*Article* **Estimation of the Rational Range of Ecological Compensation to Address Land Degradation in the Poverty Belt around Beijing and Tianjin, China**

**Haiming Yan 1,2, Wei Li 1,2,\*, Huicai Yang 2, Xiaonan Guo 2, Xingran Liu <sup>2</sup> and Wenru Jia <sup>2</sup>**


**Abstract:** Ecological compensation provides innovative ecological solutions for addressing land degradation and guaranteeing the sustainable provision of essential ecosystem services. This study estimated the ecosystem service value and the opportunity cost of land use in the Poverty Belt of China—around Beijing and Tianjin—from 1980 to 2015 on the small watershed scale, and thereafter estimated the rational range of ecological compensation in this ecologically fragile zone. Results showed that the total ecosystem service value in the study area gradually decreased from CNY 54.198 billion in 1980 to CNY 53.912 billion in 2015. Moreover, the annual total ecological compensation of the whole study area ranged between CNY 2.67 billion and 2.83 billion. More specifically, areas with higher ecological compensation standards are mainly concentrated in the northwestern and northern parts of the study area, with a lower economic development level, while areas with lower ecological compensation standards are mainly located in areas with a relatively high level of economic development, e.g., the southern and southeastern parts of the study area. These results can provide valuable decision-support information for the design and optimization of ecological compensation to address land degradation along with rapid urbanization in the Beijing–Tianjin–Hebei region.

**Keywords:** ecological compensation; ecosystem services; opportunity cost; ecological compensation priority; land degradation

#### **1. Introduction**

Ecological compensation (i.e., payments for ecosystem services or payments for environmental services) is one of the important factors of the construction of ecological civilization in China, and plays a fundamental role in addressing land degradation along with rapid urbanization [1,2]. As an innovative form of ecological solution, ecological compensation can effectively arouse the enthusiasm of ecosystem service providers to alleviate land degradation and guarantee the sustainable provision of essential ecosystem services [3,4], the effectiveness of which has been validated in a number of programs, such as the ecological compensation program in the Catskill Basin in the United States, or the PSA Project in Costa Rica [5–7]. There has also been remarkable achievement in some ecological compensation projects in China, e.g., the Beijing–Tianjin Sandstorm Source Control Project and the "Three North" shelterbelt project [8,9]. However, the theoretical research on ecological compensation in China is still in its initial stages, and is far behind the project practice, which is one of the most important reasons why some ecological compensation projects in China have not achieved their expected effects [10,11]. It is therefore of great practical significance to carry out more in-depth theoretical exploration of ecological compensation for the design and perfection of ecological compensation projects in China [11,12].

**Citation:** Yan, H.; Li, W.; Yang, H.; Guo, X.; Liu, X.; Jia, W. Estimation of the Rational Range of Ecological Compensation to Address Land Degradation in the Poverty Belt around Beijing and Tianjin, China. *Land* **2021**, *10*, 1383. https://doi.org/ 10.3390/land10121383

Academic Editor: Agata Novara

Received: 4 November 2021 Accepted: 6 December 2021 Published: 14 December 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

Rational ecological compensation standards are key to ensuring the effects of ecological compensation, but there is still a lack of universal methods for estimating ecological compensation standards [4,11,13]. In fact, scholars around the world have explored a variety of methods for estimating ecological compensation standards [14,15]. The current methods generally first use a certain method to estimate the upper and lower limits of the ecological compensation standard, and then determine the acceptable ones by making appropriate dynamic adjustments according to the actual situation of the study area and the economic conditions of stakeholders [11,13,16]. For example, the ecosystem service value and opportunity cost have been widely used as the upper and lower limits of ecological compensation standards [13,16]. The opportunity cost measures the opportunity cost of economic development for protecting the ecological environment in the compensated areas, which can be generally estimated via questionnaire surveys, empirical investigation, and indirect calculation [11,17]. However, the opportunity cost method takes less consideration of the spatial heterogeneity within the compensated areas, which often results in insufficient compensation and, consequently, limits the accuracy and applicability of the ecological compensation standard [6,18]. Nevertheless, the opportunity cost method is still the mainstream method for determining ecological compensation standards in developing countries, since it is easy to operate and relatively fair [18]. By contrast, the ecosystem service value, which can be estimated in a direct or indirect way, can provide a reliable scientific basis for determining the ecological compensation standards [13,19]. Ecosystem service value is one of the main bases for determining ecological compensation standards; however, the ecosystem service value estimated with existing methods far exceeds the actual compensation capacity of ecosystem service consumers, and can serve as the theoretical upper limit of the ecological compensation standards [11,20].

The Poverty Belt around Beijing and Tianjin provides an ideal site for the research on ecological compensation, as it is a typical contiguous poverty zone and an ecologically fragile area, but serves as an important ecological barrier in the Beijing–Tianjin–Hebei region [20,21]. The Poverty Belt around Beijing and Tianjin is located in Hebei Province a coastal province that contains the largest number of national-level poverty-stricken counties in China, including 25 of the 39 national-level poverty-stricken counties. The coordinated development of the Beijing–Tianjin–Hebei region is one of the major national development strategies of China, while the construction of the ecological environment is one of the key fields in which prior breakthroughs should be achieved according to the "Beijing–Tianjin–Hebei Coordinated Development Plan Outline" [22,23]. Establishment of a diversified ecological compensation mechanism so as to increase the provision of essential ecosystem services is a major strategic demand for the coordinated development of the Beijing–Tianjin–Hebei region, and can provide an important means of realizing regional sustainable development in the new era [22,24]. It is therefore of extremely important practical significance to promote coordinated development, ensuring ecological safety and promoting the construction of ecological civilization in the Beijing–Tianjin–Hebei region in order to carry out in-depth exploration of the ecological compensation in the Poverty Belt around Beijing and Tianjin in this macro background [25,26].

