Theoretical Analysis of Suspended Road Dust in Relation to Concrete Pavement Texture Characteristics

Abstract

Particulate matter (PM) originating from road dust is an increasing concern in urban air quality, particularly as non-exhaust emissions from tire–pavement interactions gain prominence. Existing models often focus on meteorological and traffic-related variables while oversimplifying pavement surface characteristics, limiting their applicability across diverse spatial and traffic conditions. This study investigates the influence of concrete pavement macrotexture—specifically the Mean Texture Depth (MTD) and surface wavelength—on PM₁₀ resuspension. Field data were collected using a vehicle-mounted DustTrak 8530 sensor following the TRAKER protocol, enabling real-time monitoring near the tire–pavement interface. A multivariable linear regression model was used to evaluate the effects of MTD, wavelength, and the interaction between silt loading (sL) and PM₁₀ content, achieving a high adjusted R² of 0.765. The surface wavelength and sL–PM₁₀ interaction were statistically significant (p < 0.01). The PM₁₀ concentrations increased with the MTD up to a threshold of approximately 1.4 mm, after which the trend plateaued. A short wavelength (<4 mm) resulted in 30–50% higher PM₁₀ emissions compared to a longer wavelength (>30 mm), likely due to enhanced air-pumping effects caused by more frequent aggregate contact. Among pavement types, Transverse Tining (T.Tining) exhibited the highest emissions due to its high MTD and short wavelength, whereas Exposed Aggregate Concrete Pavement (EACP) and the Next-Generation Concrete Surface (NGCS) showed lower emissions with a moderate MTD (1.0–1.4 mm) and longer wavelength. Mechanistically, a low MTD means there is a lack of sufficient voids for dust retention but generates less turbulence, producing moderate emissions. In contrast, a high MTD combined with a very short wavelength intensifies tire contact and localized air pumping, increasing emissions. Therefore, an intermediate MTD and moderate wavelength configuration appears optimal, balancing dust retention with minimized turbulence. These findings offer a texture-informed framework for integrating pavement surface characteristics into PM emission models, supporting sustainable and emission-conscious pavement design.

1. Introduction

Airborne particulate matter (PM), particularly fine particles smaller than 10 micrometers (PM₁₀), poses a significant threat to environmental quality and public health, contributing to respiratory diseases, cardiovascular issues, and neurological disorders [1]. In urban settings, road dust (RD) constitutes a major component of PM, derived from vehicle-related activities that produce both exhaust emissions (EE) and non-exhaust emissions (NEE), the latter arising from tire friction, brake wear, and surface abrasion [2,3]. Numerous regression-based models have been developed to estimate road dust emissions, typically emphasizing meteorological conditions and traffic volume as the primary predictors [4]. One prominent example, the AP-42 model by the U.S. Environmental Protection Agency (EPA), incorporates variables such as emission factors (E), vehicle weight (W), silt loading (sL), and particle size multipliers (K) for PM_2.5 and PM₁₀, relying on legacy datasets from the 1980s [5]. While subsequent modifications have accounted for rainfall frequency, they still omit critical factors like vehicle speed [6]. The HERMES model later addressed some limitations by integrating traffic volume, road length, and rainfall variability to improve spatial and temporal accuracy [7]. In European cities like Turin and Barcelona, additional empirical models introduced surface texture indicators such as the Corrected Aggregate Mode (CAM), derived from multiplying the Mean Texture Depth (MTD) by the mean aggregate size. These models underscored the influence of macrotexture geometry on road dust emissions [8]. Despite such advances, most current models inadequately reflect pavement surface conditions. Although extensions like NORTRIP have included variables such as the pavement material, tire type, and surface moisture [9], the direct impact of surface texture characteristics—namely, the texture depth and wavelength—on suspended PM₁₀ concentrations remains underexplored. Existing studies tend to focus on other tire–road interaction outcomes such as friction, noise, or wear, rather than directly examining resuspension mechanisms [10].

To bridge this gap, the present study employs a regression-based framework informed by field data to quantify the influence of the concrete pavement macrotexture—specifically the MTD and surface wavelength—on PM₁₀ resuspension. Real-time measurements were obtained using a vehicle-mounted DustTrak 8530 sensor operating under the TRAKER protocol. The model incorporates the MTD, surface wavelength, and an interaction term between sL and PM₁₀ concentration as explanatory variables. This design enables a robust analysis of how macrotexture affects non-exhaust emissions, extending the theoretical scope of traditional models such as AP-42 and NORTRIP. Recent research has also drawn attention to tire wear particles (TWPs) as an increasingly prominent source of non-exhaust PM, especially under dense urban traffic. These particles, generally composed of rubber-based materials, fall within the PM_2.5 and PM₁₀ size ranges and are generated not only by mechanical abrasion but also through aerodynamic mechanisms like air pumping during tire–pavement contact [11,12]. However, the influence of the pavement macrotexture—particularly the MTD and wavelength—on TWP resuspension remains poorly understood. By focusing on surface-induced airflow dynamics, this study aims to elucidate the role of texture parameters in mobilizing such particles. To this end, the model centers on two field-measurable macrotexture parameters: the MTD and Peak Number (used here as a proxy for the surface wavelength). The MTD represents the volumetric depth of surface voids and reflects the pavement’s dust-retention capability, while the Peak Number quantifies the frequency of aggregate exposure, influencing localized air turbulence. Both are cost-effective, interpretable, and scalable metrics suitable for real-world applications [13]. Although high-resolution technologies like laser profilometry and 3D scanning provide detailed texture data, they are limited by cost and practicality in large-scale field studies. In contrast, the Sand Patch Test (SPT), standardized under ASTM E965, offers a simple and accurate method for estimating the MTD based on volumetric texture voids relevant to PM retention and release [14].

Building on this foundation, the proposed regression model reveals that the surface texture substantially governs both suspension and resuspension processes. The shape and size of surface voids influence how dust accumulates and is subsequently dislodged through tire-induced air displacement. As tires compress over textured surfaces, air trapped within the voids is forcibly expelled, creating localized aerodynamic forces that lift fine particles into the atmosphere—a process independent of mechanical abrasion [15,16]. This mechanism is amplified by short wavelengths and deeper textures, which increase tire contact and turbulence. Although porous asphalt surfaces have shown promise in reducing emissions due to a higher void content, detailed quantification of surface geometry parameters remains necessary [17]. Accordingly, this study systematically analyzes how variations in MTD and surface wavelength across concrete pavement types correlate with measured PM₁₀ concentrations. By advancing a texture-informed modeling framework, this research contributes to both theoretical understanding and practical strategies for designing low-emission pavements in urban environments.

2. Materials and Methods

2.1. Study Area

To examine how the pavement surface texture influences suspended road dust emissions, field experiments were carried out at two designated test sites in South Korea. Site A, located at the Yeoncheon Social Overhead Capital (SOC) Center operated by the Construction Technology Research Institute, included pavement sections with Transverse Tining (T.Tining) and Exposed Aggregate Concrete Pavement (EACP) constructed using 10 mm aggregates. Site B, situated along the Yeoju test road maintained by the Korea Expressway Corporation, featured a wider variety of surfaces, including EACP constructed with 6 mm and 8 mm aggregates, additional T.Tining sections, and segments with Next-Generation Concrete Surface (NGCS).

