Mapping Invisible Risk: A Low-Cost Strategy for Identifying Air and Noise Pollution in Latin American Cities

Abstract

Urban populations in Latin America are highly exposed to traffic-related pollutants, yet monitoring networks remain limited. This study proposes a low-cost methodology to identify urban pollution hotspots in the city of Córdoba, Argentina, by categorizing 20 sites based on traffic categories using Google Traffic data. Measurements of PM_2.5, polycyclic aromatic hydrocarbons (PAHs), and equivalent sound pressure level (LA_eq) were conducted over a 21-day cold-season period. Mean PM_2.5 concentrations ranged from 7.5 to 27.3 µg/m³, and total PAHs ranged from 1.4 to 7.9 ng/m³. Sites with high and medium traffic density exhibited significantly higher PAH concentrations and noise levels, with LAeq₅ values exceeding 65 dB at all urban core locations. Conversely, PM_2.5 concentrations were higher at peripheral sites due to topography, dust resuspension, and wildfire events. Strong correlations were found between vehicular flow and noise (r = 0.94), and between heavy-vehicle proportion and noise (r = 0.60). The lifetime lung cancer risk associated with PAH exposure was classified as “low” according to USEPA criteria. This traffic-based categorization approach provides a rapid and cost-effective tool for identifying high-risk areas in resource-limited settings, supporting urban planning and public health interventions.

1. Introduction

Currently, more than 55% of the world’s population resides in urban areas, and this percentage rises to 81% in Latin America and the Caribbean [1]. In these regions, residents are exposed to a wide range of chemical and physical pollutants that negatively impact on their quality of life. Among the most concerning is air pollution, which represents a major health risk, causing 4.2 million premature deaths in 2019—most of them in low- and middle-income countries [2].

Among air pollutants, fine particulate matter (PM_2.5) is considered one of the most hazardous. It has been associated with cardiovascular conditions such as arrhythmia, infarction, and hypertension [3], asthma exacerbations, and respiratory infections in children [4,5], among other effects. The toxicological impact of PM_2.5 depends not only on its concentration but also on its chemical composition [6]. A particularly relevant component is polycyclic aromatic hydrocarbons (PAHs), of which 70–90% are typically adsorbed onto respirable particles [7]. PAHs are widely studied due to their well-documented toxic, mutagenic, and carcinogenic properties [8,9,10,11].

In addition to air pollution, high noise levels have become a growing environmental and public health concern [12,13,14]. An increasing number of studies have reported health effects linked to environmental noise exposure, including annoyance [15], sleep disturbances [16], ischemic heart diseases (IHDs) [17], elevated blood pressure, and hypertension [18]. Recent research has also explored the potential combined effects of air pollutants and noise on human health [19], although a limited number of studies have quantified these associations [20,21].

Urban road traffic is a common source of hazardous pollutants such as PM_2.5 and PAHs, as well as environmental noise [22,23,24]. In response, many developed countries have implemented extensive monitoring networks. For example, Europe has more than 3500 air quality monitoring stations [25] and detailed traffic censuses that provide information on infrastructure and vehicle flow [26]. These efforts support air quality and public health management, as well as more complex analyses and predictive models [27,28].

Despite numerous studies addressing traffic-related pollution in developed regions, a significant research gap persists in low- and middle-income countries, where continuous air quality monitoring is scarce and information on vehicle fleet is often incomplete or unavailable. Consequently, most cities across Latin America lack the capacity to identify local pollution hotspots or assess exposure-related health risks with sufficient spatial resolution [29]. As a result, the health risks faced by urban populations remain largely undocumented. In Córdoba, Argentina, this situation is further exacerbated by the absence of continuous monitoring systems and by local topographic conditions that favor frequent thermal inversions, which trap pollutants near the surface. Moreover, a large and growing motorized fleet has a considerable impact on air quality, particularly during the colder months [30].