#### **2. Materials and Methods**

#### *2.1. Study Area*

The Poverty Belt around Beijing and Tianjin expands across Zhangjiakou City, Chengde City, and Baoding City in Hebei Province (Figure 1), with a total area of 82,893.55 km2 (113◦51 47–113◦51 47 E, 39◦1 55–42◦38 7 N); it serves as a key ecological barrier in the Beijing–Tianjin–Hebei region, and plays a dominant role in ensuring national ecological safety [26]. For example, it provides approximately 81% and 93% of the water resources in Beijing City and Tianjin City, respectively [24,27]. More specifically, the forests and wetlands in Zhangjiakou City contribute ecosystem services worth CNY ~15 billion to Beijing every year [20]. However, there are widespread ecologically fragile areas in this region, where the impacts of climate change and increased human activities along with rapid urbanization have led to serious land degradation and greatly threatened the sustainable provision of a number of essential ecosystem services [20,26]. Even worse, this region has not received sufficient ecological compensation even though it has paid a huge opportunity cost of economic development in order to ensure the ecological safety of the Beijing–Tianjin–Hebei region [23,28]. This has led to a sharp contradiction between socioeconomic development and ecological protection in this region, which has seriously threatened the national ecological safety and restrained the high-quality development of the Beijing–Tianjin–Hebei region [22,26].

**Figure 1.** Location of the Poverty Belt around Beijing and Tianjin.

#### *2.2. Estimation of the Ecosystem Service Value*

This study estimated the ecosystem service value of the study area from 1980 to 2015 using the equivalent factor method at a 1 km grid scale, and thereafter summarized it on the small watershed scale, since there is generally a homogeneous internal ecological environment in a small watershed [29,30]. First, this study classified the local ecosystems into five types (Table 1), and determined the main ecosystem service types according to previous research [1,26] and the specific situation of the study area. Then, the ecosystem service value was further categorized into market and non-market value according to the supply–demand relationship between the ecosystem and human society, based on previous studies [19,31,32]. Thereafter, the ecosystem service indicators of most concern to stakeholders in the study area and related regions were determined (Table 1). Finally, this study estimated the ecosystem service value of the study area via the ecosystem service value coefficient per unit land area in 2010 (Table 2), based on the previous studies [32,33], as follows:

$$ESV = \sum\_{i=1}^{n} \sum\_{j=1}^{m} A\_j \times E\_{ij} (i = 1, \dots, n, j = 1, \dots, m) \tag{1}$$

where *ESV* is the total ecosystem service value of a certain spatial unit, and this study chose the 1 km grid and small watershed with a generally homogeneous internal ecological environment as the basic spatial units. More specifically, the small watershed boundary data were extracted from the dataset of river basins and networks of China, based on the DEM (https://www.resdc.cn/DOI/doi.aspx?DOIid=44, accessed on 31 October 2021). *Eij*

is the equivalent factor of the *i* th ecosystem service of the *j* th ecosystem (Table 2). *Aj* is the area of the *j* th ecosystem, which is obtained from the Land Use Remote Sensing Monitoring Data of China provided by the Resource and Environmental Science and Data Center, CAS (https://www.resdc.cn/Default.aspx, accessed on 31 October 2021).

**Table 1.** Classification of the ecosystem service value in the Poverty Belt around Beijing and Tianjin.


**Table 2.** Coefficients of ecosystem service value per unit of land area in the Poverty Belt around Beijing and Tianjin, based on previous studies [19,31,32] (unit: CNY/hm2).


#### *2.3. Estimation of the Range of Ecological Compensation*

This study separately estimated the ecological compensation standards based on the ecosystem service value and the opportunity cost on the small watershed scale, and thereafter determined the rational range of the ecological compensation standard and total ecological compensation value in the study area. This study first estimated the ecological compensation standards based on the ecosystem service value and gross domestic product (GDP) in the study area. On the one hand, the ecosystem service value can generally only serve as the upper limit of ecological compensation, since it far exceeds the payment ability of ecosystem service consumers, and a conversion coefficient has been widely used to make ecological compensation based on the ecosystem service value more practical and acceptable [11,20]. Meanwhile, this study took into account only the non-market ecosystem service value, since the market value of ecosystem services can contribute to regional economic development through market mechanisms [11,34,35]. On the other hand, the more heavily the economic development in a certain area depends on natural resources, the higher the opportunity cost to protect the ecological environment in that area, which can be represented by the degree of priority for ecological compensation [32]. This study accordingly represents the degree of priority for ecological compensation in a certain area with the ratio of the non-market value of ecosystem services to GDP per unit of area, based on existing studies [8,32], and estimates it in a spatially explicit way in order to further improve the practicability of ecological compensation. The ecological compensation based on the ecosystem service value was finally estimated as follows:

$$R\_{T \text{ev} \\_j} = ESV\_{T \text{\\_}} \times k \times p\_{\text{\\_}} \tag{2}$$

$$p\_{\Box i} = 2 \arctan(\frac{ESV\_{T\_{\Box i}}}{G\_{T\_{\Box i}}}) / \pi \tag{3}$$

where *RTesv\_i* is the ecological compensation value in the *i* th area based on the ecosystem service value; *ESVT\_i* is the total non-market value of ecosystem services in the *i* th area per unit of area; *k* is the conversion coefficient of the ecosystem service value, which is set to 15% based on previous studies [32]; *p\_i* is the degree of priority for ecological compensation of the *i* th area, and the higher *p\_i* is, the more urgently the ecological compensation of the *i* th area is needed; *GT\_i* is the total GDP per unit of area of the *i* th area, which is extracted from the Spatialized GDP Dataset of China provided by the Resource and Environmental Science and Data Center, CAS (https://www.resdc.cn/DOI/doi.aspx?DOIid=33, accessed on 31 October 2021); and *π* is pi.