As illustrated in Figure 1, data collection was conducted between June and October 2023—a period selected due to its relatively low and stable background PM₁₀ concentrations, thereby minimizing atmospheric interference. No major meteorological disturbances, including rainfall exceeding 0.01 inches, were recorded during this timeframe. Suspended PM₁₀ concentrations were measured in real time using a DustTrak 8530 optical particle counter (TSI Inc., Shoreview, MN, USA) equipped with a PM₁₀-selective inlet. Following the TRAKER protocol, the sensor was mounted 50 mm behind the right-front tire of the test vehicle to capture emissions directly induced by tire–pavement interaction. Data were recorded at one-second intervals while the vehicle maintained a constant speed of approximately 50 km/h under controlled driving conditions.

To isolate the effect of pavement texture from ambient environmental variables, background PM₁₀ levels were concurrently obtained from nearby urban air quality monitoring stations located within 1–3 kilometers of each test site. Additional meteorological parameters—including wind speed, temperature, humidity, and precipitation—were retrieved from the Automated Synoptic Observing System (ASOS) operated by the Korea Meteorological Administration, ensuring precise characterization of test conditions.

Throughout the measurement campaign, environmental conditions remained stable across all sites. Daily mean temperatures ranged from 20 °C to 28 °C, with relative humidity levels between 60% and 80%. Wind speeds were consistently low, averaging under 1.0 m/s, and total rainfall remained below 0.01 inches. To further eliminate the influence of transient external variables, supplemental measurements were conducted on closed road sections with restricted vehicle access. This controlled setup ensured that the observed PM₁₀ concentrations primarily reflected resuspension driven by pavement texture rather than meteorological or traffic-related fluctuations. As a result, the dataset supported a robust analysis of the relationships among suspended particulate levels, macrotexture parameters, and sL under consistent environmental conditions.

2.2. Measurement of Micro and Macro Surface Texture of Concrete Pavement

Concrete pavement surfaces exhibit a dual-scale texture structure, comprising both microtexture and macrotexture, as schematically illustrated in Figure 2. These texture components arise from a combination of material properties and construction methods, as outlined in the pavement surface classification framework developed by the National Cooperative Highway Research Program (NCHRP) [18]. In this study, microtexture refers to the fine-scale roughness formed by the angularity and surface characteristics of exposed coarse aggregates, whereas macrotexture denotes the larger-scale surface variations shaped by finishing techniques and the spatial arrangement of aggregate exposure.

Both texture levels were quantitatively assessed using standardized field measurements and image-based analysis procedures specifically adapted for concrete pavements. To further characterize macrotexture, surface profiles were decomposed into their constituent wavelength, capturing spatial variability within each pavement segment. The frequency of directional changes in the surface elevation—expressed as the number of inflection points—was measured as the Peak Number (PN), which served as a proxy for surface wavelength complexity [19]. Importantly, concrete pavements are not the result of passive wear but are purposefully engineered to exhibit desired texture characteristics through deliberate choices in mechanical treatment and aggregate configuration.

Surface texture was quantified using two primary indicators: macrotexture depth and surface wavelength, with measurements conducted across various concrete pavement sections within the designated test zones. Macrotexture depth was evaluated using the SPT, following the ASTM E 965 standard. In this method, 25 cm³ of standard-grade sand (retained on a No. 100 sieve) was evenly distributed over the pavement surface to form a uniform circular layer. The diameter (D) of the resulting sand patch was measured, and the MTD was calculated using the equation MTD = 4 V/πD², where V represents the known sand volume. The test setup is shown in Figure 3 [20].

To improve spatial representativeness, SPT measurements were conducted at 3-m intervals across the left, center, and right wheel paths of each traffic lane. The resulting MTD values from each transect were averaged to create a comprehensive texture profile for each pavement type. These averaged values were later integrated with wavelength analyses to construct composite surface characterizations. This multi-point sampling approach ensured consistency across heterogeneous pavement conditions and minimized localized measurement bias, thereby enhancing the reliability of the texture dataset.

The equation for calculating MTD is

MTD (mm) = 4V/πD²

(1)

V: Volume of sand (mm³).

D: Diameter of sand dispersed on the surface (mm).

To evaluate microtexture, this study used surface wavelength as a proxy indicator, derived from detailed mapping of aggregate exposure patterns. A custom-built imaging platform, featuring a step-motor-controlled industrial camera, captured high-resolution scans over a 100 cm × 60 cm pavement area. Each scanned section was subdivided into 24 uniform segments using proprietary image segmentation software. The system acquired multiple focal layers per segment, which were algorithmically composited to reconstruct a three-dimensional surface model that accurately captured depth-resolved aggregate contours (Figure 4) [21]. Optical distortion was minimized through fixed-angle camera alignment and near-field illumination.

Using these composite images, the Exposed Aggregate Number (EAN) was calculated by measuring the spatial frequency of aggregate emergence across the segmented surface. This metric was then used to estimate local wavelength variations. The resulting data provided a spatially continuous, exposure-based representation of the surface wavelength specifically tailored to the characteristics of concrete pavements under in situ field conditions.

To quantify surface wavelength, each pavement section image was first geometrically corrected to align with real-world dimensions. A high-resolution 300 mm longitudinal strip was then extracted from each image, and a 5 mm wide transect was analyzed to identify aggregate crest positions based on pixel intensity gradients. As shown in Figure 5, the spacing between prominent surface peaks along this transect was measured to compute the average texture wavelength. Although similar optical profiling techniques have been applied in previous surface texture studies [20,21], the current approach specifically focuses on aggregate-level geometric spacing within concrete pavement surfaces under actual field conditions. This method provides enhanced spatial resolution compared to conventional profilometers, yielding more precise characterization of surface wavelength.

2.3. sL and Suspended Road Dust Measurement

Suspended road dust comprises fine particles re-entrained into the atmosphere due to dynamic interactions between vehicle tires and the pavement surface. To quantify the amount of settled dust on pavement, this study employed the EPA AP-42 methodology, focusing on sL, g/m²—defined as the mass of particles smaller than 75 µm per unit surface area. Field sampling was conducted using the AP-42 C-1 protocol for road dust collection, while laboratory analysis followed the C-2 standard for particle size characterization [22]. These procedures enabled precise quantification of surface-retained particulates, specifically targeting the sub-75 µm fraction.

A mechanized vacuum system (TENANT V3E) equipped with a fine particulate filter (≥1.0 µm) was used in place of traditional manual sweeping, significantly improving collection efficiency. To ensure spatial and procedural consistency, the sampling method was refined in three ways: (1) samples were collected across all lateral segments of the traffic lane (left, center, and right wheel paths), (2) each 3-m pavement segment was vacuumed using 10 sequential passes, and (3) high-efficiency mechanical vacuuming was applied to maximize particulate capture and minimize loss. These refinements ensured reproducible and uniform dust sampling across varying surface conditions.

