Environmental noise inequity in the city of Barcelona

Transportation Research Part D 63 (2018) 309–319

Contents lists available at ScienceDirect

Transportation Research Part D journal homepage: www.elsevier.com/locate/trd

Environmental noise inequity in the city of Barcelona a,⁎

a

Raymond Lagonigro , Joan Carles Martori , Philippe Apparicio a b

b

T

Department of Economics and Business, University of Vic -UCC, Sagrada Família, 7, 08500 Vic Barcelona, Spain INRS Centre Urbanisation Culture Société, 385, rue Sherbrooke Est, Montréal H2X 1E3, Canada

A R T IC LE I N F O

ABS TRA CT

Keywords: Environmental inequity Noise Vulnerable populations Geographical information systems Spatial models

Environmental noise is a growing concern for urban planners and public health experts. Continuous noise exposure has implications for people’s physical and mental health. Urban planning strategies are also involved in the need for regular noise assessments within urban areas. The objective of this study is to evaluate the exposure to noise of vulnerable population groups in the city of Barcelona, and to determine whether they are affected by an environmental inequity regarding this nuisance. Assessment of noise levels was performed by two methods of analysis—real measures and simulation—in order to build the noise database at block level for the 10 districts of the city. The results obtained by various statistical tests and spatial regression analysis show that children and low-income individuals are not affected by environmental inequity. On the other hand, we found a positive relationship between noise levels and the other groups considered: namely, the unemployed and people over age 65.

1. Introduction Environmental Equity (EQ) has become an important concern for academics in many different disciplines (Sze et al., 2009; Walker, 2012; Schlosberg, 2013). The scope of EQ has evolved and expanded (Walker, 2012), with one of its main components being the “inequitable and disproportionately heavy exposure of poor, minority, and disenfranchised populations to toxic chemicals, contaminated air and water, unsafe workplaces, and other environmental hazards” (Landrigan et al., 2010). Urban areas tend to cluster individuals and families with lower levels of education, occupation, or income (Hornberg and Pauli, 2007). Therefore, socioeconomic characteristics have a direct impact on the unequal distribution of populations in cities and, thus, their unequal exposure to environmental hazards. Socioeconomic inequalities have increasingly been recognized as one of the key factors forming the basis of health inequalities (Evans and Kantrowitz, 2002; O’Neill et al., 2007). These health inequities are more marked in urban areas because of the clustering of deprived and poor populations in certain neighbourhoods (Borrell and Arias, 1995). Noise, defined as “unwanted sound” of different types and intensities, including noise from transport, industry, and neighbours, is perceived as a pollutant and as an environmental stressor that is a prominent feature of the urban environment (Stansfeld et al., 2000). Regardless of the related air pollution, exposure to noise should be considered an important environmental factor in itself that has a significant impact on health (Foraster et al., 2011; Tobías et al., 2015). Environmental noise pollution has been proved to affect human behaviour, well-being, productivity, and health (European Commission, 1996; Stansfeld and Matheson, 2003). The influence of environmental noise on public health is probably the most significant reason that environmental noise has emerged as a major issue in environmental legislation and policy in recent years (Berglund et al., 1999; World Health Organization, 2011). With the aim of defining local action plans on urban noise control, the European Commission issued the 2002/49/EC directive, also known as the Environmental Noise Directive (END), requiring major cities to gather real world data on noise exposure (European ⁎

Corresponding author. E-mail addresses: [email protected] (R. Lagonigro), [email protected] (J.C. Martori), [email protected] (P. Apparicio).

https://doi.org/10.1016/j.trd.2018.06.007

1361-9209/ © 2018 Elsevier Ltd. All rights reserved.

Transportation Research Part D 63 (2018) 309–319

R. Lagonigro et al.

Commission, 2002). A working group was also created to assess the production of data on noise exposure (European Commission Working Group, 2007). The END determines noise indicators and levels of exposure to environmental noise through those common indicators. The directive also requires competent authorities in each member state to provide estimates of the number of people living in dwellings that are exposed to different levels of noise. Environmental equity has been assessed in the cities of Barcelona and Madrid, Spain, in terms of exposure to air pollution (Moreno-Jiménez et al., 2016), access to green spaces and green gentrification (Anguelovski et al., 2017; Cole et al., 2017), playgrounds and community centers (Anguelovski, 2013), crime and security perception (Valera et al., 2018). Borrell et al. (2013) reported evidence of health inequalities in the city of Barcelona, with socioeconomic inequalities seen to have a direct impact on disease and mortality rates. A short-term relationship between socioeconomic factors and mortality caused by air pollution has also been found in Barcelona (Barceló et al., 2009; Borrell et al., 2010; Rodríguez-Fonseca et al., 2013). But in regard to the noise standards defined by the European Commission, no studies relating socioeconomic factors and urban noise have yet been conducted in Barcelona. The city of Barcelona experienced many demographic changes during the decades from 1990 to 2010 (Catalán et al., 2008; Bayona-i-Carrasco and Pujadas-i-Rúbies, 2014). Urban saturation and immigration and economic crises meant that the urban configuration of the city shifted towards a sprawl model, as observed in other Mediterranean cities. The mononuclear compact city and the accompanying continuous metropolis lost their previous and almost absolute dominance. Sub-centres with a significant historical background, located outside the traditional city centre, have taken over this urban expansion. Concerning segregation, the case of Barcelona is particularly relevant: the percentage of immigrants in the total population has increased significantly within a very short period of time (from 4.83% in 2001 to 17.22% in 2011). The percentage of immigrants from non-EU countries in the total population rose from 4% in 2001 to 12.92% in 2011. Immigrants are currently concentrated in two types of areas: in the historical centre, where housing is of poor quality, and in peripheral districts close to public transport and composed of relatively cheap housing built in the 1960s and 1970s. Martori and Apparicio (2011) demonstrated that rapid and strong population growth has resulted in significant changes in patterns of segregation and in the emergence of ethnic enclaves. This segregation has also changed the landscape of low-income groups. Most immigrants from non-EU countries only have access to a highly informal labour market associated with low levels of income (Canal-Domínguez and Rodríguez-Gutiérrez, 2008). Therefore, the aim of this study is, first, to build a mean noise level index for all the street grid blocks in the city of Barcelona and draw up a framework to assess urban planning policies related to noise reduction; second, to analyze the spatial distribution of noise pollution in the different neighbourhoods; and, third, to evaluate the exposure to noise nuisances of vulnerable population groups. The remainder of this paper is structured as follows. The second section reviews the theoretical framework on environmental inequities and noise pollution. The third section overviews the empirical data and the statistical methodology used in the study. The fourth section presents the results obtained, which are later discussed in the fifth section. Finally, the last section draws conclusions and suggests some future lines of research. 2. Theoretical framework The goal of environmental justice is to ensure that all people, regardless of race, origin, age, or income, are protected from disproportionate impacts of environmental hazards (Melnick, 2002). Environmental justice is a major concern in academic research and political decision-making, and has become a basis for urban planning (Forkenbrock and Schweitzer, 1999; Antweiler et al., 2001; Schlosberg, 2013). The term environmental justice connotes some remedial actions to correct injustice for specific group of people, while environmental equity implies an equal sharing of risk burdens without pursuing its reduction (Cutter, 1995). The impacts of environmental inequalities in health and urban policies are one of the main challenges for public health throughout Europe (Judge et al., 2006). The World Health Organization (2008) stated that people with lower levels of education, occupation, and/or income tend to die at a younger age, and to have a higher prevalence of most types of health problems. The impact of the socioeconomic characteristics of the living environment and of exposure to environmental pollution is recognized as a crucial factor in the production of health inequalities (Evans and Kantrowitz, 2002; O’Neill et al., 2007). According to the WHO and the UN-Habitat report, Hidden Cities, all urban environments have the ability to produce health inequities that are “systematic, socially produced (and therefore modifiable), and unfair” (World Health Organization. UN-Habitat, 2010). Noise is considered to be an environmental pollutant of major importance in urban environments (Stansfeld et al., 2000). Chronic noise exposure has implications for physical and mental health (Ising and Kruppa, 2004). Direct noise effects on health have been reviewed by many authors and different consequences have been pointed: hearing impairment, annoyance, sleep disturbances (awakenings or sleep-cycle shifts), hypertension, cardiovascular risks (blood pressure increase, hypertension, subcortical stress reactions, heart diseases), disturbances in mental health, impaired task performance, chronic stress (Stansfeld and Matheson, 2003; Ising and Kruppa, 2004; Goines et al., 2007; Haralabidis et al., 2008; Basner et al., 2014). Basner et al. (2014) propose a review of observational and experimental studies on various diseases caused by noise exposure, pointing to the importance of noise prevention and mitigation strategies for public health. In New York City, McAlexander et al. (2015) stated that street-level noise has the potential to cause auditory and non-auditory health effects. Tobías et al. (2015) also quantify the direct relationship between the mortality index and road traffic noise in the city of Madrid, Spain. Noise exposure consequences may be worse for particular subgroups such as children, older people, and lower socioeconomic groups (World Health Organization, 2011). Van Kamp and Davies (2013) reviewed studies on evidence of noise effects on health, suggesting that vulnerable groups, such as, for instance, children, older people and lower socioeconomic population, should be 310