This study further estimated the ecological compensation based on the opportunity cost of land use. Farmers in the study area play an important role in protecting the ecological environment, for which they sacrifice their economic development rights [24,27]. The loss of economic development due to ecological protection can be reflected in the opportunity cost, while the latter can be measured by the land rent per unit of area [36]. This study obtained the data on the land rent per unit of area of various land types in the study area by carrying out some field surveys and using querying websites (e.g., https://www.tuliu.com/, accessed on 20 October 2021). The total ecological compensation based on the opportunity cost was finally estimated as follows:

$$R\_{\text{Toc\\_j}} = OC\_{\text{\\_}} \times \lambda \tag{4}$$

where *RToc\_i* is the ecological compensation value in the *i* th area based on the opportunity cost; *OC\_i* is the opportunity cost of land use in the *i* th area based on the land rent, which is estimated based on the land rent price; and *λ* is the opportunity cost conversion coefficient, which is also set to 15% based on the results of field surveys and the ratio of the transaction price to the listing price on the websites (e.g., https://www.tuliu.com/, accessed on 20 October 2021).

#### **3. Results**

#### *3.1. Dynamics of the Ecosystem Service Value*

The results suggested that the total ecosystem service value in the Poverty Belt around Beijing and Tianjin showed an overall downward trend between 1980 and 2015 (Figure 2). Specifically, the total ecosystem service value declined most obviously between 1980 and 1990, and it recovered to a certain degree from 1990 to 1995, but thereafter showed a further gradual declining trend. In 2015, the total ecosystem service value of the study area reached CNY 53.912 billion, with a decrease of CNY 286 million compared to that in 1980. More specifically, the total ecosystem service value of forestland decreased by CNY 292 million, while that of the water body decreased by CNY 81 million, which was primarily due to the conversion of the forestland and water body with higher equivalent factors to cropland and grassland with lower equivalent factors. By contrast, the total ecosystem service value of cropland, garden plots, and grassland increased slightly between 1980 and 2015, with increases of CNY 29 million, 29 million, and 31 million, respectively.

#### *3.2. Spatial Heterogeneity of Ecological Compensation*

The results suggested that the ecological compensation standards based on the ecosystem service value in the study area showed significant spatial heterogeneity, ranging from CNY 0.47/hm2 to CNY 910.73/hm2, with an average of about CNY ~341.75/hm2 (Figure 3). The areas with low ecological compensation standards are widespread in the southern, central, and northeastern parts of the study area; for example, areas with ecological compensation standards of CNY < 150/hm2 and CNY 150–250/hm<sup>2</sup> are contiguously distributed in most parts of Laiyuan County, Yi County, and Laishui County in the southern part of the study area, and almost all of the northeastern part of the study area. By contrast, the

areas with the ecological compensation standard of CNY > 600/hm2 are mainly located in the western and northern parts of the study area, e.g., most parts of Kangbao County, the northern part of Shangyi County, most parts of Guyuan County in Zhangjiakou City, and a few parts of Fengning County and Weichang County in Chengde City. The areas with the ecological compensation standard of CNY 450–600/hm2 are generally adjacent to these areas with the ecological compensation standard of CNY > 600/hm2, e.g., most parts of Zhangbei County, the southwest part of Shangyi County, the middle part of Guyuan County, and most parts of Weichang County. The level of economic development is generally very low in these areas with low ecological compensation standards, and the ecological compensation is overall attractive to most of the local farmers in these areas. For example, the statistical data suggest that the average rural per capita income from property in Zhangjiakou City was CNY 103 in 2010. Meanwhile the cropland area per capital in Zhangjiakou City was generally 0.2133–0.2667 hm2, and the ecological compensation of CNY 600/hm2 means that the income from ecological compensation is CNY 128~160 per capita, which can generally effectively motivate the local farmers to participate in ecological conservation.

**Figure 2.** The total ecosystem service value in the Poverty Belt around Beijing and Tianjin from 1980 to 2015 (unit: CNY 100 million).

**Figure 3.** Ecological compensation standards based on the (**a**) ecosystem service value and (**b**) opportunity cost in the study area.

The spatial pattern of ecological compensation standards based on the opportunity cost was overall consistent with that based on the ecosystem service value, with some remarkable differences in a few areas. For example, the ecological compensation standard based on the opportunity cost was generally CNY > 500/hm2 in Shangyi County, Zhangbei County, and Kangbao County in the northwest of the study area, while it was generally below CNY 400/hm<sup>2</sup> in Weichang County and Fengning County in the northern part of the study area. By contrast, the areas with lower ecological compensation standards were concentrated in the southeastern and southern parts of the study area, where the ecological compensation standards were generally CNY 200–300/hm2 or below CNY 200/hm2. In general, the ecological compensation standards based on the ecosystem service value and opportunity cost were consistent overall, i.e., the areas with higher ecological compensation standards were mainly concentrated in the northwestern and northern parts of the study area, while areas with lower ecological compensation standards were mainly located in the southern and southeastern parts of the study area.

The results of this study showed that the total ecological compensation value based on the ecosystem service value and the opportunity cost in the whole study area was approximately CNY 2.83 billion and CNY 2.67 billion per year, respectively—very close to and overall consistent with the results of Xu et al. [37], i.e., CNY 3.45 billion per year. The total ecological compensation value based on the ecosystem service value on the small watershed scale ranged between CNY 2.40 thousand and 52.98 million per year, showing conspicuous spatial heterogeneity. Specifically, areas with a total ecological compensation value based on the ecosystem service value below CNY 5.00 million per year were continuously distributed in the southern, middle, and northeastern parts of the study area, where there is a relatively better ecological environment and a higher level of economic development, jointly resulting in a weak demand for ecological compensation (Figure 4). By contrast, areas with a total ecological compensation value exceeding CNY 5.00 million per year were mainly scattered in a few regions in the northern, northwestern, and central parts of the study area. More specifically, areas with a total ecological compensation value exceeding CNY 25.00 million per year were mainly located in Zhangbei County, Guyuan County, Fengning County, and Weichang County in the northwestern and northern parts of the study area, as well as in Chicheng County in the middle part of the study area. The level of economic development is very low in in these areas, where there are widespread key ecological function zones and, consequently, there is very strong demand for ecological compensation in these areas.

