006). However, particle pollution has health effects even at very low concentrations, and it has not been possible to identify any threshold below which no health damage has been observed. Several governments around the world have issued PM air quality standards (Vahlsing and Smith 2012), such as those established by the United States Environmental Protection Agency (EPA-US 2013) and by the Ministry of the Environment of the Government of Peru (PE-MINAM 2001).

Some of the greatest impacts of PM contamination on people are evident in urban environments (Baklanov et al. 2016; Baldasano et al. 2003; Molina and Molina 2002, 2004). Fifty-two percent (52%) of the world’s population lives in cities (UN 2016), and 31 cities around the world that surpass 10 million inhabitants, which in total hold approximately 700 million inhabitants (UN 2016). Five of these so-called megacities are in Latin America, namely, São Paulo, Brazil (top 12, 21.297 million inhabitants); Mexico City (top 10, 22.157 million inhabitants); Buenos Aires, Argentina (top 22, 15.334 million inhabitants); Rio de Janeiro, Brazil (top 27, 12.981 million inhabitants); and the metropolitan area of Lima-Callao (MALC) in Peru, which now has 10.072 million inhabitants and is currently positioned at number 31 of the most populated cities in the world (UN 2016). These megacities of Latin America have high PM levels (Bell et al. 2006; Reich et al. 2008).

The ambient air pollution estimated annual health impact of urban air pollution from particulate matter in MALC caused the premature death of approximately 3900 Peruvians and accounted for the loss of approximately 66,000 disability-adjusted life years per year, attributable to mortality (44%) and chronic bronchitis (13%), restricted activity days (20%), and respiratory symptoms (16%) (Awe et al. 2015).

In the MALC, the PM10 and PM2.5 levels in certain areas in the city often exceed the established safety thresholds for environmental concentrations recommended by the WHO (2006) and the Ministry of the Environment (MINAM by its acronym in Spanish) (PE-MINAM 2001). The main source of PM is associated with the growing size of the automotive fleet (1,195,353 in 2010 to 1,674,145 in 2015) (MTC 2016) and the use of fossil fuels (Dawidowski et al. 2014). Some of the older public transport vehicles have been removed from Lima’s roads, but 50% of Lima’s busses and minivans are over 18 years old (Carbajal-Arroyo et al. 2007; Underhill et al. 2015). A gas reformulation program has also been implemented to reduce the sulfur content and to convert part of the automotive fleet to natural gas (Dawidowski et al. 2014). Other sources are the industrial activity (brickworks, cement production, smelting, the fishing industry, and power generation) (PE-MINAM 2014). Additionally, 12% of municipal waste is open burned. Another factor is the prevailing anticyclonic meteorological conditions throughout the year, which added to an abrupt topography and led to a permanent subsidence and thermal inversion layer above the city, thus providing a stable atmospheric gradient that reduces the dispersion of air pollutants.

In this work, we analyze PM10 and PM2.5 measurements taken in the period of 2010–2015 to characterize the spatial and temporal distribution of the concentrations of particles on different time scales. Air pollution trends for PM10 and PM2.5 are analyzed. The results are evaluated by comparing the mean annual and mean daily concentrations of PM10 and PM2.5 with the corresponding WHO and MINAM standards. The study aims to understand PM air pollution trends to assess measures implemented to improve air quality and to determine the level of exposure of the MALC population over the past 6 years.

Materials and methods

Area of study

Lima is located at 77° W and 12° S in the middle of the western side of the South American continent (Arellano Rojas 2013; IGN 2014). It is the capital of the Republic of Peru, of the region and of the province of Lima (Fig. 1). It forms an extensive and populous urban area known as the MALC. From the coastline, in a northeasterly direction, the steep landscape ranges from a large alluvial plain formed by the valleys of the Chillón, Rímac, and Lurín rivers to a landscape of hills and finally mountains in the extreme east, which are located at more than 1000 m above sea level (and are the first foothills of the western slopes of the central Andes). It has an area of 2819.3 km2. In the summer (December through March), the average temperature (relative humidity) oscillates between 24 and 26 °C (65 and 68%) in the mornings, whereas at night, it fluctuates between 18 and 20 °C (87 and 90%). In winter (June through September), the average daytime temperature (relative humidity) oscillates between 18 and 19 °C (85 and 87%), and during nights, it ranges between 18 and 20 °C (90 and 92%). The average annual precipitation is 10 mm (Arellano Rojas 2013; SENAMHI 2016). The MALC is under the influence of persistent atmospheric stability caused by the interaction of the South Pacific Anticyclone, the Von Humboldt Current, and the presence of the Andes mountain range, which runs parallel to the coast. The thermal inversion reaches average altitudes in summer from its base of approximately 500 m above sea level, whereas in winter, its base reaches 1500 m above sea level (Arellano Rojas 2013). A detailed description of meteorological variables is provided in Fig. S1 in the supplementary material.
Fig. 1

Map of the MALC and the station locations of the Lima air quality monitoring network (right: modified after Wikimedia. https://goo.gl/n8t4MX, accessed 2017-03-10; left: google earth v7.1.5.1557, 2016-03-02, DigitalGlobe 2016. http://www.earth.google.com, accessed 2017-03-10). For more detail about the air quality monitoring network, see Table S1 in the supplementary material

PM10 and PM2.5 pollutant measurements

Since 1999, the MALC has had an air quality monitoring program, which is operated by the General Directorate of Environmental Health (DIGESA by its acronym in Spanish) of the Ministry of Health (MINSA by its acronym in Spanish); however, it was not until 2010 that the city of Lima had a real-time automatic air quality monitoring network (AAQMN) (LR 2010). Station and analyzer operation is the responsibility of the National Service of Meteorology and Hydrology (SENAMHI by its acronym in Spanish) of MINAM and is supervised by the network manager. There are currently ten stations that measure atmospheric aerosols (SENAMHI 2016) (Fig. 1).

The sites in the network are selected to represent where the population lives, plays, and works. The station meets regulations for occupational health and safety under national regulations. The electrical power system should meet electrical codes. In general, the shelter temperature should be maintained between 20 and 30 °C. Operation of the station includes regularly scheduled station visits, instrument zero and span verifications, calibrations, detection of leaks, preventive maintenance, and documentation. The verifications are conducted on a monthly or weekly basis. The data are transmitted by telemetry to SENAMHI headquarter where the data are validated to correct for null entries, duplicates, and/or anomalies in the data. The database is a 1-h time resolution. More information about the PM instrumentation and GPS coordinates and station operations is provided in Table S1.

Data analysis

A descriptive statistical analysis was performed using MS-Excel © (Microsoft Corporation, Redmond, WA, USA) in the R programming language (R Development Core Team 2012) using the Openair package (Carslaw 2013) under the open-source software RStudio: Integrated Development Environment for R (RStudio 2016). Temporal trends were estimated using the Theil-Sen approach (Sen 1968; Theil 1992). The Theil-Sen test calculates the slopes among all pairs of points, and the median slopes are selected as the Theil-Sen estimate, which is used as the pollutant trend for the given period (Carslaw and Ropkins 2012).