Transportation Research Part D 63 (2018) 309–319

R. Lagonigro et al.

targeted more specifically to analyse noise impacts on health. Most studies on the relationship between noise inequities and socioeconomic characteristics of the population consider immigrants and low-income population groups as their targets (Brainard et al., 2004; Lam and Chan, 2006; Bocquier et al., 2012; Nega et al., 2013). Individuals with low income and immigrants tend to cluster in specific urban settlements due to house pricing or mobility capacities. Other socioeconomic characteristics considered when tackling health inequities are low education and unemployment (Mackenbach and Bakker, 2003; Borrell et al., 2013; Rodríguez-Fonseca et al., 2013). As well, children and elders are physiologically more vulnerable in terms of the effects of noise on their state of health (Bolte et al., 2010; Landrigan et al., 2010; van Kamp and Davies, 2013). Kihal-Talantikite et al. (2013) found a direct impact of noise on the spatial distribution of infant mortality in the city of Lyon, France, that was clearly related to socioeconomic characteristics of the population. In Germany, people of lower socioeconomic status are proved to suffer more than others from noise pollution (Hoffmann et al., 2003). In the city of Montreal, Canada, low-income and visible minorities live in city blocks with slightly higher noise levels (Carrier et al., 2016). Most of the studies on noise effects on population consider, mainly, road traffic noise, as it is the most important source of ambient noise exposure worldwide, but, obviously, some urban areas are more exposed to noise than others (Moudon, 2009). The case of noise pollution in the city of Barcelona shows a more complex pattern as a consequence of the large concentration of tourism in the old city centre, where many streets are for pedestrians only, although high levels of noise are nonetheless detected (Casellas et al., 2010). Some urban policies in the city of Barcelona tended to promote urban sustainability, and simply displaced environmental problems and injustices to regional scales (Martínez-Alier, 2003). Later on, the city of Barcelona, like some other Southern European cities, deployed urban regeneration programs to deal with inequities in the distribution of environmental risk in urban areas. Arbaci and Tapada-Berteli (2012) stated that some of those programs have in fact increased socio-spatial inequalities. The European Commission’s 2002/49/EC directive, within the scope of environmental and health protection, defines the patterns for evaluating environmental noise and designing strategic noise maps, as well as action plans to manage noise in European cities (European Commission, 2002). The Catalan Parliament’s law 16/2002 and its regulation D176/2009 apply the European directives and establish criteria for protection against noise pollution (Parlament de Catalunya, 2002). Following the European directive and the Catalan law, the Barcelona city council approved the development of an acoustic map of the city in order to assess and manage urban noise pollution (Ajuntament de Barcelona, 2012c).

3. Data and methods The city of Barcelona is located on the central coast of Catalonia, Spain. It has a surface area of 102.16 km2, with a population of 1,610,427 inhabitants. The city has a street network that is 1368 km long, overlaying an area with 11.3 km2 of streets, roads, and highways, and 9.8 km2 of sidewalks (Ajuntament de Barcelona. Departament d’Estadística, 2012). The municipality of Barcelona developed a plan (Ajuntament de Barcelona, 2012c) to assess noise pollution in the city following the European Commission’s 2002/49/EC directive (European Commission, 2002). One of the consequences of this plan is the creation of a noise map of the city incorporating the noise level per street section (Ajuntament de Barcelona, 2012b). The noise database built by the Barcelona city council includes three noise level periods for each of the 14,343 street sections in the city. The assessment of the noise level or sound pressure level in dB(A) (A-weighted decibels), as defined in international standard IEC 61672:2003 (International Electrotechnical Commission, 2003), was carried out by two methods of analysis, real measures and simulation, during the year 2012. A total of 2309 short-term readings and 109 long-term readings were performed. Long-term measures are a minimum of 24 h long, and establish the temporary evolution of exposure to noise over the whole day, giving the information for the different periods of the day. Short-term measures are a minimum of 15 min long, and assess the daily noise level in the streets, detecting the causing sources and giving a street typology. Those measures are used with some computation methods to determine the noise levels in the whole street city network (NMPB-Routes-96, 1996; European Commission, 2003; Ajuntament de Barcelona, 2012b) Three daily noise level periods were calculated: Ld (daytime noise between 7 am and 9 pm), Le (evening noise between 9 pm and 11 pm), and Ln (nighttime noise between 11 pm and 7 am). A fourth Lden index was also computed as recommended by the Directive 2002/49/EC (European Commission, 2002), as a weighted average of the daytime level, evening level plus 5 dB(A), and night-time level plus 10 dB(A). The equation for Lden index is as follows:

Lden = 10log

Ld Le + 5 Ln + 10 1 × Td × 10 10 + Te × 10 10 + Tn × 10 10 T

(

)

where Ld, Le and Ln are day, evening and night noise levels respectively; Td, Te and Tn are day, evening and night length periods respectively (i.e. 14 h for day period, 2 h for evening period, 8 h for night period); T is the 24 h of total period time. R software (R Core Team, 2015) was used for data cleaning, aggregation, and analysis, and for geospatial mapping. The choice of the scale of analysis is one of the key issues in the assessment of spatial environmental inequities (Walker, 2010). In order to better track indicators, the study variables should be analyzed at the finest resolution possible, using the smallest spatial units available for the study area. The street section noise levels were aggregated at the city block level (i.e. a group of buildings surrounded by street sections) so as to better compare them to the usual information on housing and socioeconomic population data (Ausejo et al., 2010). The city of Barcelona is divided into 5375 city blocks, of which 4507 contain households. The Barcelona urban administration department’s digital database was used to determine the relationship between city blocks and street sections. Noise levels were summarized at the city block level by calculating a weighted average, using the street section length divided by the total block perimeter as a weight. The equation for the city block average noise level is as follows: 311

Transportation Research Part D 63 (2018) 309–319

R. Lagonigro et al.

Ldenk =

S

∑ ⎜⎛ ∑ kiS i

⎝

i

ki

× Ldenki ⎞⎟ ⎠

where Ldenk is the day-evening-night weighted noise level of block k; Ski is the street length of street i in block k; Σi Ski is the sum of the street lengths of all the streets out bounding block k (note that this is the perimeter of block k); Ldenki is the day-evening-night weighted noise level of street i in block k. According to the theoretical framework and the available data, we consider four groups based on the socioeconomic characteristics of low income, unemployment, low education, and non-European Union immigration, and two age groups: children under 15 years old and elders over 64 years old. Two data sources, both at the census tract level, were used: the Spanish National Statistics Office’s Censo de Población (2011), and the immigration data from the continuous registry of population (2001–2011) known as the Padron Municipal. The Padron Municipal is a municipal (non-state) registry that includes both regular and irregular migrant populations. It is available from 1996 onwards, and is easily accessible for research purposes. Some transformations and interpolations were performed in order to have all data on the same time and space scales. As explained above, noise data collection in Barcelona was assessed in 2012 at street section level, and we then aggregated it at the city block level. Socioeconomic variables were extracted from the 2011 Spanish census, and are only available at the census tract level. Finally, population for each city block was obtained from the municipality’s urban data. Socioeconomic data were calculated for each city block based on the proportions obtained by dividing the block population by the census tract population, as explained in (Pham et al., 2012). The spatial pattern of the six independent variables are presented in Fig. 1. Considering the various daily noise levels—Ld (daytime noise between 7 am and 9 pm), Le (evening noise between 9 pm and 11 pm), Ln (nighttime noise between 11 pm and 7 am), and Lden (corrected weighted average)—some authors suggest using only one noise indicator (e.g. Kihal-Talantikite et al., 2013). Correlations between the noise levels during the different daily intervals were computed for all city blocks in order to determine the degree of linear association between them. Because of the high correlation between the three daily intervals and the corrected weighted average (over 0.96), any results obtained would be very similar; the corrected weighted noise average (Lden) was therefore used for the study. Local indicators of spatial association (Anselin, 1995) were also computed to compare spatial cluster patterns with almost identical results for the four different daily noise intervals. Five categories for the Lden noise level were considered, as indicated in the EC directive (European Commission, 2002): under 55 dB(A), from 55 dB(A) to 59 dB(A), from 60 dB(A) to 64 dB(A), from 65 dB(A) to 69 dB(A), and over 70 dB(A). Fig. 2 shows the spatial pattern of the Lden noise level for the 4507 inhabited city blocks in the city of Barcelona. The European Union (2013) defines what is to be considered high-noise, as noise levels above 55 dB(A) Lden and 50 dB(A) Ln (nighttime noise level). Other authors have proposed different noise level limits. Lam and Chung (2012) propose high and low noise levels over 63 dB(A) and under 55 dB(A) respectively, as for noise mean levels in the city of Hong Kong. Berglund et al. (1999) consider 55 dB(A) limit for the mean noise level and the night-time noise limit to be disturbing to sleep of 55 dB(A). Once the mean noise levels and population groups had been computed within the boundaries of the city blocks, various tests were performed to measure the statistical association between this annoyance and the six population groups studied. The statistical analyses were inspired by those widely used in environmental equity studies (Rey and Janikas, 2005; Bocquier et al., 2012; Pham et al., 2012). The standard approach in most empirical work is to start with a non-spatial linear regression model (OLS) and then to determine (e.g. by using the Moran’s I test) whether or not the model needs to be extended with spatial effects. We estimated various models and applied OLS regressions to explain the global relationship between the dependent variable Lden and the different population groups studied. Models were defined by considering individuals in each group as a proportion of the total number of individuals per block. The evaluation of OLS models was assessed by multicollinearity through Variance Inflation Factor (VIF) values (O’Brien, 2007). The relative quality of the models was also assessed with the Corrected Akaike Information Criterion (AIC) (Akaike, 1973). To begin with the spatial exploration, the spatial autocorrelation was evaluated using the Moran’s I index in order to determine whether the model needed to be extended with spatial effects. The significant clusters of noise levels are presented in Fig. 3. The clusters were determined by estimating local indicators of spatial autocorrelation, which determines spatially defined patterns of spatial autocorrelation (Anselin, 1995). The results of the Lagrange Multiplier tests (LM-Lag and LM-Error) and their robust versions (RLM-Lag and RLM-Error) may be used to decide what kind of spatial dependence is the most appropriate to control for the presence of spatial dependence in the OLS residuals (Anselin, 2009). 4. Results Fig. 1 above presents the noise level map of the city of Barcelona, along with the city districts. In general, the city of Barcelona exhibits a situation of high mean noise levels. Some 48% of city blocks have a mean noise level over 65 dB(A), and only 5% of city blocks have mean noise levels under 55 dB(A). The city blocks with the highest noise levels can be found in the Eixample district in the city centre, which has the most regular-shaped blocks and high flows of street traffic. Part of the Sarrià-Sant Gervasi district, where the street network is also regular, with some sloping elevations, also has many blocks with high-noise levels (over 70 dB(A)). The results are divided into three parts. First, we examine some statistical evidence of correlation between mean noise and the selected population groups. Next, we present empirical evidence of the presence of spatial autocorrelation in the OLS residuals. Finally, we introduce the spatial effects into the model and base our decision on the robust versions of the Lagrange Multiplier tests. 312

Transportation Research Part D 63 (2018) 309–319

R. Lagonigro et al.

Fig. 1. Spatial distribution for the six explanatory variables (Unemployment, Low-education, Low-income, non-EU immigration, children under 15 years old and people 65 years old and over).