**Figure 4.** The total ecological compensation value based on the (**a**) ecosystem service value and (**b**) opportunity cost on the small watershed scale in the study area (unit: CNY 10,000).

The spatial pattern of the total ecological compensation value based on the opportunity cost was generally consistent with that based on the ecosystem service value, but with significant differences in a few areas. For example, the total ecological compensation value based on the opportunity cost was generally consistent with that based on the ecosystem service value in the northern and northwestern parts of the study area, all exceeding CNY 20.00 million per year. However, there are also areas where the total ecological compensation value based on the opportunity cost exceeds CNY 20.00 million per year in Yangyuan County and Huailai County in the middle part of the study area, as well as in Longhua County, Luanping County, and Xinglong County in the southeastern part of the study area (Figure 4). This may be because there is more cropland with a higher opportunity cost but relatively lower ecosystem service value in these areas, leading to a higher total ecological compensation value based on the opportunity cost. In general, this study showed that the total ecological compensation value of the whole study area ranged between CNY 2.67 billion and 2.83 billion per year, and that there were generally consistent spatial patterns of the total ecological compensation value based on the ecosystem service value and the opportunity cost, indicating that the ecological compensation in this study is overall reliable.

#### *3.3. Spatial Pattern of the Ecological Compensation Priority*

The results of this study showed that the ecological compensation priority in the study area ranged between 0.002593 and 0.6269, with an average value of 0.3165. This study further classified the ecological compensation priority into five levels using the Jenks natural break method, with the breakpoints of 0.1372, 0.2646, 0.3943, 0.4922, and 0.6269 (Figure 5). There was remarkable spatial heterogeneity of the ecological compensation priority in the study area, where areas with high or very high ecological compensation priority were concentrated in the Bashang region in the northern and northwestern parts of the study area (Figure 5). Specifically, areas with very high ecological compensation priority were concentrated in the northern part of the study area, e.g., most parts of Shangyi County, almost all of Kangbao County and Guyuan County, the central and northern parts of Fengning County, and most of Weichang County. Meanwhile, areas with high ecological compensation priority were concentrated in most parts of Zhangbei County and parts of adjacent Chongli County, most parts of Chicheng County and part of adjacent Fengning County, and a few parts of Shangyi County and Weichang County. There are widespread important ecological function zones with enormous ecosystem service value in these areas, all of which play an important role in guaranteeing the ecological safety of the Beijing– Tianjin–Hebei region by providing a number of essential ecosystem services, such as wind prevention and sand fixation, water conservation, and biodiversity protection. Meanwhile, these inland "ecological export" areas with a low level of economic development have paid a huge development opportunity cost for a long period in order to guarantee the ecological safety of the Beijing–Tianjin–Hebei region.

It is an arduous task to implement further ecological environmental protection by relying on the local resources in areas with high or very high ecological compensation priority, where there is an urgent need for the provision of ecological compensation by other areas [23]. On the one hand, there is generally a relatively higher level of economic development in areas with low or very low ecological compensation priority, which are generally located in the southern, central, and northeastern parts of the study area, e.g., Xuanhua District, Xuanhua County, and Huailai County in the southern part of Zhangjiakou City, and Kuancheng County and Luanping County in the southern part of Chengde City. These areas generally have a relatively greater ability to implement ecological conservation, with significant geographical advantages, and taking considerable advantage of the cheap agricultural and forestry products from those areas with high or very high ecological compensation priority. Nevertheless, these areas still need some external financial support in order to establish a more eco-friendly economic system, and cannot provide sufficient support for ecological conservation in areas with high or very high ecological compensation priority. On the other hand, some economically developed areas—e.g., Beijing City and Tianjin City—have taken enormous advantage of the ecosystem services from the study area and, therefore, should provide some ecological compensation in order to promote the ecological conservation of the study area [26,28,32]. For example, the areas with high or very high priority have provided a large amount of water resources to Beijing City and

Tianjin City, while the areas with low or very low priority have also paid considerable opportunity costs of industrial development and agricultural production in order to ensure the supply of water resources to Beijing City and Tianjin City [24,27]. There is an urgent need for ecological compensation from Beijing City and Tianjin City, which can play an important role in ameliorating the standard of living in the study area and avoiding a more intensive manner of land use with serious biodiversity degradation [23,26,32]. Overall, there is an urgent need for more ecological compensation in the study area, especially in those areas with high or very high ecological compensation priority, which should be met with financial support from areas outside the study area.

**Figure 5.** Ecological compensation priority levels in the Poverty Belt around Beijing and Tianjin.

#### **4. Discussion**

The results of this study can provide valuable spatially explicit reference information for the design and improvement of ecological compensation projects, but it is still necessary to carry out some more in-depth research. For example, this study clearly revealed the spatial heterogeneity of the ecological compensation in the study area using the small watershed scale rather than the county-level scale, which may contribute to formulating more specifically targeted policy measures and improving the feasibility of ecological compensation policies. However, this study estimated the ecosystem service value with some static parameter values, which cannot accurately reflect the time-series dynamics of the ecosystem service value. Moreover, this study considered the inflation factors, but still cannot accurately reflect the dynamic changes in the ecological compensation standard, since the latter was estimated based on the Spatialized GDP Dataset of China in 2010, which is a static dataset even though it can more accurately reveal the spatial heterogeneity of the ecological compensation. It is still necessary to reveal the time-series dynamics of the ecosystem service value and ecological compensation standards more accurately using more dynamic parameter values. Overall, this study accurately revealed the rational range of ecological compensation in the study area in a spatially explicit way, but it is still necessary to carry out further research in order to provide more reliable reference information for the design and improvement of ecological compensation projects.