Assessment of standards and thresholds

To assess the possible health impact, the WHO thresholds (GD-WHO) and PE-MINAM air quality standards are compared on an annual and daily basis. The GD-WHO values on an annual basis are 20 and 10 μg m−3 for PM10 and PM2.5, respectively, expressed as the annual mean (WHO 2006). On a daily basis, they correspond to values of 50 μg m−3 for PM10 and 25 μg m−3 for PM2.5 (WHO 2006), in both cases expressed as a 24-h moving average. The PE-MINAM standards for PM10 concentrations on an annual basis are 50 μg m−3, expressed as an arithmetic mean; on a daily basis, they correspond to a value of 150 μg m−3 as a 24-h arithmetic average (PE-MINAM 2001). To date, the PE-MINAM has not established an annual PM2.5 standard but has done so for the 24-h period (25 μg m−3 for the 24-h arithmetic mean) (PE-MINAM 2001). It should be noted that the WHO thresholds are stricter than those of PE-MINAM, which is because the former consider only health risks, whereas the other standards are established by including cost-benefit considerations.

Results and discussion

PM10 and PM2.5 annual concentration levels

Table 1 presents a summary of the annual concentrations, 98th percentile concentrations, and percentage of annual data availability for the PM10 and PM2.5 measurements at the MALC monitoring stations. For the purposes of this study, only when there is data availability greater than 60% are the data associated with a given station considered valid values and representative of the annual concentration.
Table 1

Annual PM10 and PM2.5 concentrations in μg m−3 and annual percentages of data availability by the stations under study (see Fig. 1)

Year

Variable

ATE

SBJ

CDM

STA

VMT

HCH

SJL

SMP

CBR

PPD

a) PM10

2010

Annual Conc. (μg m−3)

123

47a

42a

P98 Conc. (μg m−3)

189

68a

76a

Available data (%)

74

23a

32a

2011

Annual Conc. (μg m−3)

130

51

45

76a

P98 Conc. (μg m−3)

203

75

75

118a

Available data (%)

76

88

82

15a

2012

Annual Conc. (μg m−3)

113

49

45

80

113

P98 Conc. (μg m−3)

198

76

78

143

227

Available data (%)

78

73

82

80

76

2013

Annual Conc. (μg m−3)

109

55

48

98

134

P98 Conc. (μg m−3)

199

91

88

185

298

Available data (%)

94

89

86

62

88

2014

Annual Conc. (μg m−3)

112

51

44

73a

90a

101

87

49

80

107

P98 Conc. (μg m−3)

179

98

85

135a

200

163

145

79

137

175

Available data (%)

94

62

85

56a

59a

70

71

75

69

69

2015

Annual Conc. (μg m−3)

91

46a

42

68

153

91

82

47

77

115

P98 Conc. (μg m−3)

162

76a

75

114

270

145

143

80

124

188

Available data (%)

74

25a

70

75

32a

76

82

68

79

75

Avg

Annual Conc. (μg m−3)

113

52

45

82

133

96

85

48

79

111

P98 Conc. (μg m−3)

188

85

80

147

249

154

144

80

131

182

Available data (%)

82

78

81

72

65

73

77

72

74

72

b) PM2.5

2014

Annual Conc. (μg m−3)

44a

18a

16a

21a

23a

38

34

19

32

35

P98 Conc. (μg m−3)

68a

31a

36

30a

39a

67

59

33

50

55

Available data (%)

25a

25a

16a

8a

8a

71

66

74

69

69

2015

Annual Conc. (μg m−3)

35

18

16

27

26a

24

28

17

23

29

P98 Conc. (μg m−3)

56

30

27

47

44a

41

46

28

40

45

Available data (%)

65

73

64

78

32a

79

82

69

69

76

Avg

Annual Conc. (μg m−3)

35

18

16

27

31

31

18

28

32

P98 Conc. (μg m−3)

56

30

32

47

54

53

31

45

50

Available data (%)

65

73

64

78

75

74

72

69

73

(−) No available data

aConcentrations with low data availability. Only calculated for reference

The average PM10 concentrations for monitoring stations (see Table 1(a)) are highest for VMT (133 μg m−3), ATE (113 μg m−3), PPD (111 μg m−3), and HCH (96 μg m−3). They are followed by lower concentrations at the SJL (85 μg m−3), STA (82 μg m−3), and CRB (79 μg m−3) stations. The stations with the lowest annual PM10 concentrations are SBJ (52 μg m−3), SMP (48 μg m−3), and CDM (45 μg m−3). The 98th percentile of environmental concentrations of PM10 is between 1.6 and 2.1 times the average annual value at the stations under study and follows a pattern similar to the mean annual PM10 concentrations. It is noteworthy that the stations with the highest PM10 concentrations are located in the eastern part of the city, that is, the highest part, whereas those with a lower concentration are found in the western part, which is the lowest part of the city (see Fig. 1); this trend is due to the pattern of persistent local winds entering from the coast with a south-southwest direction, causing the pollution loads to be transferred to the eastern and northeastern areas, which are critical deposition areas. The annual PM10 concentrations for the following groups of stations are not significantly different at the 95% confidence level (p value > 0.05): [VMT, ATE, PDA, and HCH], [STA, SJL, and CRB], and [SBJ, CDM, and SMP].

For PM2.5 (see Table 1(b)), a similar pattern is generally observed, even though the average concentrations between the stations are more similar and are in the range of 35 to 16 μg m−3. The highest annual concentrations of PM2.5 are observed at ATE (35 μg m−3), PPD (32 μg m−3), HCH (31 μg m−3), SJL (31 μg m−3), CRB (28 μg m−3), STA (27 μg m−3), SBJ (18 μg m−3), SMP (18 μg m−3), and CDM (16 μg m−3). The 98th percentile value, in general, follows a pattern similar to the annual concentrations of PM2.5. The 98th percentile of the environmental concentration of PM2.5 is between 1.6 and 2.0 times the average annual value at the stations under study. Two groups of stations in terms of the magnitude of the concentration of PM2.5 are clearly observed. The first group, formed by the stations of ATE, PPD, HCH, SJL, CRB, and STA, exhibits the highest concentrations; the second group, formed by the SBJ, SMP, and CDM stations, has lower concentrations. Among these groups of stations, an analysis of variance (ANOVA) indicated that there are no significant differences at the 95% confidence level (p value > 0.05).

PM10 and PM2.5 monthly concentration level

Figure 2 shows the temporal variability of the monthly (see Fig. 2a), daily (see Fig. 2b) and hourly (see Fig. 2c) averages of the PM10 and PM2.5 concentrations obtained from hourly measurements available at the monitoring stations under study.
Fig. 2

Temporal variations in the PM2.5 and PM10 concentrations on a monthly, daily, and hourly basis for monitoring stations in the MALC (see Fig. 1)

The average monthly variability (see Fig. 2a) exhibits the highest PM10 concentrations at the VMT (168 μg m−3), PPD (138 μg m−3), ATE (136 μg m−3), HCH (117 μg m−3), SJL (116 μg m−3), CRB (108 μg m−3), STA (93 μg m−3), SBJ (57 μg m−3), SMP (56 μg m−3), and CDM (52 μg m−3) stations. In general, the maximum concentrations are observed in the months of February through April, whereas the minimum concentrations are observed between May and September and are in the range of 90 to 36 μg m−3. In addition, a greater amplitude of the variation of the monthly concentrations (expressed as the difference between the maximum and minimum monthly concentration) at the VMT station (98 μg m−3) is observed, followed by the SJL, CRB, ATE, and PPD (approximately 48 μg m−3) and HCH (34 μg m−3) stations, whereas the lowest annual variability is observed at the STA, SMP, CDM, and SBJ stations, with values between 24 and 10 μg m−3. The month-to-month variability in the PM10 concentration (Fig. 2a) exhibited significant differences at the 95% confidence level (p value < 0.05) between the maximum and minimum values.