313

Transportation Research Part D 63 (2018) 309–319

R. Lagonigro et al.

Fig. 2. Noise level map of the city of Barcelona (Lden).

To introduce these spatial effects, we use a row standardized contiguity matrix of first-order rook weights. Other forms of weights matrix were tested, yielding similar qualitative and quantitative results. Table 1 shows the univariate statistics for the variables used in the model: Lden (the corrected weighted noise average) and the six population groups considered as percentages of the total block population. It should be noted that the median of mean noise level is 65.1, so that more than half of the city blocks in the city of Barcelona are characterized by noise levels exceeding 65 dB(A). A T-test was performed to compare the means of the noise levels weighted by the individuals in the groups studied. Negative significant differences were found for the low-income (65.30 dB(A)) versus the no low-income population (66.29 dB(A)), and the loweducation population (65.64 dB(A) versus 66.12 dB(A)). Instead, positive but weaker differences were found for the unemployed (66.43 dB(A) versus 66.07 dB(A)). No significance (0.21 dB(A) or less) was found for non-EU immigrants (65.94 dB(A) versus 66.13 dB(A)), children under 15 (65.91 dB(A) versus 66.13 dB(A)), and people over 64 years old (66.20 dB(A) versus 66.07 dB(A)). Spearman correlation coefficients were calculated to verify for dependence between noise level and the proportions of the groups studied. We found significant negative correlations for the low-income group (−0.25, p < 0.00001), and children under 15 (−0.19, p < 0.00001). Conversely, a positive correlation can be observed for the unemployed group (0.2, p < 0.00001), and a weaker positive correlation for people over 64 years old (0.11, p < 0.00001). As shown in Table 1, based on Moran’s I, the model needed to be extended with spatial effects with all variables presenting spatial autocorrelation. We computed the robust versions of the Lagrange Multiplier tests (LM) for the spatially lagged dependent variable (RLM-Lag) and for error dependence (RLM-Error) (Table 2). The results obtained were significant for both the spatial lag and spatial error models, but higher values were obtained for both LM-Lag and RLM-Lag. This suggests that the spatial lag model was preferable to the spatial error model. These results were obtained using the R library, spdep (Bivand et al., 2011, 2017). The results of the OLS and spatial lag models are also reported in Table 2. First, the lower AIC and the higher log-likelihood values indicate that the spatial lag model outperforms the OLS model. Next, only the coefficients of three population groups are significant (p < 0.05) in the spatial lag model while all five coefficients are significant for the OLS model. The Moran’s I spatial autocorrelation of the residuals for the spatial lag model is reduced to 0.0224 (p = 0.009). The higher the proportions of children (B = −0.051, p = 0.001) and the low-income population (B = −0.026, p = 0.000) are in a city block, the lower the mean noise level is. On the other hand, the mean noise level increases for unemployed persons (B = 0.066, p = 0.000). In Table 3, the proportion of individuals living in city blocks classified by mean noise level is presented for the total population

314

Transportation Research Part D 63 (2018) 309–319

R. Lagonigro et al.

Fig. 3. Local indicators of spatial autocorrelation for the weighted average of noise level (Lden). Table 1 Univariate statistics for the variables used in the model.

Lden Unemployed population (%) Low-education population (%) Low-income population (%) non-EU immigrants (%) Children under 15 years old (%) People 65 years old and over (%)

Mean

SD

CV

Min

Q1

Median

Q3

Max

Moran’s Ia

64.5 5.0 1.8 7.9 17.7 12.6 20.6

5.7 5.4 2.4 7.2 10.7 3.0 4.9

0.1 1.1 1.3 0.9 0.6 0.2 0.2

22.6 0.0 0.0 0.0 1.3 6.0 3.8

61.9 1.4 0.3 2.8 10.6 10.6 17.9

65.1 3.3 0.9 5.9 14.8 12.0 20.9

68.1 6.7 2.3 11.1 20.6 14.3 23.3

76.0 49.0 24.7 42.7 62.5 29.0 50.9

0.787 0.367 0.435 0.698 0.857 0.828 0.704

Min: minimum; Max: maximum; Q1: first quartile; Q3: third quartile; SD: standard deviation; CV: coefficient of variation. a All values are significant at p = 0.0000.

and the six groups. In line with the previous analyses, the proportion of unemployed individuals and of people over 64 years old living in city blocks with higher noise levels is higher than the proportion of the total population in those city blocks. On the other hand, the proportion of children living in city blocks with lower mean noise is higher. Similarly, the proportion of low-income individuals is also higher in city blocks with lower noise levels. 5. Discussion The city of Barcelona presents a general situation of noise nuisances, with 88% of city blocks experiencing high-noise levels as defined by the European Union (2013). The available data on long-term average exposure show that 65% of Europeans living in major urban areas are exposed to high-noise levels (European Union, 2013). In contrast, in the city of Barcelona, 94.66% (considering both Lden and Ln limits) of the population lives in city blocks with high-noise levels. Berglund et al. (1999) consider the Lden limit of 315

Transportation Research Part D 63 (2018) 309–319

R. Lagonigro et al.

Table 2 OLS and Spatial lag model of the mean noise level. OLS

Spatial lag model

Coefficient

T value

p value

Intercept Unemployed population (%) Low-education population (%) Low-income population (%) non-EU immigrants (%) Children under 15 years old (%) People 65 years old and over (%) Wy (spatial lag coefficient)

68.719 0.280 −0.334 −0.162 0.039 −0.462 0.075

80.25 13.31 −7.05 −14.27 3.97 −14.68 3.57

0.000 0.000 0.000 0.000 0.000 0.000 0.000

Moran’s I RLM-Lag RLM-Err BP test Adjusted R2 Log-Lik AIC

0.716 314.39 22.730 370.74 0.169 −13839.79 27591.72

0.000 0.000 0.000 0.000

Coefficient 7.670 0.066 −0.042 −0.026 0.009 −0.051 0.015 0.884 Moran’s I Pseudo R2 Sq.cor. Log-Lik AIC

Z value

p value

12.86 6.47 −1.86 −4.63 1.84 −3.31 1.47 142.98

0.000 0.000 0.063 0.000 0.066 0.001 0.141 0.000

0.022 0.752 0.820 −11076.96 22171.92

0.009

Table 3 City block noise levels and proportions of individuals (difference from the total population in parentheses). Noise category

< 55.0 dB(A)

55.0–59.9 dB(A)

60.0–64.9 dB(A)

65.0–69.9 dB(A)

≥70.0 dB(A)

Total population (reference)

(Ref.)