In this context, there is a pressing need for practical, low-cost approaches that combine openly available data sources with short-term environmental measurements to characterize air and noise pollution patterns. Therefore, the primary objective of this study was to evaluate the effectiveness of traffic-based categorization as a tool for identifying critical areas that may pose health risks and to guide preventive and monitoring actions. Specifically, we investigated whether classifying urban sites according to traffic density could serve as a proxy for detecting pollution hotspots in the absence of an extensive monitoring network. To this end, we measured air pollutant concentrations and noise levels at 20 sites in Córdoba during the cold season and analyzed their relationship with traffic density categories defined using Google Traffic data. Finally, we estimated the additional cancer risk associated with PAH exposure. This study provides a framework that can be replicated in other urban areas facing similar data and infrastructure limitations.

2. Materials and Methods

2.1. Study Area and Sampling Sites

Córdoba is one of the largest cities in Argentina, with over 1.5 million inhabitants and an urban area of approximately 570 km² [31]. Located in the central region of Argentina (31°24′ S, 75°11′ W), Córdoba city has an irregular topography that gradually ascends from the central area to the periphery. These features reduce air circulation and promote frequent thermal inversions during the cold season [30]. Vehicular traffic is the main source of atmospheric pollutants in the city, as demonstrated by both instrumental and biomonitoring studies [32,33]. In 2019, the vehicle fleet exceeded one million units [34]. Twenty sampling sites were selected across the urban area (Figure 1), where vehicular flow, fleet composition, number of inhabitants, and other relevant parameters were recorded. At each site, PM_2.5 concentrations, particle-bound PAHs, and equivalent sound pressure level (L_Aeq) were measured over a 21-day period during the cold season.

2.2. Sites Categorization

In the absence of official traffic data, site classification was based on the Google Maps traffic application. Google uses Floating Car Data (FCD), a widely available source that collects real-time information from mobile phones and GPS devices traveling within the road network [35]. This tool is particularly useful in Argentina, where there are 61 million active mobile connections and 39.8 million internet users, representing 87.2% of the population [36]. Google Maps displays a four-color traffic layer—green, orange, red, and dark red—indicating real-time traffic conditions inferred from vehicle speed and congestion levels. While it does not provide explicit vehicle flow data (i.e., vehicles per hour or per kilometer), the color-coded scheme offers a qualitative proxy for traffic density in areas lacking direct traffic counts [37].

The typical traffic option was selected to analyze traffic patterns. To capture both peak and off-peak hours, six specific weekday time points were considered: 8:00 a.m., 10:00 a.m., 3:00 p.m., 5:00 p.m., 7:00 p.m., and 10:00 p.m. For each of these points, images were obtained for four different days to ensure temporal representativeness. The linear meters of roads within each traffic category were then quantified within a 500 m buffer around each site using QGIS (version 3.28—Firenze) [38].

For each time and day considered, a site was classified as LDT if ≥70% or more of the streets were green, as HDT if the summatory of orange and red streets were ≥50%, and as MDT in all other cases. The final classification was determined by the mode across all evaluated time points. (Step-by-step procedure for site classification is in the Supplementary Materials). Additionally, sites were categorized by geographical location into peripheral sites (PS) and central sites (CS).

To complement traffic density estimates, on-site vehicular characterization was conducted. One-hour video recordings were made twice at each site, concurrent with noise measurements (see Section 2.5), to quantify and classify passing vehicles. The two recordings—one in the morning and one in the afternoon—were carried out five days apart to better capture day-to-day variability in traffic conditions. All motorized vehicles were recorded and categorized as light vehicles (motorcycles, cars, vans) or heavy vehicles (trucks and buses). The proportion of heavy vehicles was calculated due to their usually greater noise emissions and higher PM output, especially from diesel engines [39].

2.3. PM_2.5

PM_2.5 samples were collected over 72 h using Harvard Impactors coupled with a low-volume vacuum pump (flow rate of 16 ± 4 L/min) [40] using 47-mm polytetrafluoroethylene (PTFE) filters with a 1.0 µm pore size (Whatman). The sampler was positioned at a height between 3 and 7 m above ground level at each site. A total of 100 samples were collected from 20 sites (5 samples per site) during the winter of 2022 (from 25 July to 7 October 2022). After each 72 h sampling period, a 24 h interval was reserved for filter replacement and equipment calibration. Ambient temperature and vacuum pump pressures were recorded during the sampling period for air flow correction.