In the case of the PM2.5 average monthly variability (Fig. 2 a), the maximum concentrations observed for PM2.5 per station from greatest to least rank as ATE (43 μg m−3), PPD (38 μg m−3), SJL (37 μg m−3), HCH (36 μg m−3), STA (33 μg m−3), SBJ (23 μg m−3), CDM (21 μg m−3), and SMP (20 μg m−3). The maximum concentrations are generally evident between May and September. The minimum concentrations of PM2.5 vary from season to season between 11 and 30 μg m−3, with the absolute minima generally observed between October and April. The amplitude of the variability (expressed as the difference between maximum and minimum monthly concentrations) is in the range of 7 to 14 μg m−3, according to the season. The stations with the highest variability are VMT (14 μg m−3), ATE (13 μg m−3), PPD (12 μg m−3), CRB (11 μg m−3), HCH (10 μg m−3), CDM (9 μg m−3), and SMP (7 μg m−3). Significant differences between the maximum and minimum values are observed at 95% confidence (p value < 0.05).

PM10 and PM2.5 daily and hourly concentration levels

In the case of the behavior of the daily averages within the weekly cycle (Fig. 2 b) for PM10, there is an amplitude of variation among the days of the week (expressed as the difference between days with maximum and minimum concentrations), with ranges between 8 and 24 μg m−3. The maximum amplitudes are observed at the ATE (24 μg m−3) and SJL (20 μg m−3) stations, and the lowest values are observed at the CRB (11 μg m−3), CDM (10 μg m−3), and SMP (8 μg m−3) stations. There is a significant decrease in concentrations on Sundays. According to a 95% confidence-level ANOVA, there is a significant decrease in concentrations on Sundays compared with Tuesday through Friday (p value < 0.05), when the concentrations are higher. This is due to the variability in vehicular emissions because, during weekends, vehicular traffic decreases, and the emission contributed by mobile sources to the PM is more than 68% (SENAMHI 2014). In the case of PM2.5, the amplitude of variation between the days of the week for the stations under study ranges between 2 and 6 μg m−3, which explains why there is no statistically significant difference between the average concentrations on different days of the week (p value > 0.05). This result occurs because PM2.5, due to its smaller size and mass, can remain for a longer time in the atmosphere (Konstantinos 2008). The hourly variations for PM10 and PM2.5 at the stations under study are shown in Fig. 2 c. It is observed that the maximum concentrations occur at a peak generally between 6:00 and 10:00 in the morning and a second peak between 18:00 and 23:00 h. The average hourly maximums are observed at the VMT (PM10: 154 μg m−3, PM2.5: 35 μg m−3), ATE (PM10: 152 μg m−3; PM2.5: 56 μg m−3), HCH (PM10: 119 μg m−3; PM2.5: 47 μg m−3), and PPD (PM10: 140 μg m−3; PM2.5: 39 μg m−3) stations, followed by the SBJ (PM10: 65 μg m−3; PM2.5: 24 μg m−3), CDM (PM10: 62 μg m−3; PM2.5: 22 μg m−3), and SMP (PM10: 62 μg m−3; PM2.5: 24 μg m−3) stations. These trends can be understood as follows: the MALC has a vehicular fleet of 2.2 million; at an average age of 18 years old, the residents of the MALC make a total of 10 million trips a day in Lima and Callao (SENAMHI 2014). This commuter traffic in the metropolis results in peak hours of traffic and therefore emissions, which would occur in the morning and evening hours.

Relationships between PM2.5 and PM10

The relationship between the concentrations of PM2.5 and PM10 at the same station is evaluated using Pearson’s correlation coefficient. The results are presented in Table 2 (bold font), where it can be observed that the correlation coefficients of (PM2.5)St-i vs. (PM10)St-i are in the range of 0.72 to 0.49. The greatest values of the (PM2.5)St-i vs. (PM10)St-i correlation coefficient are observed at the SMP (0.72) and ATE (0.70) stations, whereas the least are found at the CRB (0.54), VMT (0.57), and PPD (0.49) stations. In general, these results are consistent because the emission sources of PM10 and PM2.5 at the stations should be similar. In the cases in which the correlation coefficients are smaller (CRB, VMT, and PPD stations), the stations are located in zones that have an important component of industrial emissions, few green areas, and an intense vehicular activity; consequently, the sources of PM10 and PM2.5 emissions are different. For example, there is a cement plant at VMT that, as part of its production process, emits PM10, whereas the emissions of PM2.5 in the zone primarily originate from vehicles.
Table 2

Site and inter-site Pearson correlation coefficients of (PM10)St-i vs. (PM10)St-j (normal font Open image in new window ), (PM2.5)St-i vs. (PM2.5)St-j (italics font Open image in new window ) and (PM25)St-i vs. (PM10)St-j (bold font Open image in new window ) between paired monitoring stations

(PM2.5)St-i vs. (PM10)St-j (bold font)

ATE

SBJ

CDM

STA

VMT

HCH

SJL

SMP

CBR

PPD

(PM10)St-i vs. (PM10)St-j (normal font)

ATE

(PM2.5)St-i vs. (PM2.5)St-j (italics font)

0.70

0.50

0.42

0.69

0.32

0.76

0.60

0.43

0.53

0.37

SBJ

0.45

0.66

0.74

0.66

0.40

0.47

0.57

0.66

0.30

0.41

CDM

0.38

0.61

0.74

0.52

0.27

0.49

0.50

0.78

0.37

0.39

STA

0.64

0.53

0.54

0.62

0.34

0.71

0.65

0.50

0.48

0.46

VMT

0.36

0.43

0.40

0.53

0.57

0.43

0.48

0.28

0.57

0.58

HCH

0.72

0.35

0.33

0.45

0.48

0.69

0.75

0.52

0.65

0.53

SJL

0.67

0.50

0.45

0.63

0.59

0.67

0.65

0.56

0.65

0.56

SMP

0.53

0.62

0.70

0.58

0.47

0.50

0.58

0.72

0.49

0.39

CBR

0.51

0.33

0.41

0.41

0.57

0.65

0.62

0.51

0.54

0.5

PPD

0.53

0.37

0.37

0.49

0.58

0.58

0.61

0.45

0.66

0.49

The Pearson correlation coefficients for each fraction of particles at paired sites (PM10)St-i vs. (PM10)St-j and (PM2.5)St-i vs. (PM2.5)St-j are presented in Table 2 (in normal font and italic font, respectively). The values for the correlations cover the range of 0.33 to 0.72 for the coefficients of (PM10)St-i vs. (PM10)St-j and 0.37 to 0.78 for the coefficients of (PM2.5)St-i vs. (PM2.5)St-j (see Table 2). Correlations with Pearson coefficients greater than 0.6 can be considered strong correlations and demonstrate that the PM emissions have the same emissions pattern in the considered sites (Molina et al. 2017; Toro et al. 2014). According to the results shown in Table 2, the coefficients are categorized into three groups based on the degree of correlation. Thus, there are higher correlations among three groups of sites: the first consists of the ATE, STA, and HCH sites; the second the SMP, CDM, and SBJ stations; and the third the PPD, CRB, and SJL stations. In the first group, the stations are located relatively close to each other, and their environment has a high residential component. In the second group, the stations are located in the lower part of the city, and their environments have green areas and asphalted streets, with a strong component of vehicular emissions. Finally, the third group is located in Lima North, where the stations located in the upper parts of the city have similar patterns of PM10 and PM2.5 emissions; this area of the city is characterized by large industrial and commercial areas.