To assess airborne dust emissions, real-time PM₁₀ measurements were performed using a mobile sensing system mounted on a test vehicle. The system incorporated a laser-based optical particle counter (DustTrak DRX 8530, TSI Inc., USA) configured to collect PM₁₀ data at one-second intervals. The inlet was positioned 50 mm above the ground near the right front tire, directly capturing particles resuspended through tire–pavement contact. This setup, based on the TRAKER methodology, was adapted for enhanced spatial specificity, resembling the close-proximity principles of the CPX acoustic measurement protocol [23]. To prevent airflow distortion, the inlet was pressure-balanced using a vacuum pump calibrated to a matched flow profile. A data acquisition laptop recorded time-series PM₁₀ values, while a dual-view camera system monitored vehicle speed and surrounding environmental conditions, enabling synchronized cross-verification between particulate readings and physical context [24]. Two complementary metrics were used to assess surface dust potential and actual resuspension: (1) sL as the indicator of surface-bound particulates, and (2) PM₁₀ concentration as the airborne outcome. Their relationship was further analyzed through the PM₁₀ content (%), defined as the proportion of <10 µm particles within each silt sample. This derived index linked the fine particle fraction on the pavement to measured suspension levels during test runs [25]. Statistical analysis of sL and PM₁₀ distributions across test sites was conducted using the Interquartile Range (IQR) method, identifying outliers as values below Q1 − 1.5 × IQR or above Q3 + 1.5 × IQR. To determine the fine particle composition in silt samples (sL×PM₁₀), a Helos & Rodos laser diffraction analyzer (Sympatec., GmbH, Clausthal-Zellerfeld, Germany) was employed to measure particle size fractions below 10 µm and 70 µm. Particles under 75 µm were used to compute sL, while those smaller than 10 µm were categorized as PM₁₀. This integrated analysis enabled detailed correlation between surface dust characteristics and real-time resuspension behavior.

2.4. Assumption

Building on this understanding, the present study hypothesizes, as shown in Figure 6, that concrete pavement texture—defined by its macrotexture depth and surface wavelength—plays a central role in modulating resuspension through mechanical air displacement.

At a vehicle speed of approximately 50 km/h, which reflects the test conditions used in this study, tire–pavement interactions are expected to induce rapid compression and decompression within surface voids. This phenomenon, known as air pumping, generates upward airflow capable of lifting particles smaller than 10 µm into the atmosphere. Research has shown that the aerodynamic wake produced by rolling tires can mobilize fine particles from the pavement surface, with near-surface wind speeds reaching up to 2.2 m/s under typical driving conditions [12]. To translate this physical mechanism into a quantitative framework, a regression-based model was developed to relate suspended PM₁₀ concentrations to texture-specific variables. The model posits that increased texture depth and shorter surface wavelength intensify microscale aerodynamic disturbances at the tire–pavement interface, thereby elevating the likelihood of particle resuspension. Multi-variable linear regression analysis using field data confirmed that pavement texture geometry is a statistically significant predictor of PM₁₀ emissions. The resulting model offers a physically grounded and empirically supported explanation of texture-induced emission behavior, with direct implications for sustainable pavement design and particulate pollution mitigation

3. Results

3.1. Correlation Between sL and Suspended Road Dust for Validation

A field analysis of the asphalt-paved section (designated as Site C) revealed substantial spatial variation in PM₁₀ concentrations, corresponding to differences in measured sL. All measurements were conducted under uniform driving conditions at a constant speed of 50 km/h, in accordance with standardized dust monitoring protocols [25,26]. As depicted in Figure 7, a strong positive correlation (r² = 0.76) was observed between the average sL and suspended PM₁₀ concentrations across the Site C transects. The derived regression equation, y = 267.33x + 128.12, further quantifies the magnitude of this relationship, indicating that for every 0.1 g/m² increase in sL, the PM₁₀ concentration increases by approximately 26.7 µg/m³. These results underscore that reducing sL via surface cleaning or material control could serve as an effective strategy for mitigating road dust emissions.

This trend confirms that sL serves as a primary predictor of dust resuspension potential on road pavements. The observed linearity between surface-bound silt mass and airborne PM₁₀ concentrations reinforces the conceptual framework wherein sL functions as an emission reservoir embedded within the pavement surface. The reliability of this relationship is further supported by the measurement methodology employed in this study, which utilized a vehicle-mounted, real-time sensor system capable of capturing dynamic resuspension events during actual tire–pavement interaction. This mobile platform effectively minimizes external noise and ensures that measured concentrations accurately reflect texture-driven dust behavior. Although the empirical relationship between sL and PM₁₀ levels is well-documented, its interaction with pavement surface texture remains underexplored. Since sL is not uniformly distributed across the pavement but is affected by localized surface features, the spatial variability of texture characteristics likely modulates both the retention and resuspension of particulate matter. In this context, sL should be interpreted not only as a surface-level indicator of dust load but also as a texture-sensitive parameter. The findings of this study provide a quantitative foundation for incorporating sL as a co-variable in future multi-factor models of texture-induced resuspension, particularly in frameworks designed to spatially map or predict dust emissions at the sub-lane scale.

3.2. Data Eligibility Criteria for Analyzing Suspended Road Dust and Surface Texture Parameters

The road dust emission monitoring system used in this study simultaneously measured sL and airborne dust concentrations via DustTrak sensors mounted on a mobile laboratory vehicle. These two data streams were correlated to evaluate the relationship between surface-bound dust and resuspended PM₁₀ levels. However, during field measurements, the process of collecting sL—particularly through vacuum suction—may inadvertently reduce the amount of surface dust available for resuspension. This can result in artificially low PM₁₀ concentrations that fall outside the valid operational range of the monitoring system, potentially distorting the correlation analysis.

To mitigate this issue and improve data reliability, a rigorous data screening protocol was applied. Measurement sets were reviewed to ensure they represented conditions prior to any significant disruption of the pavement surface.

Table 1. Comparative results of suspended road dust concentrations before and after sL sampling using the vacuum method.

Before sL
Location	Vehicle Speed	Road Dust (RD)	Background Dust (BD)	Resuspended Dust (RD-BD)
Site C	30 km/h	114.014	97.252	16.719
	40 km/h	122.989	95.792	27.197
	50 km/h	148.3	93.033	55.267
After sL
Location	Vehicle Speed	Road Dust (RD)	Background Dust (BD)	Resuspended Dust (RD-BD)
Site C	30 km/h	124.223	122.14	2.083
	40 km/h	125.545	122.306	3.239
	50 km/h	128.75	122.889	5.861
Mean Reduction in Suspended Dust Concentration After sL Sampling
Location	Vehicle Speed	Before	After	Mean Reduction Rate (%)
Site C	30 km/h	16.719	2.083	87.55
	40 km/h	27.197	3.239	88.09
	50 km/h	55.267	5.861	89.41
	Average reduction			88.35

presents five repeated measurements of suspended dust concentrations taken before and after sL sampling, recorded at vehicle speeds ranging from 30 to 50 km/h across various test segments. At Site C, average PM₁₀ concentrations before sL collection were 16.719, 27.197, and 55.267 μg/m³. Following the sampling process, these values declined markedly—by an average of 88.35%—to 2.083, 3.239, and 5.861 μg/m³, respectively.