#### **5. Conclusions**

This study revealed the rational range of the ecological compensation in the Poverty Belt around Beijing and Tianjin based on the ecosystem service value and the opportunity cost, in a spatially explicit manner. The following conclusions were finally drawn: (1) The total ecosystem service value in the study area showed an overall downward trend between 1980 and 2015, decreasing from CNY 54.198 billion in 1980 to CNY 53.912 billion in 2015. (2) The total ecological compensation value of the whole study area ranged between CNY 2.67 billion and 2.83 billion per year, and it is feasible to estimate the ecological compensation based on the ecosystem service value and the opportunity cost. (3) Areas with a higher ecological compensation value and priority level are mainly located in areas with lower levels of economic development in the northwestern and northern parts of the study area, while areas with a lower ecological compensation value and priority level are mainly located in areas with relatively high levels of economic development in the southern and southeastern parts of the study area, but both of these areas are in urgent need of ecological compensation from other areas, e.g., Beijing City and Tianjin City. (4) It is still necessary to carry out further research on the time-series dynamics of the ecological compensation in order to provide more reliable reference information for the design and improvement of ecological compensation projects. Overall, this study accurately reveals the rational range of the ecological compensation in the study area in a spatially explicit manner, and can provide valuable information for addressing land degradation along with the rapid urbanization in the Beijing–Tianjin–Hebei region.

**Author Contributions:** H.Y. (Haiming Yan) and W.L.: Investigation, data curation, writing—original draft preparation, writing—review and editing, funding acquisition; H.Y. (Huicai Yang) and X.G.: conceptualization, methodology, supervision, project administration; X.L. and W.J.: software, validation, visualization. All authors have read and agreed to the published version of the manuscript.

**Funding:** This study was funded by the Natural Science Foundation of Hebei Province (E2019403210, D2019403022, D2019403168, C2019403114), the Science and Technology Project of Hebei Education Department (BJ2019045), the National Natural Science Foundation of China (51909052, 41807169, 42001034, 42001027), the Scientific and Technological Innovation Team Project of Hebei GEO University in 2021 (KJCXTD-2021-10), and the College Students' Innovative Entrepreneurial Training Plan Programs (202110077011, S202110077030, S202110077031).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data presented in this paper are available on request from the corresponding author.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


### *Review* **Progress in Dust Modelling, Global Dust Budgets, and Soil Organic Carbon Dynamics**

**Weixiao Chen 1,2,3,†, Huan Meng 1,4,†, Hongquan Song 1,5 and Hui Zheng 1,4,\***


**Abstract:** Dust emission is an important corollary of the soil degradation process in arid and semi-arid areas worldwide. Soil organic carbon (SOC) is the main terrestrial pool in the carbon cycle, and dust emission redistributes SOC within terrestrial ecosystems and to the atmosphere and oceans. This redistribution plays an important role in the global carbon cycle. Herein, we present a systematic review of dust modelling, global dust budgets, and the effects of dust emission on SOC dynamics. Focusing on selected dust models developed in the past five decades at different spatio-temporal scales, we discuss the global dust sources, sinks, and budgets identified by these models and the effect of dust emissions on SOC dynamics. We obtain the following conclusions: (1) dust models have made considerable progress, but there are still some uncertainties; (2) a set of parameters should be developed for the use of dust models in different regions, and direct anthropogenic dust should be considered in dust emission estimations; and (3) the involvement of dust emission in the carbon cycle models is crucial for improving the accuracy of carbon assessment.

**Keywords:** dust emission; wind erosion; dust models; dust cycle; carbon cycle

#### **1. Introduction**

The aeolian and fluvial processes play a fundamental role in earth systems and have important environmental and ecological effects at both local and global scales [1]. Wind erosion is a natural geological process involving the detachment, transport, and deposition of soil particles by strong winds [2–5], and it is a key soil degradation process in arid and semi-arid areas worldwide [6–9]. In contrast to water erosion, where the eroded material follows determined paths, wind-eroded material is widely dispersed over the landscape [10]. The mineral dust generated by soil particle emissions, in turn caused by wind erosion, is considered the most important source of atmospheric aerosols [11]. The global annual emission amount of mineral dust due to wind erosion is estimated to be around 1 to 5 billion (109) tons [11–14], which account for approximately 30–50% of the total aerosol introduced into the atmosphere [15]. Dust aerosols play important roles in regulating the Earth's radiation budget, climate, global biogeochemical cycles, terrestrial soil formation, air quality, and human health [16–30].

To assess the socio-economic and environmental effects of dust processes, it is essential to quantify the dust emission rates at different spatial and temporal scales. Dust emission involves complex interactions among soil properties, climate, vegetation, and land use regimes. The understanding of dust processes and the capability of dust emission models

**Citation:** Chen, W.; Meng, H.; Song, H.; Zheng, H. Progress in Dust Modelling, Global Dust Budgets, and Soil Organic Carbon Dynamics. *Land* **2022**, *11*, 176. https://doi.org/ 10.3390/land11020176

Academic Editors: Jinyan Zhan, Xinqi Zheng, Shaikh Shamim Hasan and Wei Cheng

Received: 8 December 2021 Accepted: 18 January 2022 Published: 21 January 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

have improved considerably over the past five decades. Based on the measured physical properties of dust emission at the field scale, several approaches have been adopted to estimate dust emission rates, such as mathematical simulation using data on the relationships between meteorological records and interacting surface parameters [31], remote sensing [32,33], and using geographic information systems (GIS) [34–37]. Numerous dust models have been developed to quantify dust emission rates and soil losses in the field [38,39], regional [40,41], continental, and global scales [42,43].

