Process of heavy metal transport between soil and the atmosphere: a review

Wind speed.

Based on the transport process discussed in section “Transport process”, the emission of soil fugitive dust requires wind power to impart an initial thrust to soil aggregates, thereby inducing their suspension, saltation, and creep (Lackóová, Kaletová & Halászová, 2023; Lin et al., 2024). Wind speed serves as an index for measuring the wind power. Some studies have shown a positive correlation between the emission fluxes of soil particles and wind speed at the same sampling height (Feng, Sharratt & Wendling, 2011; Li et al., 2015). This correlation likely accentuates the loss of soil HMs, implying that HM emission fluxes increase in tandem with those of soil particles. At lower wind speeds, the emission fluxes of particles are minimal. However, as the wind speed increases, so do the soil aggregates, which in turn cause saltation and creep; as a result, smaller aggregates are released into the air, which raises the emission threshold (Rubinstein, Ben-Hur & Katra, 2020). Nonetheless, the emission of soil particles is also influenced by other factors, thereby significantly complicating the quantitative analysis of wind force.

Concentration of HMs in soil.

The concentration of soil HMs is an important factor for atmospheric HMs, as soil HMs can be transported into the atmosphere in the form of soil fugitive dust. Several studies have found that the higher the concentration of HMs in soil, the more likely these metals are to be emitted into the atmosphere (Li et al., 2016b; Liu et al., 2022; Wu et al., 2020; Yu et al., 2022). Moreover, particles originating from anthropogenic sources tend to be smaller, accumulate easily in surface soil, and are more readily mobilized into the atmosphere by aerodynamic forces. At present, there are no direct studies demonstrating the influence of HMs concentration on this process; hence, further research of the relevant mechanisms is necessary.

Soil water.

Soil water, comprised of adsorbed and capillary water, can effectively reduces soil fugitive dust by enhancing the cohesive force of soil particles, leading to a reduction in HMs loss from soil. Many studies have shown that soil particle loss from soil is inversely proportionate to soil water content, with soil erodibility gradually decreasing towards zero as the soil water concentration increases (Baker, Southard & Mitchell, 2005; Funk et al., 2008; Gill, Zobeck & Stout, 2006; Madden, Southard & Mitchell, 2009). This decrease is due to the weakened cohesive forces between soil particles resulting from the loss of adsorbed and capillary water, along with the increased mass of soil particles (Madden, Southard & Mitchell, 2010). However, some studies have demonstrated that this phenomenon is associated with the transition of soil water from capillary to adsorbed water (Cornelis & Gabriels, 2003; Madden, Southard & Mitchell, 2010). Consequently, both adsorbed and capillary water are pivotal in reinforcing the binding force of soil particles, with capillary water contributing more substantially (Madden, Southard & Mitchell, 2010). There exists a threshold moisture level, roughly one-third of the water content retained by soil at −1.5 MPa (soil water potential), beyond which soil erodibility initially decreases gradually with increasing water content, followed by a rapid decline with each additional increment of water, ultimately reaching zero (Madden, Southard & Mitchell, 2008). Even if adsorbed water is an important binding force in the soil, especially in clayey soil (materials) and the total amount that makes sand, and in the case of clay, the capillary water is not enough to support cohesion in sand and clay (Madden, Southard & Mitchell, 2010).

Soil organic matter (SOM).

Soil organic matter (SOM) is primarily from plant and animal residues, which also has an effect on the quantity and stability of big aggregates, which is a major factor in determining the resistance of the soil to erosion (Li et al., 2015). With the accumulation of SOM, both cation exchange capacity (CEC) and the number of large aggregates increase incrementally (Aimar et al., 2012; An & Xu, 1988). In addition, SOM can effectively enhance microbial diversity (Chen, 2018) and reduce soil emissions. Li et al. (2015) found that PM loss was not significantly related to soil SOM across three wind speeds (8, 10, and 13 m/s) for crushed soil, whereas there was a significant negative correlation at lower speeds for uncrushed soil. Additionally, several studies have shown a positive relationship between SOM content and HM concentration in soil (Bastakoti, Robertson & Alfaro, 2018; Kabata-Pendias, 2017), indicating that SOM effectively immobilizes HMs. Thus, the role of SOM in the loss of soil HMs is dual-faceted (promoting and inhibiting) (Fig. 2) and requires further exploration in future studies.

Soil aggregate size.