Unemployed population (%)

0.58% (−0.56) 0.98% (−0.16) 1.60% (0.46) 1.16% (0.02) 1.51% (0.37) 1.04% (−0.10)

6.11% (Ref.) 5.39% (−0.72) 6.19% (0.08) 8.32% (2.21) 5.65% (−0.46) 6.42% (0.31) 6.06% (−0.05)

30.26% (Ref.) 27.64% (−2.62) 34.26% (4.00) 37.98% (7.72) 32.75% (2.49) 31.05% (0.79) 29.49% (−0.77)

43.14% (Ref.) 46.80% (3.66) 44.20% (1.06) 37.90% (−5.24) 42.21% (−0.93) 42.56% (−0.58) 43.33% (0.19)

19.35% (Ref.) 19.59% (0.24) 14.36% (−4.99) 14.21% (−5.14) 18.23% (−1.12) 18.46% (−0.89) 20.08% (0.73)

Low-education population (%) Low-income population (%) non-EU immigrants (%) Children under 15 years old (%) People 65 years old and over (%)

55 dB(A) and Ln limit of 55 dB(A); with this limits, the proportion of population exposed decreases to 72.66%. If we consider road traffic noise as the main noise source (Moudon, 2009), these results are in line with those obtained by Moreno-Jiménez et al. (2016) that empirically demonstrated overexposure to high concentrations of pollutants mainly caused by traffic in two major Spanish cities, Madrid and Barcelona. High mean noise level blocks in Barcelona are highly clustered (see Moran’s I in Table 1) mainly in the city centre. The highest noise levels (over 70 dB(A)) are observed in the Eixample district, where city blocks are regularly defined in grid form and thus have higher levels of street traffic and higher car speeds because of the straight streets. Such results are clearly contrary to the aims of the district’s design. The Eixample district was planned by Ildefons Cerdà during the mid-eighteenth century in order to rapidly enlarge the city. Cerdà’s plan analyzed in detail the relationship between the buildings and the streets, placing great importance on public health and well-being. Some of the initial plan’s purposes have degenerated during the 150 years since the plan’s approval (Neuman, 2011). Interior courtyards between the buildings in the block, buildings on only two sides of each block, and height limits, among others, were some of the ideas that were provided to act as urban noise reduction systems, but were subsequently dismissed. The spatial distribution of mean noise levels in Barcelona is key to the correlations between noise and the vulnerable groups considered in the study. According to our results, areas with higher percentages of children under 15 years old are relatively privileged regarding noise annoyances. In contrast, older people over 65 are slightly more present in areas with higher mean noise levels. This can be explained, first, because the mean population age in the city centre is higher than in other areas of the city, or, in other words, the population age in the city centre, mainly the Eixample district, is older. And, secondly, it is because non-EU immigrants, which are also more present in areas with lower noise levels, tend to have high birth rates, thus incrementing the rates of children in those areas (De Bustillo and Antón, 2011; Gonzalez and Ortega, 2013). Regarding unemployed population, there is a positive spatial correlation, which is explained, not only by the spatial distribution of mean noise levels, but also by the high unemployment rates—18.4% for the year 2012—due to the economic crisis. Unemployed population is scattered all over the city, but more present in the central part of the city where noise levels are the highest. The results of our work differ from those of studies carried out in other cities where a positive spatial correlation has been shown 316

Transportation Research Part D 63 (2018) 309–319

R. Lagonigro et al.

for low-income individuals and immigrants (Brainard et al., 2004; Nega et al., 2013; Carrier et al., 2016). For low-income individuals, the case of Barcelona presents an inverse spatial correlation. Low-income individuals are clearly more present in areas with lower mean noise levels. The neighbourhoods in the central part of the city experiencing high-noise have moderately high family incomes, with two neighbourhoods (Sant Gervasi-La Bonanova and Sant Gervasi-Galvany) having very high family incomes (Ajuntament de Barcelona, 2012a). Conversely, the Nou Barris district, with very low incomes, presents lower noise situations. Regarding immigration, our results did not find a significant spatial correlation, but we did find a slightly higher presence of non-EU immigrants in areas with medium noise levels. In terms of health inequities caused by air pollution, our results are in line with other studies carried out in the city of Barcelona (Borrell et al., 2010, 2013; Rodríguez-Fonseca et al., 2013). Previous studies have observed high correlations between measures of exposure to noise and air pollution (Boogaard et al., 2009; Davies et al., 2009). In the city of Girona, for instance, which is near Barcelona but is much smaller in terms of population, there is a high correlation between noise level and air pollution (Foraster et al., 2011). Contrarily, Apparicio et al. (2016) found a weak correlation between exposure to air pollution and noise in the city of Montreal, and stated that both nuisances should be measured simultaneously in order to evaluate their combined effects on the health of urban populations. Studies on air pollution in the city of Barcelona have found a positive correlation between health inequities and socioeconomic characteristics. Those studies based the socioeconomic characteristics they were examining on a social deprivation index focusing mainly on unemployment and low education, but not including income or immigration (Domínguez-Berjón et al., 2008). When introducing other socioeconomic parameters, such as low income, our work differs from studies developed in other cities, in finding an inverse correlation between noise and low-income individuals. Regarding age groups, our results are in line with those of many other earlier studies, in observing that children under 15 years old live in areas with lower mean noise levels. Older people tend to be more present in higher-noise blocks, although non-conclusive results were obtained for that group. There are some inherent limitations in our study regarding the modelling of noise and socioeconomic characteristics of the population at block level. Socioeconomic data, only available at census tract level, was transformed to city block level (Pham et al., 2012). The noise database used in our study was built with acquisition systems at street level on selected places, and simulation of noise propagation based on the real measures (Ajuntament de Barcelona, 2012b). The results of Shilton et al. (2005) on noise mapping requirements for the EU 2002/49/EC directive (European Commission, 2002) where applied on the data measuring systems to obtain a higher degree of accuracy. It is difficult to quantify the uncertainty of the calculated noise levels due to influence of many factors as weather, variation in source operating conditions, and background noise. Ausejo et al. (2011) found an overall uncertainty of ± 2 dB(A) for the noise model of the Spanish city of Palma de Mallorca, which, in their study was based on simulation using only traffic data. Overall, the noise level database seemed to be roughly coherent with the expected, as for our knowledge of the neighbourhood typology and traffic densities in the city. The height of the buildings in each area was not considered in the study. Moreover, in some areas of the city, noise has a direct impact on the lower parts of the buildings, mainly in the city centre, where there is only a short distance from the buildings to the street. This is not the case for high-income neighbourhoods of the city, where buildings are further from the street, with green areas between them.