PM2.5/PM10 ratio

The annual average of the ratio for the stations under study is 0.4 ± 0.1 (± 1σ), with a range of 0.21 to 0.44. This indicates that in the MALC, the PM2.5 concentration represents between 44 and 21% of the total PM10 and on average 40%. In the literature, it has been estimated that ratios higher than 0.7 indicate significant vehicular emissions (Xu et al. 2017), which could be assumed in the MALC because of its large vehicle fleet (1.7 million); however, in the MALC, the ratio is lower than this threshold, so there must be additional PM10 emission sources in the city. In the north, south, and east areas of the MALC, there are a large number of unpaved roads and sidewalks, which, undoubtedly, in addition to the re-suspension of PM by vehicular traffic and the presence of marine aerosols, can account for this proportion.

The month-to-month, day-to-day, and hour-by-hour variations in the PM2.5/PM10 ratio are illustrated in Fig. 3. With respect to the month-to-month variation in the PM2.5/PM10 ratio, Fig. 3a shows that the highest values are recorded in the months of May to August, and they are significantly different at the 95% confidence level with respect to the minima, which is associated with the variability of the meteorological variables (wind speed, wind direction, temperature, and relative humidity) throughout the year. This observed pattern is explained in the next sections. The day-to-day variations, which are shown in Fig. 3b, indicate an increase in the PM2.5/PM10 ratio on the weekend (mainly on Sundays), which can be attributed to a decrease in the re-suspension of coarse particles and a decrease in emissions resulting from less vehicular traffic. These patterns could be explained by the most intense vehicular activity occurring on working days, which would support the high concentrations of coarse particles and the fine particles that remain in the atmosphere for a longer time.
Fig. 3

Temporal variation in the concentrations of the PM2.5/PM10 ratio at monthly, daily, and hourly scales

In the case of daytime variation (Fig. 3c), a pronounced variation in the PM2.5/PM10 ratio is observed at most of the study sites, with a peak in the morning between 02:00 and 05:00 h. Then, a second peak of lower intensity is observed between 22:00 and 24:00 h. The hourly variation is explained by the fact that during the first hours of the day (2:00 and 4:00 h), vehicle activity is minimal, leading to decreased coarse PM, as opposed to fine PM, which remains relatively constant given its smaller size. In the morning hours, vehicular emissions during rush hour are associated with significant emissions of PM2.5, which cause the ratio to rise. In the afternoon, the development of activities leads to increases in re-suspension emissions, which contribute to PM10 emissions, thus reducing the ratio. The nocturnal peak (19:00 and 22:00) can be explained by the deposition of PM10 (Bernardoni et al. 2011).

PM2.5/PM10 ratio and meteorological variables

Regarding the monthly variation, the relationship is influenced primarily by the meteorological conditions of the MALC. Figure 4 shows the effect of the meteorological variables: wind direction (wd), wind speed (ws), temperature (T), and relative humidity (RH) on the PM2.5/PM10 ratio for all sites under study by month.
Fig. 4

Effect of the meteorological variables (wind direction, wind speed, temperature, and relative humidity) on the PM2.5/PM10 ratio. a Bivariate polar plot of the wind speed (ms−1) and direction for the 1-h mean mass concentrations of the PM2.5/PM10 ratio. b The relationship between relative humidity and c temperature versus the PM2.5/PM10 ratio for all sites under study by month

A bivariate polar plot of the ws and wd and the PM2.5/PM10 ratio with all available data for the period under study is shown in Fig. 4a. The figure shows that, despite having higher ws values, such as those observed in the months of October to February, PM10 re-suspension and the environmental concentration of this fraction are favored; thus, the observed PM2.5/PM10 ratio should decrease in value (blue to green color in Fig. 4a). In contrast to lower ws values observed in the months of March to September, the PM10 re-suspension is unfavorable, and therefore, the PM2.5/PM10 ratio observed should increase in value (green to red color in Fig. 4a). These plots account for the possible transport phenomena. In this sense, the maximum PM2.5/PM10 ratio is observed when ws is lower, which is indicative, in general, of local emission sources for PM2.5 with respect to PM10. Furthermore, stagnant wind conditions allow air pollutants to accumulate, resulting in elevated and localized concentrations of air pollutants (DeGaetano and Doherty 2004).

The correlation among T, RH vs. PM2.5/PM10 ratio for all sites under study by month is shown in Fig. 4b, c. In general, an inversely proportional relationship between T and the PM2.5/PM10 ratio (see Fig. 4b) is observed; i.e., at a lower temperature, there is a tendency for an increase in the PM2.5/PM10 ratio. Conversely, a direct relationship between RH and the PM2.5/PM10 ratio (in Fig. 4c) is observed; i.e., an RH increase is coincident with an increase in the PM2.5/PM10 ratio. These behaviors of T and HR vs the PM2.5/PM10 ratio are related to where higher T and lower HR (in summer compared to autumn-winter) enhance the transport process of coarse particles due to strong winds that re-suspend and keep the coarse particles in the atmosphere generated by vehicular activity or dust lifted by the wind (Jelić and Bencetić 2010). On the one hand, the period of lower T and higher HR (autumn-winter) favors the formation of fine particles by gas-particle conversion processes, and on the other hand, this period favors the removal processes of the coarse particles due to wet deposition (Molina et al. 2017; Toro et al. 2015). These patterns of variability for T and HR are associated with a phenomenon of inversion by subsidence and marine thermal inversion because of the influence of the Cordillera de los Andes, the Humboldt cool oceanic current, and the South Pacific anticyclone. During the summer, smaller and less permanent marine thermal inversion occurs mostly below 500 m due to the combined effects of subsidence and oceanic cooling. This weak marine thermal inversion causes a density decrease in stratiform clouds, with the consequent increase of solar irradiation, which leads to a decrease in HR and an increase of T at the surface, resulting in turbulent processes that lead to re-suspension of coarse PM, and at the same time prevents the formation of secondary particulate material (Arellano Rojas 2013). In contrast, the stratiform cloudiness increases during the winter, as does the HR, and the occurrence of drizzle is more frequent; these conditions favor a significant decrease in PM10 concentrations due to wet deposition and an increase in PM2.5 concentrations due to the formation of secondary aerosols via the process of converted gas-particulate (more details are provided in Fig. S1).