PM_2.5 mass was determined gravimetrically using a balance with a resolution of 0.01 mg (Sartorius) [6,41]. The PM_2.5 concentration was subsequently calculated from the measured mass and the volume of filtered air. Results were expressed as 24 h mean concentrations.

2.4. Polycyclic Aromatic Hydrocarbons

Determination of PAHs was conducted following the protocol by Tames et al. [42]. Briefly, PAHs adsorbed onto PM_2.5 were extracted with dichloromethane, filtered through 0.22 μm polyvinylidene fluoride (PVDF) membranes, and the solvent was changed to acetonitrile. The extract was further reduced by evaporation to approximately a volume of 200 μL. The PAH content in each filter was quantified using high-performance liquid chromatography (HPLC) (Dionex Ultimate 3000, Thermo Scientific, Waltham, MA, USA) with a thermo-stabilized C-18 reverse-phase column at 35 °C and fluorescence detection. A combination of acetonitrile and MilliQ water was used as the mobile phase under non-isocratic conditions with a flow rate of 1.2 mL/min. In all cases, the injection volume was 20 μL. The quantified PAHs included naphthalene (Nap), acenaphthene (Ace), fluorene (Fle), phenanthrene (Phe), anthracene (Ant), fluoranthene (Fla), pyrene (Pyr), benzo[a]anthracene (BaA), chrysene (Chr), benzo[b]fluoranthene (BbF), benzo[k]fluoranthene (BkF), benzo[a]pyrene (BaP), dibenzo[a,h]anthracene (DBA), indeno [1,2,3-c,d]pyrene (IDP), and benzo[g,h,i]perylene (BghiP).

Ten-point calibration curves for all PAHs were employed (Supelco) to obtain PAH concentrations (R² values for all compounds were greater than 0.9991). The Chromeleon 7.2.0.3765 integrated software (Dionex, Thermo Scientific, Waltham, MA, USA) was used for chromatogram analysis. The results were expressed both as the PAH mass per air volume (ng/m³) and as the mass of individual PAHs per mass of PM_2.5 (ng/μg). The procedure was validated through recovery tests using pure standards, with recovery percentages ranging from 85.28% and 121.46%.

Quantified PAHs were also classified into three groups by molecular weight for further analysis: low-molecular-weight (LMW), containing 2- and 3-ring PAHs (Nap, Ace, Fle, Phe, and Ant), medium-molecular-weight (MMW), containing 4-ring PAHs (Fla, Pyr, BaA, and Chr), and high-molecular-weight (HMW), containing 5- and 6-ring PAHs (BbF, BkF, BaP, DBA, BghiP, and IDP). Several diagnostic ratios were performed for source estimation.

2.5. Noise Level Determination

Traffic noise levels were measured using a Sound Level Meter (UNI-T model UT351) with a measurement range from 30 to 130 ± 1.5 dB (A-weighted). Previous studies have used measuring times ranging from 1 to 15 min for A-weighted equivalent sound pressure level (L_Aeq) [43,44,45,46]. Therefore, 5 min equivalent continuous sound pressure level measurements were obtained [47] and reported as units of A-weighted decibels (L_Aeq5) [42]. Special care was taken to isolate measurements from other urban noise sources such as pedestrians, construction sites, and stores. Two measurements were conducted at each site between 9 a.m. and 6 p.m. on weekdays [48]. The sound level meter was mounted on a tripod at a height of 1.5 m, 1 m away from the road edge, under environmental conditions of dry surfaces, no rain, and wind speed below 5 m/s [49].