6. Conclusions The results of this study only in part corroborate those of a number of other environmental equity studies. In fact, the city of Barcelona presents some complex relationships in regard to vulnerable groups. For the two age groups analyzed, children tend to be more present in areas with lower noise levels, whereas a slightly inverse situation is found for older people. In the case of low-income individuals and immigrants, the results of the spatial regression are relatively modest. On the other hand, in considering unemployment, we found a significant positive relationship, in line with other environmental equity studies carried out in the city. The city of Barcelona presents, in general, high mean noise levels experienced by most people living in the city. Dealing with excessive street-level noise should be a priority for public health and urban planning (McAlexander et al., 2015). Urban plans are being developed by the municipality of Barcelona (Ajuntament de Barcelona, 2012c) to assess noise pollution in the city, following the European Commission 2002/49/EC directive (European Commission, 2002). As indicated by our results, urban policies should be promoted to generally reduce mean noise levels, mainly in the central neighbourhoods, where most of the city blocks have high-noise nuisances (over 70 dB(A)). Clearly, the street and building designs of those neighbourhoods are a direct cause of the traffic noise impact (Sanchez et al., 2016). But, as stated by the EC, measures aimed only at reducing the health risks of high-noise levels, such as noise insulation, do not effectively reduce the total burden of disease due to road traffic noise. Noise-abatement measures should be integrated into mobility and land-use planning action programs and transportation decisions. This study is based on the noise database built by the Barcelona city council, which measured noise from all sources (Ajuntament de Barcelona, 2012b). This is appropriate to assess noise nuisances for the population, but, in future work, noise sources should be better characterized in order to promote urban policies appropriate to the origin of the noise (traffic, nightlife, temporary street work, outdoor and industrial equipment, mobile machinery, and others). The spatial characterization of noise zones in the city of Barcelona may be useful for developing urban policies to reduce mean noise levels. The framework proposed in our work could be used in future research to assess the effects of noise reduction urban action plans. Our study may help to assess the effects of actions implemented and may point to other specific strategies on noise reduction.

317

Transportation Research Part D 63 (2018) 309–319

R. Lagonigro et al.

Acknowledgements The authors most gratefully thank the Departament de Reducció de la Contaminació Acústica in the Ajuntament de Barcelona for providing us with the noise database for the city of Barcelona. We thank the Data Analysis and Modeling research group (2017SGR71) for the fruitful discussions we held together. References Ajuntament de Barcelona, 2012a. Distribució Territorial de la Renda Familiar Disponible per càpita a Barcelona. Barcelona. Retrieved from < http://ajuntament. barcelona.cat/barcelonaeconomia/sites/default/files/RFD 2012_bcn.pdf > . Ajuntament de Barcelona, 2012b. Mapa del soroll. Retrieved July 20, 2007, from < http://w20.bcn.cat/WebMapaAcustic/mapa_soroll.aspx > . Ajuntament de Barcelona, 2012c. Pla per la reducció de la contaminació acústica 2010-2020. Barcelona. Retrieved from < http://ajuntament.barcelona.cat/ ecologiaurbana/sites/default/files/Pla per la reducció de la contaminació acústica 2010-2020.pdf > . Ajuntament de Barcelona. Departament d’Estadística, 2012. Estadístiques per territori. Retrieved from < http://www.bcn.cat/estadistica/catala/dades/guies/a2012/ index.htm > . Akaike, H., 1973. Information theory and an extension of the maximum likelihood principle. In: Petrov, B.N., Csake, F. (Eds.), Second International Symposium on Information Theory. Akademai Kiado, Budapest, pp. 267–281. Anguelovski, I., 2013. New directions in urban environmental justice: rebuilding community, addressing trauma, and remaking place. J. Plann. Edu. Res. 33 (2), 160–175. Anguelovski, I., Connolly, J.J.T., Masip, L., Pearsall, H., 2017. Assessing green gentrification in historically disenfranchised neighborhoods: a longitudinal and spatial analysis of Barcelona. Urban Geogr. 1–34. Anselin, L., 1995. Local indicators of spatial association—LISA. Geogr. Anal. 27 (2), 93–115. Anselin, L., 2009. Spatial regression. SAGE Handbook Spat. Anal. 1, 255–276. Antweiler, W., Copeland, B.R., Taylor, S.M., 2001. Agyeman, Julian, environmental justice and sustainability. In: Atkinson/Dietz/Neumayer (Ed.), Handbook of Sustainable Development. Edward Elgar, Cheltenham, UK/Northampton, MA, USA, 2007, reprinted 2008, 171–188 Alam, Shawkat, Sustainable Development and Fr. Environment, 91, 877–908. () Apparicio, P., Carrier, M., Gelba, J., Séguin, A.-M., Kingham, S., 2016. Cyclists’ exposure to air pollution and road traffic noise in central city neighbourhoods of Montreal. J. Transp. Geogr. 