Trend analysis

The temporal trends were calculated for the stations with the highest number of PM10 measurement years, i.e., the ATE, VMT, STA, SBJ, and CDM stations. The results show that at the 95% confidence level, there is a significant trend of decreasing PM10 concentrations at the ATE and STA stations. The trend indicates a decrease on the order of −4.3 (CI −5.3, −3.2)% year−1 (equivalent to −5.51 (CI −7.1, −4.0) μg m−3 year−1) for the ATE station and on the order of −5.3 (CI −8.1, −1.6)% year−1 (equivalent to −4.7 (CI −7.5, −1.3) μg m−3 year−1) for the STA station. On the other hand, at the other stations at which it was possible to calculate the Telson estimator (i.e., the VMT, SBJ, and CDM stations), the results indicate that the trend is not significant at the 95% confidence interval, which means that the confidence interval exhibits a slope equal to zero. The results obtained for each of the stations are as follows: VMT: 6.3 (CI −4.5, 17.6)% year−1, SBJ: −5.4 (CI −8.1, −1.6)% year−1, and CDM: −1.5 (CI −3.5, 0.9)% year−1.

These results indicate that control measures such as street paving, exclusive roads for mass public transport, incentives for the use of natural gas, afforestation programs, restrictions on the import of used vehicles, and required vehicular technical inspections have led to reductions in PM10 concentrations in high-elevation areas of the city, mainly residential sectors, where the ATE and STA stations are located. In the rest of the city, there was no evidence of improvement in terms of trends of reductions in the PM10 concentration. Here, it should be noted that in the MALC, there are approximately 2.2 million vehicles, almost 70% of the total number of vehicles in the nation, which explains why the greatest contribution of this pollutant is associated with the automobile fleet (SENAMHI 2014).

Air pollution assessment

Figure 5 shows a comparison of the situation of the MALC relative to other world cities. Figure 5 shows the PM10 and PM2.5 concentrations of different cities around the world according to the 2014 WHO database (WHO 2016b), the PM concentration in the MALC, and a statistical summary of concentrations (maximum, minimum, mean, and standard deviation); see Fig. 5a. A comparison with megacities around the world is also presented. Figure 5b shows the list of megacities listed in the WHO database and the gross domestic product of the respective country to which the megacity belongs (WB 2017).
Fig. 5

a Concentrations rankings of particulate matter (PM10 and PM2.5) in different cities around the world (WHO 2016b). b Comparison of annual PM10 and PM2.5 concentrations (bottom left) per capita GDP (bottom right) and the number of inhabitants of the city (top) for global megacities (WB 2017).

When comparing the MALC with the cities listed in the WHO database, it can be observed that it ranks in positions 202 and 397 among more than 1500 cities in terms of PM10 and PM2.5, respectively (see Fig. 5a). The PM10 levels observed are approximately 1.95 times the average concentration of all cities listed in the database. The PM10 concentrations are similar to those observed in Linfen, China (84 μg m−3); Changzhi, China (83 μg m−3); Muscat, Oman (82 μg m−3); Nalbari, India (82 μg m−3); Mianyang, China (82 μg m−3); Tripoli, Libya (81 μg m−3); Bhilai, India (81 μg m−3); Kolhapur, India (81 μg m−3); Jilin, China (81 μg m−3); Toluca, Mexico (80 μg m−3); Shanghai, China (79 μg m−3); and Fresno, USA (74 μg m−3). The PM2.5 concentration in the MALC corresponds to approximately 1.2 times the average concentration of all cities in the database. Its concentration is similar to those observed in Ostrow, Poland (28 μg m−3); Verona, Italy (28 μg m−3); Shantou, China (28 μg m−3); Napoli, Italy (27 μg m−3); Rio de Janeiro, Brazil (27 μg m−3); Budapest, Hungary (27 μg m−3); Medellin, Colombia (27 μg m−3); and Bogota, Colombia (27 μg m−3).

When comparing the MALC with global megacities, i.e., cities of more than 10 million inhabitants, we can see that Lima is located in positions 12 and 16 in terms of PM10 and PM2.5, respectively, out of 39 megacities (see Fig. 5b). It is important to note that the city in the 13th ranking is Shanghai, China, which has more than double the population of the MALC. The relationship between the degree of contamination and economic development (see Fig. 5b) shows that at lower gross domestic product (GDP) per capita, in general, a higher concentration of PM10 and PM2.5 is observed. This accounts for the possibility of investing in pollution reduction policies that countries with greater economic development have. The MALC currently has a GDP lower than USD 10,000, which corresponds to the GDP of the countries in which the megacities with the highest pollution are located.

According to the WHO thresholds (GD-WHO) and Peruvian regulation (PE-MINAM), the average annual concentration of PM10 in the MALC exceeds the limits of 20 and 50 μg m−3 for PM10 (see Table 2). Thus, the WHO recommendation is exceeded by more than a factor of 2 at all stations and by more than a factor of 5 at some of them (ATE, VMT, and PPD). The national standard is exceeded at a minimum of seven stations of the MALC by factors between 1.6 and 2.4; the exceptions are the SBJ, CDM, and SMP stations. For PM2.5, the WHO threshold is 10 μg m−3 on an annual basis, which is exceeded at all stations by factors of between 1.8 and 4; the station that most exceeds the threshold is ATE (4.0 times), with CDM (1.6 times), SBJ and SMP (1.8 times) exceeding by a lesser degree.

Table 3 lists the percentage of days for which the PE-MINAM and WHO thresholds for the PM10 and PM2.5 concentrations are exceeded. For this calculation, the 24-h arithmetic average is used.
Table 3

Exceedances (in % of days per year) of the 24-h Peruvian air quality standards (PE-MINAM) and World Health Organization guidelines (GD-WHO) at the monitoring sites under study

Year

ATE

SBJ

CDM

STA

VMT

HCH

SJL

SMP

CBR

PPD

PM10

GD-WHO

2010

99.6

a

a

2011

98.6

49.1

27.6

a

2012

98.9

45.1

30.2

91.8

90.9

2013

91.3

56.6

40.3

96.5

91.9

2014

99.4

43.6

23.7

a

a

97.6

95.3

40.8

89.3

98.8

2015

94.8

a

22.6

82.4

a

96.1

83.6

35.2

92.1

100

Avg

97.0

49.1

29.0

89.8

91.6

96.6

100.0

38.3

90.8

100

PE-MINAM

2010

17.2

a

a

2011

17.4

0.00

0.00

a

2012

12.7

0.00

0.00

1.37

23.9

2013

16.3

0.00

0.00

9.29

34.4

2014

11.6

0.00

0.00

a

a

6.3

1.16

0.00

0.79

7.91

2015

4.43

a

0.00

0.00

a

2.2

1.34

0.00

0.00

16.7

Avg

13.2

0.00

0.00

3.03

29.5

4.1

1.54

0.00

0.37

12.7

PM2.5

GD-WHO

2014

a

a

a

a

a

100

94.6

33.8

96.0

99.2

2015

98.3

31.1

17.6

77.5

a

62.3

84.7

20.7

59.9

94.2

Avg

98.3

31.1

17.6

77.5

a

79.9

89.3

27.7

77.8

96.6

PE-MINAM

2014

a

a

a

a

a

89.1

77.9

10.8

78.5

89.6

2015

99.6

11.6

4.00

62.0

a

41.9

60.3

7.17

47.6

84.5

Avg

99.6

11.6

4.00

62.0

a

65.3

74.4

9.23

63.1

86.8

(−) No available data

aConcentrations with low data availability

The results indicate that for PM10, the GD-WHO limits are systematically exceeded at all the stations of the monitoring network. In general, the percentage of days per year in which the 24-h average daily concentration of PM10 is greater than 50 μg m−3 is 50%. The stations with the highest average number of days for which the PM10 concentration exceeds the GD-WHO threshold are the ATE (97%), VMT (91.7%), PPD and SJL (100%), HCH (96.6%), CRB (90.8%), and STA (89.8%) stations. In the case of the national standard for PM10 (PE-MINAM 150 μg m−3), seven of the ten stations of the air quality network studied exceed the threshold; the exceptions are the SBJ, CDM, and SMP stations (0% exceeded days). The number of days for which the threshold is exceeded on average by station is, from highest to lowest, as follows: VMT (29.5%), ATE (13.2%), PPD (12.7%), HCH (4.1%), SJL (1.54%), and CRB (0.37%).