2.6. Health Risk Assessment

The free software AirQ+ (Version 2.2) [50] was used to express the carcinogenicity of the analyzed PAHs. For this, BaP equivalents were obtained employing Toxic Equivalency Factors (TEFs), as described by Nisbet and Lagoy [51]. BaP was selected as the reference due to its status as the most well-known and documented PAH and its classification as a Group 1 carcinogen (carcinogenic to humans) by the IARC. The BaP equivalent concentration (BaP_eq) was calculated using the following:

$${\mathrm{B}\mathrm{a}\mathrm{P}}_{\mathrm{e}\mathrm{q}}={\sum }_{\mathrm{i}=1}^{\mathrm{n}}{\mathrm{C}}_{\mathrm{i}}\times {\mathrm{T}\mathrm{E}\mathrm{F}}_{\mathrm{i}}$$

${\mathrm{C}}_{\mathrm{i}}$ is the concentration, and ${\mathrm{T}\mathrm{E}\mathrm{F}}_{\mathrm{i}}$ is the Toxic Equivalent Factor for the i^th target compound, respectively.

Once the BaP_eq was determined, the AirQ+ software was employed to assess the lifetime lung cancer risk (LLCR) associated with it using the following equation:

$$\mathrm{L}\mathrm{L}\mathrm{C}\mathrm{R}=\sum {\mathrm{B}\mathrm{a}\mathrm{P}}_{\mathrm{e}\mathrm{q}}\times {\mathrm{U}\mathrm{R}}_{\mathrm{B}\mathrm{a}\mathrm{P}}$$

where BaP_eq represents the sum of individual BaP_eq, and UR_BaP is the inhalation cancer unit risk. The results were expressed as the number of additional cancer cases per 100,000 inhabitants over 70 years. According to the USEPA, LLCR for an individual can be categorized as very low (LLCR < 10⁻⁶), low (10⁻⁶ ≤ LLCR < 10⁻⁴), moderate (10⁻⁴ ≤ LLCR < 10⁻³), high (10⁻³ ≤ LLCR < 10⁻¹), and very high (LLCR > 10⁻¹) [52].

2.7. Statistical Analysis

The normality of the variables (PM_2.5, PAHs, and L_Aeq5) was assessed using the modified Shapiro–Wilk test. When the normality assumption was not met, non-parametric statistical methods were applied. Differences among traffic density categories (LDT, MDT, and HDT) were evaluated using ANOVA, followed by Fisher’s LSD post hoc test. Spearman’s correlation analyses were performed to investigate relationships between all pollutants and vehicular data. A significance level of 5% (p < 0.05) was applied in all cases. All statistical analyses were conducted using the Navure Team (2023) software (Version 2.7.1) [53].

3. Results and Discussion

3.1. Sampling Sites and Categorization

The results of the categorization based on traffic density are shown in Figure 2. A clear gradient can be observed, with LDT sites located in the peripheral areas of the city, while MDT and HDT categories emerge as one approaches the downtown area. This was an expected pattern since the vehicle fleet converges to the downtown. In fact, Site 2 was the only peripheral location that showed medium-density traffic due to the city’s spatial expansion towards the northwest. As shown in

Table 1. Mean hourly number of vehicles and proportion of heavy vehicles recorded at each site, categorized by traffic density (LDT: low-density traffic; MDT: medium-density traffic; HDT: high-density traffic) and geographical location (PS: peripheral sites; CS: central sites).

Location Category	Site Number	Traffic Density Category	Mean Hourly Number of Vehicles	Heavy Vehicles (%)
PS	10	LDT	1608	5.22
	19		864	5.56
	20		660	9.09
	11		410	7.32
	16		360	6.67
	13		210	7.14
	12		48	0
	1		36	0
	7		11	0
	15		7	0
	17		0 *	-
	2	MDT	1776	3.38
CS	18	MDT	996	18.67
	8		768	3.13
	3		720	0
	5	HDT	1896	16.46
	6		1380	5.43
	4		1070	0.93
	9		864	1.39
	14		739	1.39

* No vehicles were recorded at this site, as it was located within a military facility.

, this area has a high number of vehicles, which corresponds with one of the largest residential neighborhoods in the city.