This sharp decline in post-sampling PM₁₀ levels illustrates that such data may not accurately reflect natural dust emission conditions. To preserve analytical integrity, only measurements recorded prior to sL sampling were used to evaluate the relationship between sL, surface texture, and road dust suspension. This approach ensures that the resulting correlations reflect the undisturbed interaction between pavement conditions and suspended particulate matter. Based on these initial findings, the dataset was restructured to enhance the consistency and efficiency of the emission evaluation process. As shown in Figure 8, standardized data processing protocols were applied to the calculation of sL, and the overall data acquisition workflow was organized into four sequential stages to minimize data loss and maintain methodological consistency:

• Suspended Dust Measurement: PM₁₀ concentrations were recorded prior to sL sampling. If a measured value was lower than the background concentration—indicating potential interference from nearby vehicles or environmental sources—the corresponding data point was excluded.
• sL Measurement: sL was then quantified using the vacuum-based sampling method described earlier to determine the mass of deposited dust per unit area on the pavement.
• Road Surface Texture Measurement: After dust collection, the surface texture depth and wavelength were measured to characterize the pavement texture under low dust-loading conditions.
• Data Filtering: When averaging suspended dust concentrations for a given road segment, only stable measurements collected over approximately 10 seconds of non-accelerating motion were used. Data collected during turns or immediately following a U-turn were excluded based on road alignment.

For sL analysis, dust was collected over a 3-m segment using an enhanced sampling protocol designed to directly link suspended dust levels with localized sL values. This 3-m section consisted of ten successive 0.3-m samples, in alignment with the EPA’s industrial vacuum inlet specifications. A minimum of 10 sampling repetitions was performed to ensure data reliability and consistency in surface dust quantification [25,26]. Texture characterization involved two complementary methods. The average surface wavelength was computed by analyzing aggregate exposure patterns across concrete pavement types using high-resolution imaging techniques. The MTD was measured via the SPT. To ensure spatial representativeness, approximately 100 measurements were taken along each 100-m test section, with the front and rear portions independently evaluated to minimize the influence of continuity effects in extended roadway segments.

3.3. MTD and Average Wavelength Results for Different Types of Concrete Pavement

As shown in

Table 2. Concrete surface texture measurement results on the test road sections.

Measurement Section	Type of Pavement	Average Wavelength (mm) (Standard Deviation)	Average Mean Texture Depth (mm) (Standard Deviation)
Site A	EACP (10 mm)	5.3 (0.08)	1.43 (0.2)
	T.Tining	30 (0.01)	1.14 (0.19)
Site B	EACP (6 mm)	3.4 (0.06)	1.23 (0.11)
	EACP (8 mm)	3.7 (0.06)	1.27 (0.15)
	NGCS	4.5 (0.05)	1.4 (0.18)
	T.Tining	30 (0.01)	0.84 (0.1)

, the T.Tining treatment applied to the concrete pavement was originally designed to produce a surface wavelength of 30 mm and a texture depth of 3 mm per groove. However, field measurements indicated that while the intended wavelength remained consistent at 30 mm, the actual texture depth diminished to 1.14 mm, attributed to reduced aggregate exposure and surface wear over time. For the EACP 10 mm pavement, the measured wavelength was 5.3 mm, with an MTD of 1.43 mm.

At Site B, the T.Tining surface was likewise intended to maintain a wavelength of 30 mm and a depth of 3 mm, but surface abrasion led to a significantly reduced texture depth of 0.84 mm—closely matching the MTD and average wavelength observed at Site A (1.14 mm). In the case of the EACP (8 mm) section at Site B, the surface wavelength was measured at 3.7 mm, and the MTD was 1.27 mm. These results reflect the impact of long-term surface degradation on texture geometry and underscore the need for ongoing monitoring to accurately evaluate pavement-induced emission behavior.

Suspended road dust concentrations were measured using a TRAKER-based system, with the test vehicle traveling at a constant speed of 50 km/h and the sampling inlet mounted directly on the vehicle. Background PM₁₀ levels during the measurement period were obtained from the nearest air quality observatory, supplemented by on-site measurements using DustTrak equipment [27]. As presented in

Table 3. Road dust concentration results for concrete pavement sections on the test road.

Measurement Section	Type of Pavement	sL (mg/m2)	Suspended Road Dust (µg/m3)		PM10 Contents (%)	Background Dust (OB)	Background Dust (TR)
Site A	EACP (10 mm)	44.4	255.16		2.1615	61	68.01
	T.Tining	80	306		3.775	61	67.99
Site B	EACP (6 mm)	152.3	192.35		4.1245	62	64.25
	EACP (8 mm)	200.6	167.7		4.1345	62	64.25
	NGCS	151.5	172.16		3.9335	62	64.11
	T.Tining	50.1	277.29		1.8830	65	70.28
OB: Background concentration measured by the observatory closest to the test site TR: PM10 concentration measured by the TRAKER-mounted vehicle during testing
Sampling locations				sL collected Specimens
Exposed aggregate concrete pavement (10 mm)
Next-generation concrete pavement
Transverse tining concrete pavement

, background concentrations recorded by the DustTrak were consistently higher—by approximately 2 to 7 µg/m³—compared to those reported by the observatory, highlighting the added sensitivity of the mobile monitoring system to localized particulate conditions.

This discrepancy is primarily attributed to the proximity of the TRAKER-based DustTrak system, which measures particulate concentrations directly above the pavement surface—where resuspension caused by tire–pavement interaction is most pronounced. In contrast, observatory data are collected at greater distances and elevated positions, capturing broader ambient background conditions. Rather than indicating measurement bias, the elevated DustTrak readings are interpreted as localized indicators of surface texture effects and tire-induced turbulence. These near-surface readings offer a more accurate depiction of the micro-scale resuspension dynamics driven by pavement geometry and vehicle movement. This methodological distinction underscores the value of proximity-based sensors like DustTrak in assessing how pavement surface characteristics influence non-exhaust emissions. Based on resuspended dust resulting from tire–road interaction, the dust loads for T.Tining and EACP were measured at 44.4 mg/m² and 80 mg/m², respectively. The average PM₁₀ content was 2.16% for T.Tining and 3.775% for EACP, suggesting the importance of stratified analysis by surface type to better isolate the impact of specific pavement features.

Dust emission factors and resuspended particulate loads were also measured over 100-m sections of NGCS, T.Tining, and EACP pavements at Site B, with all tests conducted at a driving speed of 50 km/h. The highest suspended dust concentration was observed for T.Tining at 277.29 µg/m³, followed by EACP at 167.7 µg/m³. For EACP with 6 mm aggregate, sL was recorded at 192.35 mg/m² with a PM₁₀ content of 4.1245%, while NGCS showed similar values of 172.16 mg/m² and 3.93%, respectively. Notably, although T.Tining at Site B exhibited a comparable macrotexture depth and wavelength, its sL and PM₁₀ content were significantly lower, at 50.1 mg/m² and 1.8830%, respectively. These findings further highlight the complex interplay between surface texture, aggregate structure, and the mechanisms driving dust resuspension.