In the case of the PM2.5 concentration, GD-WHO and PE-MINAM both have a daily value of 25 μg m−3. Regarding PM10, the results indicate that the GD-WHO and PE-MINAM thresholds are systematically exceeded at all stations under study. The percentage of days for which the GD-WHO threshold is exceeded for each station, from highest to lowest, is as follows: ATE (98.3%), PPD (96.6%), SJL (89.3%), HCH (79.9%), CRB (77.8%), STA (77.5%), SBJ (31.1%), SMP (27.7%), and CDM (17.6%). In general, the number of days for which the GD-WHO threshold is exceeded corresponds to at least 50% of the days of the year at five stations. Similarly, in the case of the PE-MINAM threshold, it is observed that all stations exceed the daily standard. The number of days for which the threshold is exceeded at each station, from highest to lowest, is as follows: ATE (99.6%), PPD (86.8%), SJL (74.4%), HCH (65.3%), CRB (63.1%), STA (62.0%), SBJ (11.6%), SMP (9.23%), and CDM (4.00%). At the stations with the greatest number of days for which the threshold is exceeded, the standard is exceeded at least 30% of the days of the year.

Finally, it is important to note that a large part of the population of Lima is exposed to concentrations that exceed the protection thresholds. Hence, the development of control measures and reduction of pollution are essential to reduce pollution by PM. The government of Peru has developed a pollution control and prevention plan that aims to improve air quality in the MALC and considers measures related to reduction of vehicular emissions, exclusive roads, road improvements, automobile fleet renewal, industrial emissions reduction, improvement of fuels, incentive for the use of clean fuels, increased green areas, and green taxes, among other measures (CIAL 2010).

Conclusions

The urban atmosphere of the MALC is contaminated by PM10 and PM2.5. The annual average PM10 concentrations range from 133 to 45 μg m−3, with a mean of 82 μg m−3, for the period from 2010 to 2015. The highest annual PM10 concentrations are observed at monitoring stations located to the east of the city, namely, ATE (113 μg m−3), VMT (112 μg m−3), PPD (111 μg m−3), HCH (96 μg m−3), SJL (85 μg m−3), STA (82 μg m−3), and CRB (79 μg m−3). In contrast, the lowest annual concentrations for PM10 are observed at the stations located west of the city, namely, SBJ (52 μg m−3), SMP (48 μg m−3), and CDM (45 μg m−3). For PM2.5, the average annual concentrations range from 35 to 16 μg m−3, with a mean of 27 μg m−3, in the period under study. In terms of annual PM2.5 concentration, from highest to lowest, the stations are ranked as follows: ATE (35), PPD (32), HCH (31), SJL (31), CRB (28), STA (27), SBJ (18), SMP (18), and CDM (16). In general, this pattern of distribution of PM concentrations, i.e., greater in the eastern part of the city, is related to the characteristic wind pattern of the city, which has prevailing winds from the south and southwest. The variation in the ratio (PM2.5/PM10) ranges from 0.21 to 0.44, resulting in an average concentration of PM2.5 representing 40% of that of PM10. This result indicates the presence of significant emitter sources, mainly of coarse particles, which come from the extensive uncovered areas, where re-suspension occurs owing to winds and intense vehicular activity, in addition to activities of production, transportation, and commercialization of cement, brick, aggregates, debris, etc.

The highest monthly concentrations of PM10 are observed in the summer and early autumn (February–April). The minimum monthly concentrations are observed in late autumn and early winter (May to September). On the other hand, regarding the average monthly concentrations of PM2.5, the highest concentrations are recorded between late autumn and winter (May and September), and the minimum occurs between mid-spring and early autumn (October–April). This temporal pattern occurs because subsidence thermal inversion weakens in the middle of spring and early fall and because the humidity decreases, which is detrimental to the formation of secondary particulate matter and contrary to what occurs during the cold and humid period (May–September). Regarding the hourly variability of the daily concentrations of both particle sizes, it was possible to identify two periods of maximum concentration mainly influenced by vehicular activity in the MALC, one between 6:00 and 10:00 h and the second between 18:00 and 23:00 h. This pattern can be explained if one considers that the MALC has 2.2 million motor vehicles, which make a total of nine million trips a day; thus, the maximum vehicular emissions are generated during the peak hours of traffic in the city in the morning and evening.

From the long-term trend analysis based on the Theil-Sen estimator, there is a trend of decreasing PM10 concentrations at ATE (5.51 μg m−3 year−1) and STA (4.7 μg m−3 year−1). However, to estimate the long-term trends, continuity in the measurement of PM environmental concentrations is required. The beginning of the installation of the monitoring network has made it possible to perform these first evaluations, which will be improved by having in the future more and better information from the stations of the air quality monitoring network.

The MALC is one of the most polluted cities in the world. The MALC is ranked 202 out of 1524 cities and ranked 397 out of 1615 cities with the highest environmental concentrations listed in the WHO database for PM10 and PM2.5, respectively. The observed concentration for PM10 (84 μg m−3) in the MALC is approximately twice the average concentration of all cities listed in the WHO databases, whereas the average concentration of PM2.5 corresponds to 1.2 times the average concentrations of cities listed in the database.

According to WHO thresholds (GD-WHO) and Peruvian regulations (PE-MINAM), the annual and daily threshold concentrations for more PM fractions will be systematically exceeded in the MALC. For the WHO thresholds, the fractions of days exceeded per year for PM10 and PM2.5 were greater than 65%. The stations that achieved the greatest number of days of exceedance of the WHO thresholds were the ATE (287 days) and CDM (86 days) stations. For PM2.5, these stations were PPD (256 days) and CDM (41 days). On the other hand, the maximum rates of exceedances for both PM fractions with respect to PE-MINAM regulations were approximately 7% for PM10 and more than 70% for PM2.5, with VMT (88 days) being the station with the greatest number of days of exceedance of the national standard of air quality for PM10. For PM2.5, all the stations under study exceed the standard, reaching a maximum of days of exceedance of the standard at ATE (235 days) and a minimum number of days of exceedance at CDM (10 days). Although the national standards and WHO thresholds have been exceeded, a tendency has been observed for some stations to exhibit a decrease in the PM10 concentration (namely, the ATE and STA stations).