Soil is the main terrestrial reservoir of organic carbon and contributes substantially to the global carbon cycle [1,44] . Small changes in the soil organic carbon (SOC) stock may result in large changes in atmospheric carbon dioxide (CO2) concentration [4,45]. Dust emission is an essential component of the carbon budget; it removes carbon from vast areas and, if the wind is strong enough, readily transports carbon dust offshore [46,47]. Thus, soil redistribution through dust mobilisation is an important mechanism underlying carbon cycling in terrestrial ecosystems, the atmosphere, and oceans. The active component of SOC and the organic carbon combined with the fine fraction of the soil are easily removed from terrestrial ecosystems via dust emission [48]. Wind-driven mobilisation of carbon augments the net loss of carbon from terrestrial systems.

In this review, we discuss empirical and physical dust models at multiple spatial scales, developed worldwide over the past five decades; the effects of dust emission on global dust budgets and SOC dynamics; and the link between dust processes and the global carbon cycle.

#### **2. Dust Models Adopted Worldwide**

#### *2.1. Factors Influencing Dust Emissions at Multiple Scales*

Dust emission is a dynamic natural process regulated by complex interactions among the climate, soil properties (grain size, aggregation, structure, moisture, and surface roughness), vegetation (cover, distribution, and height), and land use at different spatial and temporal scales [3,34,49–54]. This process is recognised as a major source of uncertainty in climate models [55,56].

Dust emission is essentially a flow process in which soil is detached from an erodible surface and transported in various ways (surface creep, saltation, and suspension) in response to wind shear stress (Figure 1) [57]. Dust transport mechanisms redistribute soil and associated nutrients and organic materials at different spatial scales (Figure 1). The mechanism by and the distance to which soil particles are transported are determined by their size. Large (>500 μm) and medium-sized (100–500 μm) particles are more likely to be transported via surface creep and saltation, respectively, over relatively short distances; smaller particles (<100 μm) can be transported via suspension over longer distances, across regions, continents, and the world [32,57–61].

**Figure 1.** Dust processes and their controls at different spatial scales.

The development of dust models requires an understanding of the factors affecting dust emission at different spatial scales. At the grain scale (<10−<sup>2</sup> m), dust emission is controlled by wind shear speed and the structure, texture (particle size distribution), moisture content, mineral composition, electrostatic forces, chemistry, and microbiota composition of the soil (Figure 1) [57,62–70]. Together, these factors determine the weight, drag, and interparticle cohesion of soil aggregates and threshold friction velocity (u\*t) [57]. The u\*t, which controls both frequency and intensity of erosion events, is the minimum friction velocity required to initiate the movement of soil particles, representing the strength of forces among the soil particles and the capacity of an aeolian surface to resist wind erosion [60,71–73]. This crucial parameter controls the frequency and intensity of dust emission. Soil erodibility is defined as the susceptibility of soils to detach and transport by erosive agents, namely water or wind. Soil erodibility is also dependent on the intrinsic properties of soils (include texture, mineralogy, chemistry, and organic matter content) and the combined influence of temporal soil properties, namely moisture, aggregation, surface crusting, and the availability of loose erodible material [57,66,68,69,74–77]. At the field scale, the grain-scale conditions of soil texture, soil moisture, and inter-particle bonding control soil aggregation and crusting, and thus, influence soil particle movement and the potential for dust emission [74,75]. Aggregation and crusting affect soil surface roughness, u\*t, and the availability of loose erodible soil particles. The latter parameters affect soil erodibility at the landscape (103 m) scale [62].

At the landscape scale, dust emission is determined by soil type, vegetation cover, cultivation practices, soil surface roughness, u\*t, and the availability of loose erodible material [57]. However, at the regional to global scales (>10<sup>4</sup> m), the transport, transformations, and deposition of dust particles, and their chemical reaction with air pollutants are affected by soil type, landforms, climate, ecoregional environmental conditions, and practices of land use and land management [76]. Together, these factors determine the

relative influence of soil moisture, aggregation, and crusting on the soil surface, as well as the spatial and temporal variations in soil erodibility at the field scale. The regional climate, other ecoregional conditions, and land use practices may, in turn, be affected by dust transport and deposition. This interdependence generates feedback that affects soil erodibility at various scales, from the landscape to the microscopic [57].

#### *2.2. Dust Models at Multiple Spatial and Temporal Scales*

To understand the role of dust in the earth system, numerous models that simulate dust emission at various spatial and temporal scales have been developed since the 1960s [38,78–81]. Most of these models are used to predict dust emission rates. Dust models can be divided into empirical and physical types [7]. Empirical models are based on functions derived from field or wind tunnel experiments under a wide variety of soil types and soil surface roughness conditions. Physical dust models focus largely on the physical mechanisms of dust movement and predict patterns of dust emission, transport, and deposition driven by climate, land use, and/or the land management measures being employed. The evolution of the dust models reviewed in this study is illustrated in Figure 2. We systematically reviewed a representative selection of 18 dust models developed over the past 60 years.

Dust models usually concentrate on smaller (<100 μm in diameter) soil particle emissions, as such particles can be suspended in the atmosphere and transported over long distances [61]. One such model is the wind erosion equation (WEQ) developed by Woodruff and Siddoway [38] from empirical functions that describe the effects of environmental factors on the rate of soil loss. Physical models were developed mainly in the 1980s (Figure 2). As shown in Figure 2, before 2000, dust models mainly constituted dust emission modules at the field scale (e.g., WEQ, wind erosion prediction system (WEPS), Texas erosion analysis model (TEAM), revised wind erosion equation (RWEQ), wind erosion assessment model (WEAM), and wind erosion stochastic simulator (WESS)) [38,39,82–85]. With increasing awareness of the role of dust at regional scales, regional dust emission models were developed and forced by climate datasets (e.g., wind erosion on European light soils (WEELS) and Australian land erodibility model (AUSLEM)) or dust transport models were developed by integrating dust emission modules with regional- to global-scale climate models (e.g., integrated wind erosion modelling system (IWEMS), computational fluid dynamics wind erosion model (CFD-WEM), computational environmental management system (CEMSYS), global ozone chemistry aerosol radiation and transport (GOCART), GOCART-Air Force Weather Agency (GOCART-AFWA), GOCART—University of Cologne (GOCART-UoC), dust entrainment and deposition (DEAD), community aerosol research model (CARMA-MM5), and global transport model of dust (GMOD), and Lund-Potsdam-Jena dynamic global vegetation model–dust (LPJ-Dust)) [11,40–43,53,86–91].