As shown in Table 1, sites categorized as MDT and HDT exhibited the highest number of vehicles, ranging from 120 to 316. The LDT category showed a wider range of vehicle numbers, from 0 to 268, due to the heterogeneity of the sites. In this category, the sites with the lowest vehicle counts were located in residential areas, distant from major roads or heavy-traffic streets, whereas those with higher counts (sites 10, 19, and 20) were situated along roads connecting rural and industrial areas with the city. These results show a correspondence between the Google Traffic categories and the number of vehicles recorded at each monitoring site. This suggests that, in the absence of direct vehicle count data, Google’s color-coded traffic scheme could serve as a reliable estimator of vehicular flow.

3.2. PM_2.5 and PAHs Concentration

PM_2.5 levels were highest in the LDT category (33.28 μg/m³), showing a significant difference compared to the MDT and HDT categories (21.17 μg/m³ and 22.54 μg/m³, respectively) (Figure 3). The recorded concentrations ranged from a minimum of 3.50 μg/m³ in HDT to a maximum of 116.43 μg/m³ in LDT. The highest mean values were detected at sites 13 (53.14 μg/m³), 6 (47.96 μg/m³), and 12 (45.31 μg/m³), all located in the western area of the city. The elevated PM_2.5 levels observed in LDT sites may be partly explained by their location in peripheral zones characterized by large areas of exposed soil, highly susceptible to dust resuspension processes—particularly under dry and windy conditions. In addition, these areas generally have fewer and lower-rise buildings, enhancing pollutant dispersion driven by wind.

These findings are consistent with previous studies in Córdoba. Achad et al. [54] and Lanzaco et al. [55], who worked near site 18, analyzed PM_2.5 composition and reported high levels of mineral/crustal elements (e.g., Al, Si, Ca, Fe, Ti), indicating that dust resuspension and atmospheric transport as one of the most relevant sources of PM_2.5. They also attributed compositional variability to traffic, industrial activity, and biomass burning, including long-range transport of emissions from forest fires. Amarillo et al. [56], monitoring PM_2.5 at nine sites from May to August 2018, found higher concentrations in peripheral areas, linked to industrial and rural emissions. Likewise, Della Ceca et al. [57], using MODIS satellite data (2003–2015), reported seasonal peaks in aerosol optical depth—especially in late winter and spring—mainly driven by biomass burning and pollutant transport, over rural and agricultural areas surrounding the city.

During the study period, 73% of the samples (n = 64) exceeded the WHO 24 h air quality guideline value of 15 µg/m³. Notably, the highest concentrations were mainly recorded after a wildfire that burned 1.52 km² in a protected area, located less than 2 km west of the city’s urban area (Figure A1—Appendix A). Particle transport by westerly winds could also have contributed to these elevated PM_2.5 levels. Furthermore, in September, several wildfires were recorded south of the city, affecting over 115 km² [58] (Figure A2—Appendix A). The transport of smoke from these fires by prevailing southern winds during this period also likely contributed to the observed increase in particle concentrations within the city.

Regarding total PAH concentrations, little variation was observed among traffic categories, with averages of 2.54 ng/m³ for LDT, 2.99 ng/m³ for MDT, and 2.19 ng/m³ for HDT, with no statistically significant differences. PAHs diagnostic ratios are commonly used to identify emission sources, distinguishing between petroleum- and combustion-related contributions [59]. The diagnostic ratios presented in

Table 2. Diagnostic ratios used to identify potential PAH sources according to traffic categories (LDT: low-density traffic; MDT: medium-density traffic; HDT: high-density traffic).

Ratio	Category	Value	Reference	Possible Source	Author
BaP/(BaP + Chr)	LDT	0.38	0.5~Diesel 0.73~Gasoline	Diesel	Ravindra et al., [61]
	MDT	0.41		Diesel
	HDT	0.39		Diesel
BbF/BkF	LDT	2.43	>0.5~Diesel	Diesel	Ravindra et al., [61]
	MDT	2.49		Diesel
	HDT	2.44		Diesel
BghiP/BaP	LDT	1.95	0.86–0.91~Road dust 1.2–2.2~Diesel 2.5–3.3~Gasoline	Diesel	Rogge et al., [62]; Oda et al., [63]
	MDT	2.03		Diesel
	HDT	2.30		Diesel
IDP/(IDP + BghiP)	LDT	0.35	0.21–0.22~Gasoline 0.35–0.7~Diesel 0.56~Coal	Diesel	Ravindra et al., [61]
	MDT	0.37		Diesel
	HDT	0.39		Diesel
(BbF + BkF)/BghiP	LDT	1.31	0.33~Gasoline 1.6~Diesel 2.18~Wood burning	Diesel	Li & Kamens [64]
	MDT	1.22		Diesel
	HDT	1.16		Diesel