In consideration of the above, it is important to understand that the improvement in air quality is due to different mitigation measures that must be applied systematically over time. However, often, these measures are too late, and the problem of contamination could be intensified. At the global level, different governments have implemented mitigation and prevention measures for decontamination, such as renewal of the public transit system, updating of emission standards, enforcement of regulations to certify emissions from stationary sources, mobile and vehicular restrictions during critical episodes of pollution, afforestation and creation of green areas, paving programs, and street washing. In general, these measures are applied in the MALC. For example, exclusive roads have been developed for public transit, and a metro is currently under construction in the city. Likewise, reforestation programs and the promotion of natural gas-fueled vehicles have been developed in the city. Notwithstanding the above in the city and given the evaluation performed in the present work, it is necessary to strengthen and extend the application of policies and strategies for air quality management, including strategic planning, design of sustainable urban areas, periodic evaluation of past policies, and strengthening association with health impacts and many others. Likewise, adequate monitoring and evaluation of the measures applied should be sought, and the population should be made aware of the implications of air quality in their daily lives.

Notes

Acknowledgments

We acknowledge the financial support of the National Meteorology and Hydrology Service of Peru–SENAMHI-Perú, for project SNIP N° 199842 “Expansion and Improvement of the Monitoring Network for the Forecasting of Air Quality in the Metropolitan Area of Lima” and Program 096—PPR096-Air Quality Management. MALG acknowledges the support of the National Commission for Scientific and Technological Research CONICYT/FONDECYT 2016 grant no. 1160617. RAT acknowledges the partial support of the National Commission for Scientific and Technological Research CONICYT/FONDECYT INICIACION 2015 grant no. 11150931. The funders had no role in the study design, data collection and analysis, and decision to publish or the preparation of the manuscript.

Compliance with ethical standards

Competing interests

The authors declared that they have no competing interests.

Supplementary material

10661_2017_6327_MOESM1_ESM.docx (456 kb)
ESM 1 (DOCX 456 kb).