3.4. Analysis of Dust Generation Based on MTD

To quantify suspended dust concentrations in relation to surface texture parameters, variations in dust levels across different texture classifications were analyzed, as summarized in

Table 4. Classification-based analysis using MTD, sL, and surface wavelength levels.

Factors	Low	Middle	High
MTD (mm)	Under 1	1~1.4	More than 1.4
Silt loading (g/m2)	Under 0.06	0.06~0.15	More than 0.15
Wavelength (mm)	Under 4	4~30	30

. This analysis enabled the identification of emission trends associated with specific surface characteristics, particularly MTD. Assuming constant dust concentrations for each MTD category, the relationship between MTD and suspended particulate matter was assessed to propose MTD-based control strategies for dust mitigation.

MTD values were classified into three categories: low (<1.0 mm), middle (1.1–1.4 mm), and high (>1.4 mm). This classification was informed by prior domestic studies indicating that 13 mm EACP with 8 mm aggregate typically yields MTDs in the range of 1.4–2.0 mm, while 10 mm EACP exhibits a benchmark MTD of around 1.0 mm [28,29]. For sL, middle and high levels were defined as values exceeding 0.05 g/m², reflecting typical urban road dust conditions in South Korea [30]. In terms of surface wavelength, EACP generally presented values under 4 mm, whereas T.Tining demonstrated longer wavelengths near 30 mm. These thresholds were used to construct a three-tier classification framework for texture parameters. The resulting system provides a basis for future research and practical strategies aimed at managing concrete pavement surfaces to reduce road dust emissions and improve environmental performance.

Figure 9 through 14 present suspended dust concentration patterns across different MTD categories under low-, medium-, and high-sL conditions. The objective of this analysis is to determine whether an inflection point exists—indicating that beyond a certain MTD threshold, increases in texture depth no longer correspond to proportional increases in airborne dust levels. This would suggest that once sL reaches a critical level where readily resuspendable particles are depleted, additional surface voids created by deeper texture no longer significantly impact dust emission behavior. Consequently, managing texture depth in relation to sL becomes a key factor in effective emission control. In these figures, the legend is organized from left to right to represent low, medium, and high MTD levels, respectively.

As illustrated in Figure 9, the type of surface treatment appears to influence dust resuspension as much as, if not more than, the texture depth itself. A correlation analysis was conducted between the concrete pavement texture depth and suspended dust concentrations while also considering the role of surface treatment methods.

For T.Tining, texture is formed through transverse grooves, which alter air dynamics at the tire–pavement interface. Despite comparable or even lower sL and PM₁₀ content relative to EACP, T.Tining consistently exhibited higher suspended dust concentrations. This discrepancy is attributed to the intensified air-pumping effect associated with its grooved structure. As the tire compresses over the surface, air trapped within the grooves is rapidly expelled—creating localized airflow speeds of up to 2.2 m/s. This mechanism significantly increases the resuspension potential of deposited particles, elevating airborne PM₁₀ levels regardless of nominal texture depth μ.

As shown in Figure 10, suspended dust concentrations on concrete pavement generally increased with a rising MTD. However, under middle-range sL conditions (0.06–0.15 g/m²), dust levels on EACP pavement began to plateau once MTD exceeded approximately 1.4 mm. This suggests the presence of a potential saturation threshold, beyond which additional increases in texture depth no longer contribute meaningfully to resuspension. The result supports the hypothesis that, under moderate dust-loading conditions, the capacity of the pavement surface to release particles may become limited, emphasizing the importance of managing both sL and MTD in tandem for effective dust emission control.

This pattern indicates the existence of a threshold texture depth beyond which further increases in MTD do not significantly elevate PM₁₀ emissions. Specifically, the 1.4 mm depth appears to represent an optimal balance between dust retention and resuspension under the test vehicle’s operating speed of 50 km/h. The observed plateau likely results from the saturation of the air-pumping mechanism, where the aerodynamic uplift generated at the tire–pavement interface reaches a functional limit. These findings suggest that maintaining the MTD near this inflection point can serve as an effective strategy for controlling dust emissions on EACP-type concrete surfaces, particularly under moderate sL conditions.

As illustrated in Figure 11, a similar trend was observed under high-sL conditions (0.15–0.25 g/m²), where suspended dust concentrations on EACP pavement stabilized once MTD reached approximately 1.4 mm. This suggests that accumulated surface dust becomes effectively contained within the pavement’s texture voids at this depth, limiting further resuspension despite higher overall dust loads. In this sL range, resuspension behavior appears to be governed more by pavement geometry than by incremental increases in silt accumulation.

While PM₁₀ content remained below 5%, as reported in Table 3, its influence on airborne dust levels was not statistically significant under high-sL conditions. Instead, sL itself emerged as the dominant factor influencing PM₁₀ concentrations, reinforcing the conclusion that emission levels are more sensitive to the total quantity of surface dust than to the proportion of fine particles in this range [25]. These findings further validate 1.4 mm as a practical MTD threshold for managing emissions on EACP pavements in scenarios of elevated dust accumulation.

As shown in Figure 12, suspended dust concentrations were significantly higher on T.Tining pavements compared to other concrete surfaces, even under similar sL conditions. This is likely due to the increased spatial voids associated with high-MTD surfaces, which amplify localized airflow and enhance dust mobilization. The concentrated transverse grooves characteristic of T.Tining reduce the effective contact area while intensifying air displacement, thereby increasing PM₁₀ emissions despite identical sL levels.

To further isolate the influence of surface geometry, pavements with similar sL and MTD values were grouped for separate analysis, excluding T.Tining sections. Under conditions of a short surface wavelength (<4 mm) and moderate sL (0.06–0.15 g/m²), suspended dust concentrations consistently increased with a higher texture depth. Notably, the difference in PM₁₀ concentrations between the low and high MTD groups exceeded 100 µg/m³. Since the PM₁₀ content remained relatively constant across these samples, its impact was considered negligible in this comparison. These results confirm that, under controlled conditions of wavelength and dust load, MTD is the primary factor influencing dust resuspension on concrete pavements.

Figure 13 illustrates the relationship between suspended dust concentrations and the MTD under high-sL > 0.15 g/m² conditions. The analysis reveals a strong positive correlation (R² = 0.7323), indicating that increases in MTD are associated with significant rises in suspended dust levels. The difference in PM₁₀ concentrations between low and high MTD categories exceeded 100 µg/m³, underscoring the pronounced influence of the surface texture depth on dust resuspension.

These findings support the hypothesis that deeper textures enhance dust mobilization by expanding the surface void volume, which in turn intensifies the air-pumping effect during tire–pavement interaction. As tires compress and release over textured surfaces, a greater MTD facilitates stronger localized airflow, enabling more particles to become airborne. Consequently, reducing the MTD may serve as an effective emission control strategy in high-sL environments by limiting the aerodynamic uplift that drives particulate resuspension.