suggest that diesel vehicles were the primary source of PAHs, with no significant differences observed between traffic categories, as was reported by Tavera Busso et al. [41]. Although diagnostic ratios do not provide definitive conclusions in this study, they remain a valuable tool for preliminary assessment of potential PAH sources [60].

Molecular weight analysis reveals that, for each traffic category, the concentrations of MMW PAHs are statistically higher than the others, while LMW PAHs contributed the least (

Table 3. Mean PAH levels grouped by molecular weight (ng/m³) (±S.D.) for each traffic category.

Molecular Weight	LDT		MDT		HDT
LMW	0.22 ± 0.11	C	0.32 ± 0.41	C	0.20 ± 0.15	C
MMW	1.07 ± 0.50	A	1.12 ± 0.56	A	0.89 ± 0.40	A
HMW	0.52 ± 0.33	B	0.61 ± 0.36	B	0.45 ± 0.27	B

Note: Categories with different capital letters differ significantly (p < 0.05).

). This trend is in line with previous studies reporting that MMW PAHs such as Fla, Pyr, and Chr are dominant in emissions from diesel engine vehicles, whereas particles emitted by gasoline engine vehicles contain a higher proportion of HMW PAHs (BbF, BaP, DBA, and IDP) [65]. On the other hand, the relatively low concentrations of LMW PAHs may be partly explained by their greater tendency to undergo a particle-to-gas phase transformation and subsequently condense onto pre-existing coarse particles [66].

The analysis of PAH concentrations in PM_2.5 revealed that the highest contributions to total particle mass were observed in the HDT (0.15 ng/µg) and MDT (0.17 ng/µg) traffic categories, showing no statistically significant difference between them. However, both categories exhibited significantly higher values compared to the LDT (0.10 ng/µg).

These findings align with previous studies, including those by Stein and Toselli [67], Olcese and Toselli [32], and, more recently, Carreras et al. [6], which identified vehicular traffic as the primary source of PAHs in the city. With a registered fleet of 600,000 private cars in Córdoba city and a total of 1,725,667 vehicles across the province, there is approximately 1 vehicle for every 2.5 residents [31,68]. This extensive fleet, combined with nearly 1000 diesel-powered public transportation units [34], converges predominantly in the downtown area and its surrounding periphery. This may explain, as shown in Figure 4, why the highest individual values of PAH-to-PM_2.5 ratios (ng/µg) were recorded at sites 5, 18, and 8. Another unexpectedly high concentration was observed at site 11, which could be related to its proximity to one of the city’s public transportation depots or the ring road. Nevertheless, these are preliminary interpretations, and further source apportionment analyses would be required to more accurately determine the origin of the elevated levels observed.

3.3. Equivalent Sound Pressure Level (L_Aeq5)

The highest mean L_Aeq5 values were recorded in the HDT and MDT categories (67.87 dB and 67.67 dB, respectively), significantly exceeding those observed in the LDT category (58.84 dB). This same pattern was observed when analyzing PAH concentrations. As shown in Figure 5, elevated sound levels are concentrated in the city center, with additional isolated high-value points appearing in peripheral areas within LDT (such as sites 10, 19, and 20). These sites, located near major roads connecting rural and industrial zones, are characterized by a high volume of vehicles traveling at greater speeds, thus contributing to elevated sound pressure levels [69,70,71]. In contrast, other sites within this category are situated in residential neighborhoods, where both vehicle speeds and traffic density are lower. This is directly related to the lower sound levels observed in these areas and explains the higher standard deviation associated with this category. In fact, Espadaler-Clapés et al. [72] recently reported a strong correlation between noise and the accumulation of heavy and medium–heavy vehicles (e.g., buses and trucks). Although a relationship was also observed by light motor vehicles (e.g., cars) and two-wheel powered vehicles (e.g., motorcycles), it was notably weaker. Moreover, under conditions of high traffic density, substantial variability in noise levels was observed relative to vehicle speed, with no clear trend emerging. These findings are consistent with those of the present study and help explain why downtown sites exhibit the highest noise levels. Additionally, the high proportion of heavy vehicles in these areas is closely linked to the convergence of public transport units (buses), thereby increasing the number of commuters exposed to elevated noise levels.