**Figure 2.** Evolution of wind erosion models. AUSLEM: Australian Land Erodibility Model [91]; CARMA-Dust: Community Aerosol Research Model, by Ames/NASA [86]; CEMSYS: Computational Environmental Management System [87]; CFD-WEM: Computational Fluid Dynamics Wind Erosion Model [53]; DEAD: Dust En-trainment and Deposition [42]; DENAPAP: Dust Emission Model, supported by the National Acid Precipitation Assessment Program [92]; DPM: Dust Production Model [93]; GMOD: Global Transport Model of Dust [43]; GOCART: Global Ozone Chemistry Aerosol Radiation and Transport [94]; GOCART—AFWA: GOCART—Air Force Weather Agency [88]; GOCART—UoC: GOCART—University of Cologne [11,89,90]; IWEMS: Integrated Wind Erosion Modeling System [40]; LPJ—dust version 1.0: Lund–Potsdam–Jena dynamic global vegetation model—dust version 1.0 [95]; RWEQ: Revised Wind Erosion Equation [82]; TEAM: Texas Erosion Analysis Model [84]; WEAM: Wind Erosion Assessment Model [85]; WEELS: Wind Erosion on European Light Soils [41]; WEPS: Wind Erosion Prediction System [39]; WEQ: Wind Erosion Equation [38]; WESS: Wind Erosion Stochastic Simulator [83].

To account for the complex interaction between the physical processes and anthropogenic factors of wind erosion, dust models are drawn from field and laboratory measurements. Owing to the differences in model complexity, required inputs, and outputs [61], these models provide variable dust process simulations at specific spatial and temporal scales. The spatio-temporal scales, input parameters, and outputs of the dust models reviewed in this study are summarised in Table 1. The early dust models are mainly focused on the development of dust emission models at a field scale. However, dust models after 2000 have mainly concentrated on dust transport models at the regional and global scales. Field-scale dust models can be used to assess soil losses due to wind erosion under different land management regimes. Physical models require more detailed inputs and are difficult to implement owing to the lack of soil and land-surface parameters. In addition, these emission models are mainly applied at a field scale and usually cannot estimate spatial variations of dust emissions for a region. Regional and global-scale dust models integrated

into dust emission modules and climate models should be employed to predict spatial and temporal variations of dust processes, such as dust emission, transport, and deposition.

**Table 1.** Summary of reviewed dust models, including spatial scale, inputs, and outputs.



**Table 1.** *Cont.*

Note: F, R, C, and G represent field, regional, continental, and global, respectively.

Evidently, u\*t is the key factor affecting dust emission simulations. Dust emission will occur when the wind friction velocity over the land surface (u\*) exceeds u\*t. Generally, there are two approaches for representing the factors that influence the soil's susceptibility to wind erosion in dust models: (1) constructing empirical relationships between soil surface conditions, soil moisture, and vegetation cover to predict rates of soil loss (e.g., models WEQ and RWEQ); and (2) integrating physical processes and theoretical relationships among soil properties, land surface conditions, and u\*t (e.g., models CFD-WEM, GMOD, TEAM, and WEAM). Empirical models can account for dynamic variations in soil erodibility [96], but they largely depend on field measurements, which are not available at large spatial scales [38]. Physical models enable the inclusion of large-scale spatial inputs and are not restricted to specific environments [40,42]. The complexity of the description of u\*t and dust emission has increased as the development of dust models progressed over the past five decades (Table 1). The definitions of u\*t and the soil's susceptibility to wind erosion differ between these models; therefore, calculations and predictions obtained from different models are not directly comparable. Thus, there is a need to integrate the two types of models to reduce the uncertainty of dust models overall.

Precisely modelling the spatial and temporal variability of dust emissions is a prerequisite to estimate and forecast atmospheric dust concentrations and their effects. Current dust emission models mostly include main physical processes of dust production, which can reproduce the spatial and temporal variability of dust emissions if the model inputs are accurately described. Studies have confirmed that the accuracy of ground surface condition data is the key determinant of spatial and temporal variability accuracy of dust emissions models [9,77,97]. In addition, the accuracy of temporal variability of dust emission models is also determined by specific model parameter values, such as the Kawamura coefficient value in a dust scheme [90] and the roughness correction factor to u\*t in the dust schemes of the Community Earth System Model (CESM) [77]. However, it is difficult to define or customise these values owing to the spatial heterogeneity of ground surface conditions and the dearth of dust observations. To evaluate and improve the performance of temporal variations of dust emission models, it is essential to improve the accuracy of surface parameters in dust emission models and strengthen the collection of dust observation globally.

Dust models are important tools to account for the complex interaction between the physical processes and anthropogenic factors of wind erosion. However, there are no universally accepted parameters for these models in different regions/countries. Therefore, most of these models have to be parameterised before they can be applied to other regions. For example, models IWEMS and CFD-WEM have been successfully applied to the simulation of dust emissions in Asia after calibration of their parameters [9,96,98–101]. Model parameterisation is essential to ensure the accuracy of the estimation results. Dust models can be evaluated using in-situ measurements of dust and other required inputs. To ensure the accuracy of the simulation, before the application of a dust model in a region, the model's empirical variables can be adjusted by comparing the model's predictions with field measurements [101]. However, this comparison is challenging because of the difficulty in obtaining dust data. Several studies have attempted to validate dust model predictions against measured and observed data. The performance of some of the selected dust models

in Europe, Australia, and China (Table 2) have shown considerable differences in the accuracy among different dust models and among estimations in different regions using the same model. This also proved that the localisation of model parameters is important for the simulation accuracy of dust models.