57, 63–69. Arbaci, S., Tapada-Berteli, T., 2012. Social inequality and urban regeneration in Barcelona city centre: reconsidering success. Eur. Urban Reg. Stud. 19 (3), 287–311. Ausejo, M., Recuero, M., Asensio, C., Pavón, I., López, J.M., 2010. Study of precision, deviations and uncertainty in the design of the strategic noise map of the macrocenter of the city of Buenos Aires, Argentina. Environ. Model. Assess. 15 (2), 125–135. Ausejo, M., Tabacchi, M., Recuero, M., Asensio, C., Pagán, R., Pavón, I., 2011. Design of a noise action plan based on a road traffic noise map. Acta Acustica United with Acustica 97 (3), 492–502. Barceló, M.A., Saez, M., Saurina, C., 2009. Spatial variability in mortality inequalities, socioeconomic deprivation, and air pollution in small areas of the Barcelona Metropolitan Region, Spain. Sci. Tot. Environ. 407 (21), 5501–5523. Basner, M., Babisch, W., Davis, A., Brink, M., Clark, C., Janssen, S., Stansfeld, S., 2014. Auditory and non-auditory effects of noise on health. Lancet 383 (9925), 1325–1332. Bayona-i-Carrasco, J., Pujadas-i-Rúbies, I., 2014. Movilidad residencial y redistribución de la población metropolitana: los casos de Madrid y Barcelona. EURE (Santiago) 40 (119), 261–287. Berglund, B., Lindvall, T., Schwela, D. H., et al., 1999. Guidelines for community noise. In: Guidelines for community noise. OMS. Bivand, R., Altman, M., Anselin, L., Assunção, R., Berke, O., 2017. Package “spdep”. Bivand, R., Anselin, L., Berke, O., Bernat, A., Carvalho, M., Chun, Y., et al., 2011. spdep: Spatial dependence: weighting schemes, statistics and models. R package version 0.5-31, < http://CRAN. R-project. org/package= spdep > . Bocquier, A., Cortaredona, S., Boutin, C., David, A., Bigot, A., Chaix, B., Verger, P., 2012. Small-area analysis of social inequalities in residential exposure to road traffic noise in Marseilles, France. Eur. J. Public Health 23 (4), 540–546. Bolte, G., Tamburlini, G., Kohlhuber, M., 2010. Environmental inequalities among children in Europe—evaluation of scientific evidence and policy implications. Eur. J. Pub. Health 20 (1), 14–20. Boogaard, H., Borgman, F., Kamminga, J., Hoek, G., 2009. Exposure to ultrafine and fine particles and noise during cycling and driving in 11 Dutch cities. Atmos. Environ. 43 (27), 4234–4242. Borrell, C., Arias, A., 1995. Socioeconomic factors and mortality in urban settings: the case of Barcelona, Spain. J. Epidemiol. Community Health 49 (5), 460–465. Borrell, C., Marí-Dell’Olmo, M., Serral, G., Martínez-Beneito, M., Gotsens, M., et al., 2010. Inequalities in mortality in small areas of eleven Spanish cities (the multicenter MEDEA project). Health Place 16 (4), 703–711. Borrell, C., Pons-Vigués, M., Morrison, J., Díez, È., 2013. Factors and processes influencing health inequalities in urban areas. J. Epidemiol. Community Health 67 (5), 389–391. Brainard, J.S., Jones, A.P., Bateman, I.J., Lovett, A.A., 2004. Exposure to environmental urban noise pollution in Birmingham, UK. Urban Stud. 41 (13), 2581–2600. Canal-Domínguez, J.F., Rodríguez-Gutiérrez, C., 2008. Analysis of wage differences between native and immigrant workers in Spain. Span. Econ. Rev. 10 (2), 109–134. Carrier, M., Apparicio, P., Séguin, A.-M., 2016. Road traffic noise in Montreal and environmental equity: what is the situation for the most vulnerable population groups? J. Transp. Geogr. 51, 1–8. Casellas, A., Dot, E., Pallares-Barbera, M., 2010. Image creation, visibility and tourism. Growth strategies for the future (Creación de imagen, visibilidad y turismo como estrategias de crecimiento económico de la ciudad). Finisterra XLV (90), 153–172. Catalán, B., Saurí, D., Serra, P., 2008. Urban sprawl in the Mediterranean?: Patterns of growth and change in the Barcelona Metropolitan Region 1993–2000. Landscape Urban Plann. 85 (3), 174–184. Cole, H., Triguero, M., Connolly, J., Anguelovski, I., 2017. A longitudinal and spatial analysis assessing green gentrification in historically disenfranchised neighborhoods of barcelona: implications for health equity. J. Transp. Health 5, S44. Cutter, S.L., 1995. Race, class and environmental justice. Prog. Hum. Geogr. 19 (1), 111–122. Davies, H.W., Vlaanderen, J.J., Henderson, S.B., Brauer, M., 2009. Correlation between co-exposures to noise and air pollution from traffic sources. Occup. Environ. Med. 66 (5), 347–350. De Bustillo, R.M., Antón, J.-I., 2011. From rags to riches? Immigration and poverty in Spain. Popul. Res. Policy Rev. 30 (5), 661. Domínguez-Berjón, M.F., Borrell, C., Cano-Serral, G., Esnaola, S., Nolasco, A., Pasarín, M.I., Escolar-Pujolar, A., 2008. Constructing a deprivation index based on census data in large Spanish cities [the MEDEA project] [Construcción de un índice de privación a partir de datos censales en grandes ciudades españolas (Proyecto MEDEA)]. Gaceta Sanitaria 22 (3), 179–187. European Commission, 1996. The Green Paper on Future Noise Policy. COM 96, 540. Commission, European, 2002. Directive 2002/49/EC of the European parliament and the Council of 25 June 2002 relating to the assessment and management of environmental noise. Off. J. Eur. Commun. 189 (12), 12–26. European Commission, 2003. Recommendation concerning the guidelines on the revised interim computation methods for industrial noise, aircraft noise, road traffic noise and rail- way noise, and related emission data. Directive 2003/613/EC. European Commission Working Group, 2007. Assessment of exposure to noise (WG-AEN). Good Practice Guide for Strategic Noise Mapping and the Production of