From a health perspective, the average noise levels recorded in this study exceeded the World Health Organization’s [73] recommended limit of 53 dB for road traffic noise. All sites classified as HDT or MDT exhibit L_Aeq5 values above this threshold, whereas residential areas within the LDT category generally remain within the guideline values.

3.4. Relationships Among Measured Variables

Table A1. Spearman’s correlation matrix for air pollutants (PM_2.5 and PAHs), total vehicle count, heavy vehicles proportion (%), and L_Aeq5.

	Vehicle Number	Heavy Vehicles (%)	PM2.5 (µg/m3)	LAeq5 (dB)	PAHs (ng/µg)
Vehicle number	-
Heavy Vehicles	0.44 ***	-
PM2.5	−0.24 *	ns	-
LAeq5	0.94 ***	0.60 ***	ns	-
PAHs	0.27 **	ns	−0.79 ***	ns	-

* Significant at 0.05 probability level; ** Significant at 0.01 probability level; *** Significant at 0.001 probability level.

(Appendix A) presents the Spearman correlation coefficients between noise levels, air pollutants (PM_2.5, PAHs), total vehicle count, and the proportion of heavy vehicles (buses and trucks) across the city. Noise levels showed a strong positive correlation with the total number of vehicles (r = 0.94) and a moderate positive correlation with the proportion of heavy vehicles (r = 0.60). Additionally, weak but statistically significant correlations were identified between total vehicle count and both PM_2.5 (r = −0.24) and PAH concentration (r = 0.27). A notable moderate-to-strong negative correlation was also observed between PAH and PM_2.5 concentrations (r = −0.79).

The positive correlation between noise levels, total vehicle count, and the proportion of heavy vehicles is widely documented in the literature. Studies by Stephenson and Vulcan [74], Bodsworth and Lawrence [75], and, more recently, Alves Filho et al. [44] have highlighted the predominant role of vehicle count and fleet composition in determining final sound levels. Conversely, correlations among L_Aeq5, PM_2.5, and PAHs are less straightforward due to the diverse dispersion mechanisms affecting these pollutants. While noise levels decrease gradually with distance from the emission source [76] and are minimally influenced by meteorological conditions [77], PM_2.5 and PAHs are strongly influenced by such factors [78,79,80]. Consequently, no significant correlation was observed between L_Aeq5 and PAHs, despite their shared peak values in the same traffic categories. Moreover, fire events were found to significantly increase PAH concentration without affecting noise levels. Finally, the inverse correlation between PM_2.5 and PAHs, as discussed in Section 3.2, reflects their dominant sources, with PM_2.5 levels being largely driven by atmospheric transport mechanisms such as wind.

3.5. Risk Analysis

Mean BaP and BaPeq concentrations did not differ significantly among traffic categories (

Table 4. Mean BaP levels (±S.D.) and estimated additional cancer cases.

Traffic Category	BaP (ng/m3)		BaPeq (ng/m3)		Carcinogenic Risk (Total)		Carcinogenic Risk (per µg PM2.5)
LDT	0.14 ± 0.09	A	0.66 ± 0.39	A	5.70 ± 3.43	A	0.22 ± 0.17	A
MDT	0.16 ± 0.10	A	0.89 ± 0.52	A	7.77 ± 4.53	A	0.43 ± 0.26	B
HDT	0.12 ± 0.07	A	0.53 ± 0.34	A	4.65 ± 3.00	A	0.30 ± 0.19	B

Note: Categories with different capital letters differ significantly (p < 0.05). Carcinogenic risk is expressed as the number of additional cancer cases per 100,000 inhabitants over 70 years.