As shown in Figure 14, suspended dust concentrations were significantly influenced by the surface wavelength, even when sL and MTD were held constant. Specifically, concrete pavements with shorter wavelengths exhibited higher dust concentrations than those with longer wavelengths. This pattern suggests that a reduced contact area—characteristic of shorter wavelengths—increases the intensity of the air-pumping effect, thereby promoting greater dust resuspension. In contrast, longer wavelengths offer a broader tire–pavement contact surface that disperses compressed air, reducing localized aerodynamic force and resulting in lower PM₁₀ emissions. Although this study did not include direct noise-level measurements, previous acoustic research supports the proposed mechanism. Studies have demonstrated that shorter wavelengths and narrower aggregate spacing elevate contact frequency and generate higher acoustic pressure peaks [29]. These conditions are also associated with increased localized airflow disturbances at the tire–pavement interface, which enhance particulate resuspension. In this study, the surface wavelength was quantified through image-based analysis and used as a geometric proxy for aggregate spacing and contact frequency.

Based on these findings, shorter surface wavelengths—particularly those resembling the artificial groove pattern of T.Tining pavement (approximately 30 mm)—may intensify air pumping despite their benefits in surface drainage. As a result, T.Tining may be inherently more susceptible to suspended dust generation than pavement types with longer wavelength profiles. These results underscore the importance of incorporating surface wavelength as a critical design parameter in strategies aimed at managing non-exhaust particulate emissions from concrete pavements.

Based on the analyses presented in Figure 9 through 14, suspended dust concentrations were found to be governed by the combined effects of MTD, sL, and surface wavelength. In general, deeper textures and higher sL levels contributed to increased dust resuspension; however, a threshold depth of approximately 1.4 mm—particularly for EACP pavements—was identified, beyond which suspended PM₁₀ levels began to stabilize. Notably, T.Tining surfaces exhibited consistently higher PM₁₀ concentrations despite having comparable or lower sL and MTD values. This outcome is attributed to the artificial groove patterns of T.Tining, which intensify the air-pumping effect and promote greater particulate uplift. Additionally, a shorter surface wavelength was associated with higher dust concentrations under identical MTD and sL conditions, emphasizing the role of tire–surface contact dynamics. These findings underscore the importance of optimizing pavement texture by maintaining MTD within functional limits and selecting surface treatments that promote longer wavelengths. Such design considerations offer a practical pathway for reducing non-exhaust particulate emissions from concrete pavements.

3.5. Results of sL According to MTD

When analyzing the relationship between sL and suspended road dust on EACP and NGCS concrete pavements, the data were modeled using nonlinear regression functions incorporating threshold points defined by texture depth categories. As illustrated in Figure 15, the regression curves exhibit distinct inflection points corresponding to each MTD group, reflecting how varying levels of sL interact with the surface texture. Specifically, the identified threshold MTD values—approximately 0.8 mm for the low group, 1.3 mm for the middle group, and 1.6 mm for the high group—represent the minimum depths beyond which increases in sL begin to significantly influence dust resuspension.

The shape of the curves suggests that deeper textures are more responsive to increased sL, amplifying dust emission potential when sufficient particulate accumulation exists. In contrast, shallower textures tend to dampen this effect, likely due to a reduced void volume and lower air-pumping capacity. These findings highlight the nonlinear dynamics of texture–dust interactions and reinforce the importance of depth-sensitive design thresholds in managing particulate emissions from concrete pavements.

Suspended road dust concentrations on concrete pavements generally increased with a greater surface texture depth. In low-MTD sections, sL was typically low, contributing to reduced PM₁₀ levels. However, this trend may not hold when low MTD is accompanied by a short wavelength and high dust content, which can still promote resuspension. In contrast, pavements with higher MTD values retained more silt, resulting in greater PM₁₀ emissions. This pattern suggests that deeper surface voids enhance air pumping during tire–pavement contact, facilitating the uplift of fine particles into the atmosphere. However, the relationship between MTD and resuspension does not apply uniformly across all pavement types.

As shown in Figure 16, the relationship between sL and suspended PM₁₀ varies markedly across different MTD groups. Under low-MTD conditions, even when sL increases, the broader contact area appears to suppress the air-pumping effect, resulting in a logarithmic decrease in suspended dust (R² = 0.7642). This implies that smoother surfaces with lower MTD reduce resuspension due to enhanced tire–surface contact and diminished airflow generation. Conversely, in high-MTD sections (>1.4 mm), especially on T.Tining pavements, the regression slope flattened considerably (R² = 0.1422), indicating that PM₁₀ concentrations became less responsive to texture depth and more influenced by silt accumulation. This is attributed to the unique groove-based design of T.Tining, where 3 mm-deep transverse grooves reduce the contact area and concentrate airflow within confined voids. As the tire compresses and releases over these grooves, strong localized airflow is generated, intensifying the air-pumping mechanism and producing consistently elevated dust levels even at similar sL values.

These findings highlight a key distinction between concrete and asphalt pavements. In asphalt, surface texture is defined by the natural arrangement and depth of aggregates, which tend to generate a relatively uniform air-pumping effect across different MTDs. In contrast, concrete textures—particularly those created through mechanical grooving—are structurally optimized for drainage or friction but may inadvertently intensify dust resuspension. T.Tining, with its deep, narrow grooves, exemplifies this effect by amplifying air expulsion and reducing the contact area. While low-MTD concrete surfaces tend to suppress resuspension due to increased contact and lower turbulence, asphalt pavements with smoother textures at a low MTD may actually exhibit higher dust emissions due to greater particle mobility. These results suggest that the intensity of the air-pumping mechanism—not texture depth alone—is the dominant factor governing dust resuspension on concrete pavements. Effective dust mitigation strategies must therefore focus on optimizing groove geometry and controlling silt accumulation, especially on surfaces like T.Tining that are structurally predisposed to a high emission potential.

A related study reported that road dust emissions can be mitigated through targeted pavement design and maintenance strategies [31]. In terms of particle size distribution, the findings of this study for concrete pavements are broadly consistent with those previously reported for asphalt surfaces. In both pavement types, particles generated by surface abrasion were predominantly smaller than 0.1 mm. Specifically, particle size distributions ranged from 12.8% to 3.4% for asphalt and from 12.0% to 6.5% for concrete. Additionally, the proportion of particulate matter (PM) within the total road dust mass was slightly higher for concrete pavements (9.5%) than for asphalt (9.3%). This subtle yet consistent difference may be attributed to the more uniform surface geometry of concrete, which facilitates the concentrated deposition of fine particles and increases their vulnerability to aerodynamic uplift. While concrete generally offers greater structural durability under traffic loading, asphalt surfaces are more susceptible to mechanical degradation due to frequent braking and thermal stress. In pavement sections with coarser aggregates and pronounced surface textures, concrete pavements exhibited comparable—or in some cases higher—levels of PM₁₀ concentration and total dust load than asphalt pavements subjected to significant wear and enriched with fine particulate content (<10 µm) [32]. These results reinforce the conclusion that, under specific geometric and material conditions, concrete pavements can retain more dust and pose a greater risk of particulate resuspension compared to asphalt surfaces.