References

  1. Arellano Rojas, C. S. (2013). Meteorological conditions and pollution levels in the metropolitan region of Lima - Peru (in Portuguese). Master Thesis. Institute of Astronomy, Geophysics and Atmospheric Sciences, Universidade de São Paulo, São Paulo, Brazil. [WWW Document]. URL https://goo.gl/qw9BOv (accessed 3.8.17).
  2. Awe, Y., Nygard, J., Larssen, S., Lee, H., Dulal, H., Kanakia, R., (2015). Clean air and healthy lungs: enhancing the World Bank’s approach to air quality management (No. No3), Environment and natural resources global practice discussion paper. Washington, DC.Google Scholar
  3. Baklanov, A., Molina, L. T., & Gauss, M. (2016). Megacities, air quality and climate. Atmospheric Environment, 126, 235–249.  https://doi.org/10.1016/j.atmosenv.2015.11.059.CrossRefGoogle Scholar
  4. Baldasano, J., Valera, E., & Jimenez, P. (2003). Air quality data from large cities. Sci. Total Environ., 307, 141–165.  https://doi.org/10.1016/S0048-9697(02)00537-5.CrossRefGoogle Scholar
  5. Bell, M. L. L., Davis, D. L. L., Gouveia, N., Borja-Aburto, V. H. H., & Cifuentes, L. A. A. (2006). The avoidable health effects of air pollution in three Latin American cities: Santiago, São Paulo, and Mexico City. Environmental Research, 100, 431–440.  https://doi.org/10.1016/j.envres.2005.08.002.CrossRefGoogle Scholar
  6. Bernardoni, V., Vecchi, R., Valli, G., Piazzalunga, A., & Fermo, P. (2011). PM10 source apportionment in Milan (Italy) using time-resolved data. Sci. Total Environ., 409, 4788–4795.  https://doi.org/10.1016/j.scitotenv.2011.07.048.CrossRefGoogle Scholar
  7. Carbajal-Arroyo, L., Barraza-Villarreal, A., Durand-Pardo, R., Moreno-Macías, H., Espinoza-Laín, R., Chiarella-Ortigosa, P., & Romieu, I. (2007). Impact of traffic flow on the asthma prevalence among school children in lima, Peru. The Journal of Asthma, 44, 197–202.  https://doi.org/10.1080/02770900701209756.CrossRefGoogle Scholar
  8. Carslaw, D. C. (2013). The openair manual—open-source tools for analysing air pollution data. Manual for version 0.9–0, King’s College London. London, UK. [WWW Document]. URL https://goo.gl/iigCI9 (accessed 3.8.17).
  9. Carslaw, D. C., & Ropkins, K. (2012). Openair—an R package for air quality data analysis. Environ. Model. Softw., 27–28, 52–61.  https://doi.org/10.1016/j.envsoft.2011.09.008.CrossRefGoogle Scholar
  10. CIAL. (2010). Second Integral Plan of Atmospheric Sanitation for Lima-Callao 2011–2015 (in spanish). Management Committee Clean Air Initiative (IAL) Lima-Callao, Lima, Peru. [WWW Document]. URL https://goo.gl/O9eaB3.
  11. Dawidowski, L., Sánchez-Ccoyllo, O., Alarcón, N.. (2014). Estimation of vehicular emissions in Metropolitan Lima. Final report (in Spanish). Lima, Peru: Servicio Nacional de Meteorología e Hidrología del Perú (SENAMHI). Report within the framework of the project South American Emissions, Megacities and Climate.Google Scholar
  12. DeGaetano, A. T., & Doherty, O. M. (2004). Temporal, spatial and meteorological variations in hourly PM2.5 concentration extremes in new York City. Atmospheric Environment, 38, 1547–1558.  https://doi.org/10.1016/j.atmosenv.2003.12.020.CrossRefGoogle Scholar
  13. EPA-US. (2009). Integrated Science Assessment (ISA) for Particulate Matter (PM). U.S. Environmental Protection Agency, Washington, DC, USA. EPA/600/R-08/139F. [WWW Document]. URL https://goo.gl/PCnvbj (accessed 3.13.17).
  14. EPA-US, 2013. National Ambient Air Quality Standards for Particulate Matter; Final Rule. 78 Federal Register 3086 (Document No2012–30946), Office of Air Quality Planning and Standards, United States Environmental Protection Agency, Research Triangle Park, NC, USA. [WWW Document]. URL https://goo.gl/3KB3SX.
  15. IGN. (2014). National Geographic Institute, Republic of Peru [WWW Document]. URL http://www.ign.gob.pe (accessed 3.13.17).
  16. Jelić, D., Bencetić, Z. (2010). Air quality in Rijeka, Croatia 27.Google Scholar
  17. Konstantinos, M. (2008). Quantification and evaluation of dust resuspension pm10 emissions in two large urban centers in Greece Quantifica, 543–547.Google Scholar
  18. Lelieveld, J., Evans, J. S., Fnais, M., Giannadaki, D., & Pozzer, A. (2015). The contribution of outdoor air pollution sources to premature mortality on a global scale. Nature, 525, 367–371.CrossRefGoogle Scholar
  19. LR. (2010). Senamhi inaugurates First Air Quality Monitoring Station in Ate | News from Peru | [WWW Document]. TheRepublica.pe. URL https://goo.gl/5k3qc8 (accessed 3.14.17).
  20. Molina, L. T., & Molina, M. J. (2002). Air quality impacts: local and global concern. In L. T. Molina & M. J. Molina (Eds.), Air quality in the Mexico megacity: an integrated assessment (pp. 1–19). Dordrecht: Springer Netherlands.  https://doi.org/10.1007/978-94-010-0454-1_1.CrossRefGoogle Scholar
  21. Molina, M. J. J., & Molina, L. T. T. (2004). Megacities and atmospheric pollution. Journal of the Air & Waste Management Association (1995), 54, 644–680.CrossRefGoogle Scholar
  22. Molina, C., Toro A. R., Morales, S. R. G. E., Manzano, C., Leiva-Guzmán, M. A. (2017). Particulate matter in urban areas of south-central Chile exceeds air quality standards. Air Qual. Atmos. {&} Heal. 1–15. doi: https://doi.org/10.1007/s11869-017-0459-y.
  23. MTC. (2016). Estimated National Vehicle Park, according to Department: 2007–2016. Lima, Peru.Google Scholar
  24. PE-MINAM. (2001). Supreme Decree No. 074–2001-PCM Regulation of National Environmental AIr Quality Standards (in spanish), Ministry of the Environment (MINAM), Republic of Peru. [WWW Document]. doi:D.S No 074–2001 –PCM.Google Scholar
  25. PE-MINAM. (2014). National air quality report 2013–2014 (in spanish). Ministry of the Environment (MINAM), Republic of Peru. [WWW Document]. URL https://goo.gl/czqcNg (accessed 3.15.14).
  26. Reich, S., Robledo, F., Gomez, D., & Smichowski, P. (2008). Air pollution sources of PM10 in Buenos Aires City. Environmental Monitoring and Assessment, 155, 191.  https://doi.org/10.1007/s10661-008-0428-x.CrossRefGoogle Scholar
  27. RStudio. (2016). RStudio: Integrated development environment for R [Computer software].Google Scholar
  28. Sen, P. K. (1968). Estimates of the regression coefficient based on Kendall’s tau. Journal of the American Statistical Association, 63, 1379–1389.  https://doi.org/10.2307/2285891.CrossRefGoogle Scholar
  29. SENAMHI. (2014). Estimation of vehicular emissions in Metropolitan Lima - Final report (in spanish). National Service of Meteorology and Hydrology of Peru (SENAMHI), Ministry of the Environment (MINAM), Republic of Perú. [WWW Document]. Natl. Serv. Meteorol. Hydrol. Peru (SENAMHI), Minist. Environ. (MINAM), Repub. Perú.Google Scholar
  30. SENAMHI. (2016). Air quality assessment in Lima metropolitan area 2015 (in spanish). National Service of Meteorology and Hydrology of Peru (SENAMHI), Ministry of the Environment (MINAM), Republic of Peru. [WWW Document]. URL https://goo.gl/db2JgU (accessed 3.8.17).
  31. Stanek, L. W., Brown, J. S., Stanek, J., Gift, J., & Costa, D. L. (2011). Air pollution toxicology—a brief review of the role of the science in shaping the current understanding of air pollution health risks. Toxicological Sciences, 120, S8–S27.CrossRefGoogle Scholar
  32. Tager, I. (2012). Health Effects of Aerosols. In Aerosols Handbook (pp. 565–636). CRC Press. doi: https://doi.org/10.1201/b12668-24.
  33. Theil, H. (1992). A rank-invariant method of linear and polynomial regression analysis. In B. Raj & J. Koerts (Eds.), Henri Theil’s contributions to economics and econometrics: Econometric theory and methodology (pp. 345–381). Dordrecht: Springer Netherlands.  https://doi.org/10.1007/978-94-011-2546-8_20.CrossRefGoogle Scholar
  34. Toro, A. R., Morales, S. R. G. E., Canales, M., Gonzalez-Rojas, C., & Leiva, G. M. A. A. (2014). Inhaled and inspired particulates in metropolitan Santiago Chile exceed air quality standards. Building and Environment, 79, 115–123.  https://doi.org/10.1016/j.buildenv.2014.05.004.CrossRefGoogle Scholar
  35. Toro, A. R., Córdova, J. A., Canales, M., Morales, S. R. G. E., Mardones, P. P., Leiva, G. M. A., 2015. Trends and threshold exceedances analysis of airborne pollen concentrations in Metropolitan Santiago Chile. PLoS One 10. doi: https://doi.org/10.1371/journal.pone.0123077.
  36. UN. (2016). The World’s Cities in 2016 – Data Booklet (ST/ESA/ SER.A/392). United Nations, Department of Economic and Social A airs, Population Division, New York, NY, USA. [WWW Document]. URL https://goo.gl/pVdtwy (accessed 3.8.17).
  37. Underhill, L., Bose, S., Williams, D., Romero, K., Malpartida, G., Breysse, P., Klasen, E., Combe, J., Checkley, W., & Hansel, N. (2015). Association of Roadway Proximity with indoor air pollution in a Peri-Urban Community in lima, Peru. International Journal of Environmental Research and Public Health, 12, 13466–13481.  https://doi.org/10.3390/ijerph121013466.CrossRefGoogle Scholar
  38. Vahlsing, C., & Smith, K. R. (2012). Global review of national ambient air quality standards for PM(10) and SO(2) (24 h). Air Quality, Atmosphere and Health, 5, 393–399.  https://doi.org/10.1007/s11869-010-0131-2.CrossRefGoogle Scholar
  39. Valavanidis, A., Vlachogianni, T., Fiotakis, K., & Loridas, S. (2013). Pulmonary oxidative stress, inflammation and cancer: Respirable particulate matter, fibrous dusts and ozone as major causes of lung carcinogenesis through reactive oxygen species mechanisms. International Journal of Environmental Research and Public Health, 10, 3886–3907.  https://doi.org/10.3390/ijerph10093886.CrossRefGoogle Scholar
  40. WB. (2017). GDP Online Database. World Bank, Washington, DC, USA. [WWW Document]. URL https://goo.gl/Bgn2X (accessed 3.13.17).
  41. WHO. (2006). Air quality guidelines. Global update 2005. Particulate matter, ozone, nitrogen dioxide and sulfur dioxide. World Health Organization Copenhagen, Denmark. [WWW Document]. URL https://goo.gl/hz79bI (accessed 1.23.15).
  42. WHO. (2016a). Fact sheet: Ambient (outdoor) air quality and health. World Health Organization, Media Centre, Copenhagen, Denmark. [WWW Document]. URL https://goo.gl/iJomX (accessed 1.23.15).
  43. WHO. (2016b). Ambient (outdoor) air pollution database, by country and city (xlsx file). World Health Organization (WHO), Department of Public Health, Environmental and Social Determinants of Health. Geneva, Switzerland. [WWW Document]. URL https://goo.gl/nPsdIq (accessed 3.8.17).
  44. Xu, G., Jiao, L., Zhang, B., Zhao, S., Yuan, M., Gu, Y., Liu, J., & Tang, X. (2017). Spatial and temporal variability of the PM2.5/PM10 ratio in Wuhan, Central China. Aerosol and Air Quality Research, 17, 741–751.  https://doi.org/10.4209/aaqr.2016.09.0406.CrossRefGoogle Scholar

Copyright information

© Springer International Publishing AG 2017

Authors and Affiliations

  1. 1.National Meteorology and Hydrology ServiceLimaPeru
  2. 2.Center for Environmental Science and Department of Chemistry, Faculty of ScienceUniversity of ChileSantiagoChile

Personalised recommendations