). All measured values complied with Directive (EU) 2024/2881 [81], which establishes 1 ng/m³ as the maximum allowable annual mean concentration. The highest concentrations were recorded toward the end of July and were closely associated with fire events, underscoring the compounded contribution of both vehicular emissions and biomass burning—two major BaP sources in the region [82].

When compared with other Latin American cities, BaP levels in Córdoba were slightly lower than those reported for Mexico City (0.283 ng/m³) [83] and La Plata, Argentina (0.381 ng/m³) [84], and considerably lower than those observed in Arequipa, Peru (1.9 ng/m³) [85], and Santiago, Chile (11.1 ng/m³) [86].

Carcinogenic potential based on BaPeq values likewise showed no statistically significant differences in cancer risk among traffic categories (Table 4). This pattern reflects the differing characteristics of the sites: MDT and HDT areas exhibited a higher proportion of PAHs per unit of PM_2.5, while LDT sites presented a higher total PM_2.5 mass. Although statistical differences were not evident, the magnitude of the estimated lifetime risk indicates that unit exposure to PM_2.5 (µg/m³) nearly doubles the additional cancer risk at MDT and HDT sites compared to LDT sites. Specifically, estimated additional cancer cases per 100,000 inhabitants over a 70-year lifetime increased by approximately 50% to nearly 100%. Furthermore, MDT and HDT sites are located predominantly in central areas with denser populations, implying greater potential public health impact. According to USEPA risk classification, the lifetime lung cancer risk (LLCR) remained within the “low” category across all traffic groups, with mean values of 4.3 × 10⁻⁵, 6.6 × 10⁻⁵, and 5.0 × 10⁻⁵ for HDT, MDT, and LDT sites, respectively (Figure 6).

The B(a)Peq levels associated with PM_2.5 measured in this study were comparable to those reported in Araraquara, southeastern Brazil, where median values ranged from 0.65 to 1.0 ng/m³ during the non-harvest season and from 1.2 to 1.4 ng/m³ during sugarcane harvesting [87]. In contrast, higher values have been documented in Arequipa, Peru, ranging from 1.2 to 2.8 ng/m³ in downtown and industrial areas [85]. Comparisons across cities should be interpreted with caution, as differences in urban structure, emission sources, meteorology, and sampling methodologies can substantially influence PAH concentrations. Additionally, the scarcity of available data for South American cities limits the ability to establish regional baselines.

4. Conclusions

This study evaluated the relationship between traffic density and three traffic-related pollutants—PM_2.5, PAHs, and noise—in Córdoba, Argentina, using a simple traffic-based classification applied to 20 monitoring sites. The results showed that PM_2.5 exhibited its highest mean concentrations at peripheral low-density traffic (LDT) sites, consistent with the presence of nearby bare-soil areas and greater wind-driven resuspension. In contrast, noise levels increased with traffic density and showed strong positive correlations with both total vehicle counts and the proportion of heavy vehicles.

Although total PAH concentrations did not differ significantly among categories, diagnostic ratios and molecular weight patterns indicated vehicular traffic—particularly diesel-powered engines—as the main source. Fire events further influenced PM_2.5 and PAH levels, amplifying potential health risks. Despite these variations, the lifetime lung cancer risk (LLCR) remained within the “low” category across all traffic classes.

Overall, our findings demonstrate that a simplified traffic density classification, informed by openly available Google Traffic data, can effectively identify pollution hotspots and capture key spatial patterns in noise and air pollutants. This low-cost framework is particularly valuable for cities lacking continuous monitoring infrastructure and provides a basis for prioritizing preventive actions and guiding future air quality assessments. Although this study did not include other key vehicular emission markers such as NO_x or black carbon, which would refine source attribution, the methodology developed here provides a robust basis for prioritizing preventive actions, identifying high-risk areas, and supporting future monitoring programs. This methodology also holds strong potential for adaptation in other Latin American cities facing similar data and infrastructure limitations.