3.6. Results of Suspended Road Dust Across Different Concrete Pavement Types

The results of suspended road dust across different concrete pavement types are shown in Figure 17, where sL values on concrete pavements consistently exceeded 0.05 g/m² across all MTD categories. This level is comparable to values reported for asphalt-paved urban roads in the Seoul metropolitan area [30], indicating that concrete surfaces retain similar or even greater amounts of fine particulate material. These findings are consistent with prior studies showing that the proportion of particulate matter within the total road dust mass is slightly higher on concrete pavements than on asphalt. A plausible explanation lies in the relatively uniform surface geometry of concrete, which facilitates the even deposition of fine particles and increases their susceptibility to aerodynamic resuspension.

Although concrete generally exhibits higher structural durability under repeated traffic loading, asphalt pavements are more prone to mechanical degradation due to stronger braking forces and greater thermal deformation. In sections with similar aggregate exposure, concrete surfaces may demonstrate higher PM₁₀ levels due to more consistent particle retention and more efficient air pumping. This comparison reinforces the importance of recognizing material-specific differences in dust accumulation and release mechanisms. Therefore, strategies for mitigating road dust emissions should consider not only surface texture parameters but also the intrinsic dust retention characteristics associated with each pavement type. Tailoring surface treatments and maintenance practices to account for these material-dependent behaviors will be essential for designing effective and sustainable dust control interventions.

Based on the comparative analysis presented in

Table 5. Underlying mechanism of suspended road dust generation across different concrete pavement types.

Type of Pavement	Dust Load Potential/Results	Suspended Dust Potential/Results	Principle and Discussion
EACP (Removal of surface mortar from randomly exposed aggregates to surfaces)	Low/Low	Low/Low	Surface wear may increase the accumulation of dust; however, the random exposure of coarse aggregates on the pavement can disrupt airflow continuity and reduce the air-pumping pressure generated between the tire and the surface, thereby lowering the amount of dust
NGCS (Formation of texture via grooving treatment after surface treatment)	Middle/Middle	Middle/Middle	It forms a groove with a certain standard, but the depth is shallower than T.tining, which may affect road dust collection. The road dust concentration can be lower than T.Tining
T.Tining (Horizontal texture with arbitrary lines spaced laterally)	High/Irregularity	Low/High	Although it forms a groove of a certain standard, it may not affect the collection of road dust using the Vacuum Sweep Method, and air-pumping generates high levels of dust

, the mechanisms and trends of suspended dust generation were found to differ significantly between asphalt and concrete pavements. These differences are closely linked to each material’s MTD and aggregate size characteristics. In particular, sL and the resulting PM₁₀ concentrations exhibited distinct behaviors depending on surface structure, with concrete pavements generally demonstrating greater dust retention and resuspension potential under similar sL conditions. Despite these findings, the role of the pavement surface texture—particularly the distinction between microtexture and macrotexture—remains insufficiently quantified in terms of its influence on particulate resuspension.

To address this gap, a schematic framework was developed to categorize the texture-induced mechanisms contributing to dust emissions. While rough-textured pavements are known to improve skid resistance and vehicular safety through enhanced friction, further research is needed to understand how surface properties interact with environmental factors such as moisture, wind, and particle size to affect dust generation and uplift. Among all surface types analyzed, T.Tining exhibited the most pronounced artificial grooving, with a spacing of approximately 3 mm. This design intensified the air-pumping effect during tire–pavement interaction and contributed to the highest observed PM₁₀ concentrations under high-sL conditions. These results underscore the importance of not only material selection but also groove geometry and texture design in managing road dust emissions.

The model exhibited strong explanatory power, with an adjusted R² of 0.765 and an overall F-statistic of 80.6 (p < 0.01), confirming high statistical significance described in

Table 6. Correlation analysis results between suspended road dust and influencing factors.

Correlation Analysis Results Between Suspended Road Dust and Influencing Factors
Right MTD	3 Points MTD		Wavelength	sL		PM10 Contents
−0.287	−0.300		0.840	−0.750		−0.510
Results of Multi-regression Analysis between suspended road dust
Adjust R2		F			p-value
0.765		80.6			<0.01
Category	Coefficient		Standard error	t-statistics		p-value
y-intercept	180.177		24.819	7.260		<0.001
Right MTD	19.122		13.730	1.393		0.047
Wavelength	3.532		0.410	8.617		<0.001
sL * PM10 (Contents of PM10 in sL)	48.954		12.162	4.025		<0.01

. At the individual predictor level, all included variables demonstrated p-values ≤ 0.05, indicating statistical significance at the 95% confidence level. Among these, the surface wavelength (t = 8.617, p < 0.001) and the interaction between sL and PM₁₀ content (t = 4.025, p < 0.01) emerged as the most influential factors driving suspended dust concentrations. While the coefficient for the Right MTD was positive (β = 19.122, p = 0.047), its t-statistic (1.393) did not meet the conventional threshold of |t| ≥ 2 for strong individual significance. This may reflect the localized nature of texture depth influence or variability in conditions specific to the right wheel path when measuring suspended road dust using the TRAKER method. Nonetheless, the positive direction of the coefficient indicates a possible trend toward increased dust resuspension with greater texture depth, warranting further investigation into spatial variability across pavement lanes.

4. Discussion

This study empirically demonstrated that the concrete pavement texture—specifically MTD and surface wavelength—plays a significant role in the generation of suspended road dust. Conventional emission models such as AP-42, NORTRIP, and various regional frameworks primarily emphasize meteorological factors and traffic intensity, often oversimplifying pavement surface conditions as categorical inputs. Such limitations hinder the accurate modeling of aerodynamic processes occurring at the tire–pavement interface. The results of this study indicate that suspended dust concentrations increase with a greater MTD, although the trend plateaus beyond approximately 1.4 mm. This saturation suggests a threshold in the air-pumping mechanism, where further increases in texture depth no longer contribute meaningfully to resuspension. Air pumping, which results from the compression and release of air between tire treads and textured surfaces, generates localized upward airflow sufficient to dislodge fine particles [15,16]. The artificially grooved structure of T.Tining amplifies this effect, explaining its consistently higher PM₁₀ levels despite a comparable silt loading (sL). In contrast, surfaces such as EACP and NGCS, which feature longer wavelengths and moderate texture depths, showed lower suspended dust concentrations. These textures reduce contact frequency and local turbulence, thereby weakening the air-pumping effect. Our findings support the hypothesis that shorter wavelengths enhance tire contact frequency and micro-scale aerodynamic disturbances, leading to greater dust resuspension. This is consistent with acoustic studies reporting increased peak noise levels on shorter-wavelength surfaces due to more frequent tire–surface interaction [29,30].