The size of soil aggregates is utilized to evaluate the soil’s erodibility (Sirjani et al., 2024). The Wind Erosion Prediction System (WEPS) categorizes soil loss size classes into three groups: saltation and creep (ranging from 0.1 to 2.0 mm), suspension (smaller than 0.1 mm), and smaller than 0.01 mm. These classifications are useful for evaluating the environmental air quality impacts of wind erosion (Feng, Sharratt & Wendling, 2011; Tatarko & Wagner, 2007). Based on this, a more detailed partition, as introduced in section “Transport process”, was subsequently proposed. Both the percentage of aggregates with <0.84 mm diameter and the geometric mean diameter (GMD) are pivotal parameters for evaluating soil susceptibility to wind erosion (Hagen, Wagner & Skidmore, 1999). Li et al. (2015) indicated that the GMD and percentage of aggregates with <0.84 and <0.42 mm diameter bear no significant relationship with the loss of PM₁₀ (particulate matter with diameters below 10 µm in soil fugitive dust) and PM_2.5 (particulate matter with diameters below 2.5 µm in soil fugitive dust) for crushed soil, potentially due to man-made alteration of the original aggregate composition (Fig. 3A). With respect to uncrushed soil, the proportion of aggregates with <0.84 and <0.42 mm diameter correlates positively with the loss of PM₁₀ and PM_2.5, but these correlations were not statistically significant (p > 0.05) (Fig. 3B). The GMD of uncrushed soil (natural soil) was inversely correlated with the loss of PM₁₀ and PM_2.5, but this correlation was not significant (Fig. 3B).

Soil particle size.

Soil particle size (dispersed size) signifies the primary particle composition of soil, while aggregate size (nondispersed size) is employed to evaluate the state of soil particles in situ (Feng, Sharratt & Wendling, 2011). The potential of soil to produce PM₁₀ or PM_2.5 has been gauged by the content of ≤10 µm and ≤2.5 µm in soil. By reviewing the related literature, the correlations between the content of ≤10 µm and ≤2.5 µm, and the ≤10μm/≤2.5 µm ratio with the emission of PM10 and PM2.5 were established. The conclusions are presented in Fig. 4. The contents of ≤10 µm and ≤2.5 µm in soil are negatively associated with the loss of PM10 and PM_2.5. The relationship was significantly negative for all experiments, with PM₁₀ loss more readily influenced by the contents of ≤10 µm and ≤2.5 µm. Moreover, the uncrushed soil at low wind velocities were more susceptible to the impact of ≤10 µm and ≤2.5 µm. However, the correlation between 2.5 µm/10 µm and PM loss varied noticeably with the wind velocity. Moreover, the ratio of 2.5 µm/10 µm exhibited a negative correlation in Fig. 4A for crushed soil. Moreover, in the case of uncrushed soil (Fig. 4B), there was no significant association seen between the 2.5 µm/10 µm ratio and PM loss.

Soil texture refers to the mechanical composition of the soil, which, based on international classification, is divided into sand (0.02–2 mm), silt (0.0002–0.02 mm), and clay (<0.0002 mm). Figure 5 shows the correlation of PM loss with sand, silt, and clay in the soil. The summary reveals a negative correlation between clay content and the loss of PM10 and PM_2.5 for both crushed soil and uncrushed soil at all three speeds, especially notable for PM₁₀ loss. In contrast, silt and sand were positively correlated with the loss of PM₁₀ and PM_2.5 for crushed soil. For uncrushed soil, silt and sand were positively correlated with PM₁₀ loss but demonstrated no significant correlation with PM_2.5 loss. Thus, anthropogenic activities can markedly increase the potential for PM loss. The main components of sand are silicon dioxide and various other minerals, resulting in sand particles that show minimal to no cohesiveness, making them easily dislodged and carried by the wind (Li et al., 2015; Rubinstein, Ben-Hur & Katra, 2020). Gelbart & Katra (2020) affirmed that there was a negative and positive correlation between the amount of sand and clay and PM loss. Considering that the primary constituents of clay are secondary clay minerals, which are distinguished by their high specific surface areas and stable cation exchange properties (Ben-Hur et al., 2009; Eden et al., 2020; Miguel Reichert, Darrell Norton & Favaretto, 2001), clay exhibits exceptional cohesive characteristics (Zuo et al., 2024). It forms wind-resistant clods that provide greater resistance to wind erosion and contribute more wind-resistant materials for abrasion compared to sand and silt (Haynes & Swift, 1990). This, in turn, helps to reduce the transportation of HMs from soil into the atmosphere. Although clay improves stability against wind erosion, it can also absorb a considerable amount of HMs (Proust, Fontaine & Dauger, 2013; Sipos et al., 2021), leading to increased HM loss. Therefore, clay exerts a dual effect, both promoting and inhibiting the loss of HMs in soil (Fig. 6).

Thus, soil particle and aggregate sizes influence the emission of soil fugitive dust, which, in turn, indirectly impacts the emission of soil HMs. The HM content in different particle sizes varies, serving as a critical determinant of HM loss from soil. Several studies have discovered that the concentration of HMs in particle rises as particle size decreases (Bi, Liang & Li, 2012; Shao et al., 2022). This is attributed to smaller particles having a higher specific surface area,which amplifies the presence of iron and manganese oxides, sulfides, organic matte, and clay minerals. Consequently, the HMs were adsorbed more easily onto these fine particles. Given that small particles are more susceptible to wind-driven atmospheric entry, the emission fluxes of HMs are also expected to increase accordingly. Nevertheless, as no studies have yet reported on the mechanisms of HMs transportation from soil to the atmosphere, the impact of soil particles and aggregate size necessitates further investigation.

Farming practices.