318

Transportation Research Part D 63 (2018) 309–319

R. Lagonigro et al.

Associated Data on Noise Exposure, Ver, 2, 13. European Union, 2013Decision No 1386/2013/EU of the European Parliament and of the Council. Retrieved from: < http://data.europa.eu/eli/dec/2013/ 1386/oj > . Evans, G.W., Kantrowitz, E., 2002. Socioeconomic status and health: the potential role of environmental risk exposure. Annu. Rev. Public Health 23 (1), 303–331. Foraster, M., Deltell, A., Basagaña, X., Medina-Ramón, M., Aguilera, I., Bouso, L., et al., 2011. Local determinants of road traffic noise levels versus determinants of air pollution levels in a Mediterranean city. Environ. Res. 111 (1), 177–183. Forkenbrock, D.J., Schweitzer, L.A., 1999. Environmental justice in transportation planning. J. Am. Plann. Assoc. 65 (1), 96–112. Goines, L., Hagler, L., et al., 2007. Noise pollution: a modern plague. South. Med. J.-Birmingham Alabama 100 (3), 287. Gonzalez, L., Ortega, F., 2013. Immigration and housing booms: Evidence from Spain. J. Regional Sci. 53 (1), 37–59. Haralabidis, A.S., Dimakopoulou, K., Vigna-Taglianti, F., Giampaolo, M., Borgini, A., Dudley, M.-L., et al., 2008. Acute effects of night-time noise exposure on blood pressure in populations living near airports. Eur. Heart J. 29 (5), 658–664. Hoffmann, B., Robra, B.P., Swart, E., 2003. Social inequality and noise pollution by traffic in the living environment–an analysis by the German Federal Health Survey (Bundesgesundheitssurvey). Gesundheitswesen (Bundesverband Der Arzte Des Offentlichen Gesundheitsdienstes (Germany)) 65 (6), 393–401. Hornberg, C., Pauli, A., 2007. Child poverty and environmental justice. Int. J. Hyg. Environ. Health 210 (5), 571–580. International Electrotechnical Commission, 2003. IEC 61672:2003: electroacoustics sound level meters. Tech. Rep. IEC. Ising, H., Kruppa, B., 2004. Health effects caused by noise: evidence in the literature from the past 25 years. Noise Health 6 (22), 5–13. Judge, K., Platt, S., Costongs, C., Jurczak, K., et al., 2006Health inequalities: a challenge for Europe. An Independent, Expert Report Commissioned by the UK Presidency of the EU, London. Kihal-Talantikite, W., Padilla, C.M., Lalloue, B., Rougier, C., Defrance, J., Zmirou-Navier, D., Deguen, S., 2013. An exploratory spatial analysis to assess the relationship between deprivation, noise and infant mortality: an ecological study. Environ. Health 12 (1). Lam, K.C., Chan, P.K., 2006. Socio-economic status and inequalities in exposure to transportation noise in Hong Kong. Transportation 2 (1), 107–113. Lam, K., Chung, Y.T., 2012. Exposure of urban populations to road traffic noise in Hong Kong. Transp. Res. Part D: Transp. Environ 17 (6), 466–472. Landrigan, P.J., Rauh, V.A., Galvez, M.P., 2010. Environmental justice and the health of children. Mount Sinai J. Med.: J. Transl. Personal. Med. 77 (2), 178–187. Mackenbach, J.P., Bakker, M.J., 2003. Tackling socioeconomic inequalities in health: analysis of European experiences. Lancet(October 25). Martínez-Alier, J., 2003. Scale, environmental justice, and unsustainable cities. Cap. Nat. Soc. 14 (4), 43–63. Martori, J.C., Apparicio, P., 2011. Changes in spatial patterns of the immigrant population of a southern European metropolis: the case of the Barcelona metropolitan area (2001–2008). Tijdschrift Voor Economische En Sociale Geografie 102 (5), 562–581. McAlexander, T.P., Gershon, R.R.M., Neitzel, R.L., 2015. Street-level noise in an urban setting: assessment and contribution to personal exposure. Environ. Health : Glob. Access Sci. Source. Melnick, A.L., 2002. Introduction to Geographic Information Systems in Public Health. Jones & Bartlett Learning. Moreno-Jiménez, A., Cañada-Torrecilla, R., Vidal-Domínguez, M.J., Palacios-García, A., Martínez-Suárez, P., 2016. Assessing environmental justice through potential exposure to air pollution: a socio-spatial analysis in Madrid and Barcelona, Spain. Geoforum 69, 117–131. Moudon, A.V., 2009. Real noise from the urban environment: how ambient community noise affects health and what can be done about it. Am. J. Prev. Med. 37 (2), 167–171. Nega, T.H., Chihara, L., Smith, K., Jayaraman, M., 2013. Traffic noise and inequality in the twin cities, Minnesota. Hum. Ecol. Risk Assess.: Int. J. 19 (3), 601–619. Neuman, M., 2011. Centenary paper: Ildefons Cerdà and the future of spatial planning: the network urbanism of a city planning pioneer. Town Plann. Rev. 82 (2), 117–144. NMPB-Routes-96, 1996. (SETRA-CERTU-LCPC, CSTB), is the French national computation method, referred to in ‘Arrêté du 5 mai 1995 relatif au bruit des infrastructures routières. Journal Officiel Du 10 Mai 1995, Article 6. O’Brien, R.M., 2007. A caution regarding rules of thumb for variance inflation factors. Qual. Quant. 41 (5), 673–690. O’Neill, M.S., McMichael, A.J., Schwartz, J., Wartenberg, D., 2007. Poverty, environment, and health: the role of environmental epidemiology and environmental epidemiologists. Epidemiology 18 (6), 664–668. Parlament de Catalunya, 2002. Llei de protecció contra la contaminació acústica. Offi. J. Catalan Parliament(3675). Pham, T.-T.-H., Apparicio, P., Séguin, A.-M., Landry, S., Gagnon, M., 2012. Spatial distribution of vegetation in Montreal: an uneven distribution or environmental inequity? Landscape Urban Plann. 107 (3), 214–224. R Core Team, 2015). R: A Language and Environment for Statistical Computing. Vienna, Austria. Retrieved from < http://www.r-project.org/ > . Rey, S.J., Janikas, M.V., 2005. Regional convergence, inequality, and space. J. Econ. Geogr. 5 (2), 155–176. Rodríguez-Fonseca, M., Palència, L., Marí-Dell’Olmo, M., Gandarillas, A., Domínguez-Berjón, M.F., Gotsens, M., Borrell, C., 2013. Evolution of socio-economic inequalities in mortality in small geographical areas of the two largest cities in Spain (Barcelona and Madrid), 1996–2007. Public Health 127 (10), 916–921. Sanchez, G.M.E., Van Renterghem, T., Thomas, P., Botteldooren, D., 2016. The effect of street canyon design on traffic noise exposure along roads. Build. Environ. 97, 96–110. Schlosberg, D., 2013. Theorising environmental justice: the expanding sphere of a discourse. Environ. Polit. 22 (1), 37–55. Shilton, S., van Leeuwen, H., Nota, R., 2005. Error propagation analysis of XPS 31-133 and CRTN to help develop a noise mapping data standard. In: Forum Acusticum, Budapest, Hungary. Spanish National Statistics Office, 2011. Censo de población y Viviendas, 2011. INE, Madrid. Stansfeld, S., Haines, M., Brown, B., 2000. Noise and health in the urban environment. Rev. Environ. Health 15 (1–2), 43–82. Stansfeld, S., Matheson, M.P., 2003. Noise pollution: non-auditory effects on health. Br. Med. Bull. 68 (1), 243–257. Sze, J., London, J., Shilling, F., Gambirazzio, G., Filan, T., Cadenasso, M., 2009. Defining and contesting environmental justice: socio-natures and the politics of scale in the delta. Antipode 41 (4), 807–843. Tobías, A., Recio, A., Díaz, J., Linares, C., 2015. Health impact assessment of traffic noise in Madrid (Spain). Environ. Res. 137, 136–140. Valera, S., Pérez-Tejera, F., Anguera, M.T., Sicilia, L., 2018. Evaluating the uses and environmental characteristics of 40 public parks and squares in Barcelona by means of systematic observation. Psyecology 9 (2), 118–151. van Kamp, I., Davies, H., 2013. Noise and health in vulnerable groups: a review. Noise Health 15 (64), 153–159. Walker, G., 2010. Environmental justice, impact assessment and the politics of knowledge: the implications of assessing the social distribution of environmental outcomes. Environ. Impact Assess. Rev. 30 (5), 312–318. Walker, G., 2012. Environmental Justice: Concepts, Evidence and Politics. Routledge, New York. World Health Organization, 2011. Burden of disease from environmental noise: Quantification of healthy life years lost in Europe. In: Burden of Disease from Environmental Noise: Quantification of Healthy Life Years Lost in Europe, p. 126. World Health Organization. UN-Habitat, 2010. Hidden Cities: Unmasking and Overcoming Health Inequities in Urban Settings. World Health Organization, Geneva. World Health Organization, C. on the S. D. of H, 2008. Closing the Gap in a Generation: Health Equity through Action on the Social Determinants of Health. World Health Organization, Geneva.

319