Figure 18 compares the relationship between MTD and suspended PM₁₀ concentrations for asphalt and concrete pavements. The results show that concrete surfaces exhibit a steeper increase in resuspended dust as MTD increases, while asphalt surfaces tend to plateau. This comparative trend reinforces the unique behavior of concrete in amplifying PM₁₀ emissions through a greater surface depth. While asphalt surfaces often reach a saturation point beyond which the MTD has limited impact, concrete pavements maintain a positive slope, consistent with the regression model findings that identified the MTD as a statistically significant predictor of PM₁₀ concentrations (p = 0.047). These differences likely stem from the more rigid and uniform structure of concrete surfaces, which intensifies air expulsion. This distinction is rooted in the air-pumping mechanism, a central driver of dust resuspension on textured pavements. As a tire rolls across the surface, air trapped in voids is rapidly expelled, generating localized aerodynamic forces capable of lifting fine particles—particularly those smaller than 10 µm. These texture-induced aerodynamic effects may also contribute to the mobilization of TWPs, as recent studies suggest that both air pumping and contact frequency influence TWP resuspension. In our field measurements, concrete pavements—especially T.Tining—exhibited more rigid and consistently grooved textures, intensifying air displacement and elevating PM₁₀ levels. In contrast, asphalt pavements presented more porous and irregular textures, which appeared to mitigate the air-pumping effect and explain the plateau observed at higher MTD levels. Thus, resuspension behavior is not solely a function of texture depth but also of how surface geometry interacts with tire dynamics. These findings indicate that MTD thresholds for dust mitigation should be tailored to pavement material, especially in urban environments where concrete surfaces are predominant. Additionally, the surface wavelength appears to exert minimal influence on PM₁₀ concentrations when sL is low (e.g., <0.1 g/m²), but under higher-sL conditions, longer wavelengths offer a greater spatial capacity for dust mobilization, thereby increasing the resuspension potential.

Importantly, the variability in dust levels under identical sL conditions highlights the limitations of single-variable predictive models. A combined evaluation of the MTD and surface wavelength offers a more robust predictor of dust resuspension behavior. Furthermore, the application of three-dimensional surface imaging in this study enabled more precise texture quantification than conventional two-dimensional techniques. Future studies should incorporate CFD modeling to simulate airflows at the tire–pavement interface under different surface geometries [34,35].

From a design perspective, maintaining an MTD below 1.4 mm and avoiding short-wavelength textures can help reduce PM emissions without compromising safety or drainage performance. Beyond the technical implications, these results have important environmental and public health relevance. Pavement texture should therefore be viewed not only as a structural or safety-related feature but also as a controllable environmental parameter in urban infrastructure planning [33]. Nevertheless, trade-offs must be carefully managed. While deeper textures improve skid resistance and drainage, they may also elevate resuspension and noise. Conversely, smoother textures reduce dust emissions but may compromise friction and durability. A multi-objective optimization approach is essential—one that integrates environmental performance with mechanical function and safety standards to achieve sustainable pavement design.

5. Conclusions

This study investigated the relationship between suspended road dust and pavement surface attributes, focusing on the roles of the MTD, sL, surface wavelength, and surface treatment methods. The findings confirmed that the MTD is a key determinant of dust resuspension, with PM₁₀ concentrations increasing as the MTD increases and stabilizing beyond a threshold of approximately 1.4 mm—particularly on EACP surfaces. This saturation point provides a functional benchmark for pavement design and maintenance strategies aimed at dust mitigation.

(1)

Interaction Between the MTD and sL

The combined effect of the MTD and sL was found to be highly significant. Under high-sL conditions, deeper textures (MTD ≥ 1.4 mm) resulted in notably elevated PM₁₀ concentrations due to intensified air-pumping effects. These results highlight the importance of considering both the texture depth and surface dust levels in tandem when developing emission reduction strategies.

(2)

Relationship Between the MTD and Road Dust Concentration

Suspended dust concentrations showed a strong positive correlation with the MTD up to the 1.4 mm threshold. Beyond this point, the influence of a greater texture depth diminished, suggesting that excessive roughness no longer contributes meaningfully to resuspension. This threshold offers a quantifiable design parameter for optimizing pavement surface profiles.

(3)

Impact of Surface Treatment Methods on Road Dust Concentration

Surface treatments such as T.Tining, which create deep artificial grooves (~3 mm), were associated with the highest levels of suspended PM₁₀. This is attributed to a reduced contact area and enhanced air pumping, with localized airflow reaching up to 2.2 m/s. While such treatments are beneficial for drainage and skid resistance, they can substantially increase dust emissions. Thus, surface treatment design must carefully balance functional performance with environmental impact.

(4)

Effect of Wavelength

A shorter surface wavelength was associated with increased dust concentrations due to more frequent tire contact points and stronger localized aerodynamic disturbances. In contrast, longer wavelengths mitigated these effects by reducing contact frequency and turbulence. Optimizing wavelength—by minimizing sharpness and increasing inter-aggregate spacing—can support both environmental goals (dust reduction) and mechanical objectives (riding comfort and noise control), though trade-offs with skid resistance and durability must be considered.

(5)

Effect of the Dust Load (sL) on Suspension

Higher sL values (above 0.1 g/m2) significantly increase dust emissions, emphasizing the need for targeted dust control measures in areas with high particulate accumulation.

Overall, the analysis suggests that greater MTD values enhance dust resuspension by increasing tire contact and promoting air pumping, whereas shallower MTDs reduce these effects. The surface wavelength also plays a critical role: very short wavelengths, with dense aggregate peaks, amplify localized pumping and increase PM₁₀ emissions. The optimal configuration appears to be an intermediate MTD combined with a moderate wavelength—sufficient to retain dust without excessively promoting resuspension. Additionally, this study confirmed that concrete pavements tend to retain a higher proportion of fine particulate matter compared to asphalt, likely due to more uniform surface geometry and wear behavior. This underscores the need to manage concrete pavement texture carefully to reduce environmental impact while maintaining structural and safety performance.

Despite its contributions, this study has several limitations. The geographic scope was limited, and measurements were collected over a short-term period without accounting for seasonal variability (e.g., weather, humidity, and maintenance cycles) that can affect sL and dust dynamics. The absence of long-term data also limits the ability to assess how surface degradation, vehicle behavior, and emissions evolve over time. Future research should address these gaps by incorporating multi-seasonal and long-term observations across diverse road types and environmental conditions and by integrating simulation techniques such as CFD for deeper insights into tire–texture interactions.

In conclusion, controlling the pavement surface texture—through targeted adjustment of the MTD, surface wavelength, and aggregate configuration—is essential to reducing suspended road dust while ensuring adequate skid resistance and long-term durability. A multi-objective optimization approach is recommended to balance environmental performance with mechanical and safety requirements in sustainable pavement design. These findings highlight the need to optimize road surface texture to mitigate suspended dust emissions. The proper management of the aggregate size and surface wavelength can play a crucial role in controlling airborne dust, particularly in urban environments with high particulate matter concentrations. Furthermore, while rougher road surfaces enhance friction and braking performance, their role in dust suspension warrants further investigation. Future research should explore the long-term environmental implications of road surface modifications and develop additional mitigation strategies to balance road safety and air quality concerns.