Land use patterns can reflect the effects of anthropogenic activities on soil owing to differences in physicochemical properties (Baker, Southard & Mitchell, 2005; Eden et al., 2020; Funk et al., 2008). Among the various land-use types, agricultural land is more susceptible to tillage (Gantulga et al., 2023; Yu et al., 2022; Wang et al., 2022). Conventional tillage alters the mechanical composition of soil, leading to the disintegration of soil aggregates and production of smaller and looser particles prone to saltation (Swet & Katra, 2016; Katra, 2020). This enhances the emission of soil fugitive dust, subsequently increasing the release of HMs from soil (Funk et al., 2008; Goossens, Gross & Spaan, 2001; Sharratt, 2011; Li et al., 2015). Moreover, a number of different farming methods, including those related to farming and cropping, also result in different dust emissions (Baker, Southard & Mitchell, 2005; Madden, Southard & Mitchell, 2008; Madden, Southard & Mitchell, 2009; Sharratt & Schillinger, 2014), thereby influencing the emission of soil HMs. In addition, the application of agricultural inputs, such as straw, soil conditioners, soil amendments, fertilizers, green manure, and livestock manure, can alter the size of soil aggregates and soil texture, influencing the emission of soil particles (Li, Bair & Parikh, 2018). Zhang et al. (2022) determines concentrations of HMs in fertilizer and livestock manure, and their concentration of Cd are 0.12 and 8.31 mg/kg, accounting for 0.19% and 21.15% of Cd in soil, respectively. Therefore, these agricultural inputs also introduce considerable quantities of HMs into soil, thereby increasing the possibility of soil HMs being released into the atmosphere.

Dry deposition.

Dry deposition is the direct deposition of atmospheric particles without rainfall or snowfall (Feng et al., 2019a). The variation in particle size directly influences the duration of the atmospheric residence (Deshmukh, Deb & Mkoma, 2012), subsequently leading to a discrepancy in dry deposition rates and, correspondingly, dry deposition flux. It is important to note that fine particles exhibit a lower dry deposition rate compared to coarse particles (Fang et al., 2004; Sakata & Asakura, 2011). This deposition phenomenon is of particular concern in urban and industrialized areas, which not only have higher emissions but also lack effective scavenging factors such as vegetation (Weerasundara et al., 2017). Numerous studies on dry deposition worldwide have been reported and summarized by Vithanage et al. (2022) and Wu et al. (2018). According to several studies, traffic activity has been found to be a key factor in the accumulation of atmospheric HMs in urban areas (Tasdemir et al., 2006; Weerasundara et al., 2017; Werkenthin, Kluge & Wessolek, 2014). Furthermore, long-range transport of particles is another origin of dry deposition, with particle size playing a determinant role. Considerable dust production from the Taklimakan Desert, the Loess Plateau, and the Gobi Desert transports both natural and anthropogenic HMs along migration pathways, which have a significant negative effect on the water bodies and the soil of East Asia (Li et al., 2016a). For example, there have been reports of dust storms in Korea that were linked to the movement of particles from the Gobi Desert (Alashan semi-desert), the Loess Plateau, and large industrial areas of China (Han et al., 2004). Inevitably, dry deposition is also influenced by meteorological parameters such as wind direction, seasonality, and topographic profiles (Tositti et al., 2020; Negral et al., 2021; Zhong et al., 2021). Because of the interception effect of mountains, Pb in atmospheric particles can enter alpine systems during long-range transport (Bing et al., 2016; Zhang et al., 2013). Zhong et al. (2021) and Zhang et al. (2025a) also showed that the deposition of atmospheric Pb in alpine forest is modulated by terrain. Low altitudes are susceptible to local anthropogenic sources, whereas high altitudes are more prone to be influenced by long-range transport (Zhang et al., 2022).

Wet deposition.

Wet deposition, another pathway of atmospheric deposition, represents a process where atmospheric particles, scavenged within and below cloud formations, descend to the ground surface during rainfall or snowfall events after dissolution in clouds and adsorption onto droplets (Azimi et al., 2005; Feng et al., 2019a). Generally, the concentrations of HMs in wet deposition correlate positively with their solubility, which depends on the pH of rainwater and the origin of the HMs. Studies have indicated that crustal elements such as Al, Fe, Si, Co, and Mn in wet deposition are less soluble, whereas anthropogenic elements (e.g., Cd, Pb, Zn, Cu, and Ni) show higher solubility (Vithanage et al., 2022). Consequently, the concentration of anthropogenic HMs in wet deposits usually exceeds that of naturally sourced HMs (Vithanage et al., 2022; Wu et al., 2018). Moreover, the concentration of HMs in wet deposition is influenced not only by the preceding weather but also by the duration of precipitation. Muezzinoglu & Cizmecioglu (2006) and Weerasundara et al. (2017) suggested that an upward trend in the concentration of HMs during wet deposition when rainfall followed an extended dry period. Interestingly, a significant amount of HMs is washed away during the initial phase of rainfall (first few minutes), resulting in higher concentrations of HMs in wet deposition over short durations compared to longer durations (Vithanage et al., 2022).