**Table 2.** Performance of selected dust models.

#### **3. Global Dust Budgets**

#### *3.1. Dust Sources and Sinks*

Global dust source regions have been identified using different approaches, such as information gathering from dust weather records [99], remote sensing [32,100], dust monitoring networks [106], and dust models [42,55,86,107]. The seven main dust source regions of the world are North Africa, Middle East/Central Asia, East Asia, North America, South America, South Africa, and Australia (Figure 3).

Some studies simulated the global dust emission, deposition, and budgets over the past three decades (Table 3). The map of global dust emission and deposition in different regions (Figure 3) generated based on the data from previous studies represented in Table 4 shows that North Africa is the largest dust source region in the world. Because of the Sahara, the world's largest desert, North Africa accounts for approximately 60% of the global dust emissions and approximately 65% of the global atmospheric dust load [55,108]. The second largest dust source region is Asia, comprising Arabia, Central Asia, and East Asia. Dust emissions and atmospheric dust loads in Asia account for approximately 30% of the global values [42,43,55]. Specifically, the dust emission and atmospheric dust load in East Asia are approximately 214 and 1.1 Tg yr<sup>−</sup>1, respectively [55]. Australia is the largest contributor to dust emissions in the Southern Hemisphere, accounting for approximately 6% of the global dust emissions [33,42,55,108–110] and 5% of the global atmospheric dust load [43,55]. The smallest dust source regions are North and South America, accounting for 0.3% and 2.5% of the global dust emissions, respectively [33,42,55,109].


**Table 3.** Research periods of global dust emission, deposition, and budgets in several studies.

**Table 4.** SOC erosion associated with dust emission in major regions of the world.


The amount of dust deposition over land is around three orders of that deposition over oceans [52]. Although dust deposition measurements are relatively scarce and incomplete worldwide, existing dust deposition rates records show large variations on land and oceans [52]. The estimates of dust deposition on the ocean shown in Figure 3 illustrate a considerable discrepancy among different studies. Nevertheless, according to most estimates, the region of maximum dust deposition is the North Atlantic due to the Saharan dust, which accounts for nearly 43% of the total dust deposited worldwide [18]. The second largest deposition centre is the Indian Ocean, receiving dust from North Africa, Arabia, Central Asia, and Australia, accounting for approximately 25% of the total dust deposition worldwide. Dust deposition in the North Pacific, South Pacific, South Ocean, and South Atlantic is 15%, 6%, 6%, and 4% of the global total, respectively.

#### *3.2. Dust Budgets*

The estimated global dust emission ranges from 895 to 8079 Tg yr−1, and the global atmospheric dust load is estimated to be between 8 and 41.65 Tg yr−<sup>1</sup> (Figure 4). Similarly, there is uncertainty regarding the lifetime of the global atmospheric dust load and the ratio of dry to wet deposition. Evidently, there are large discrepancies among the dust models. These discrepancies can be attributed to the following: differences in the description of dust processes in different dust models; different particle size ranges utilised in each model (particle size is a fundamental parameter for simulating soil particle processes and estimating the effect of dust particles on radiation and cloud processes); and different meteorological/climatic data that form a part of the model input.

**Figure 3.** Map of global (**a**) dust emissions, deposition, and (**b**) dust budgets at different regions, estimated by several dust models. Grey and black arrows in (**a**) denote dust emission and deposition (percent of the total dust deposition worldwide), respectively. Horizontal and vertical bars in (**b**) denote annual dust emission (from land regions) and deposition (in oceans), respectively, estimated by different studies. The particle size ranges (r) of dust emissions are Werner et al. [109], 0.1 ≤ r ≤ 219 μm; Luo et al. [108], 0.1 ≤ r ≤ 10 μm; Zender et al. [42], 0.1 ≤ r ≤ 10 μm; Ginoux et al. [33], 0.1 ≤ r ≤ 6 μm; Miller et al. [110], r < 10 μm; Tanaka and Chiba [55], 0.2 ≤ r ≤ 20 μm; Kok et al. [80], 0.2 ≤ r ≤ 20 μm; Checa-Garcia et al. [117] and Aryal et al. [116], multi-model with different maximum dust particle size.

**Figure 4.** Global and annual mean dust budget according to several dust models. A: Takemura et al. [113]; B: Ginoux et al. [94]; C: Chin et al. [114]; D: Tegen et al. [115]; E: Werner et al. [109]; F: Zender et al. [42]; G: Luo et al. [108]; H: Miller et al. [110]; I: Tanaka and Chiba [55]; J: Yue et al. [43]; K–M: Huneeus et al. [13]; N–S: Wu et al. [14]; T: Zhao et al. [107].

In Africa, the estimated rates of dust emission and deposition are 1112 and 685 Tg yr<sup>−</sup>1, respectively (Figure 3). In Asia, 736 Tg of dust are suspended in the atmosphere, and 611 Tg of dust are deposited on the land surface annually. In Australia, the largest dust source in the Southern Hemisphere, the dust emission and deposition rates are 73 and 46 Tg yr<sup>−</sup>1, respectively. However, in North and South America, the dust emission rate is considerably low (approximately 14 Tg yr<sup>−</sup>1), whereas the dust deposition rates are 27 (in North America) and 11 Tg yr−<sup>1</sup> (in South America). Europe has the smallest rate of dust emission (~1 Tg yr−1), but the dust deposition rate is approximately 50 Tg yr−1. It can be concluded that Africa, Asia, and Australia are net dust sources, whereas North and South America and Europe are net dust sinks. Approximately 25% of the global dust emissions are deposited into the open ocean [118]. The deposition of dust is a primary source of micronutrients, such as nitrate, phosphates, and iron, to the sea surface. This has important implications for the CO2 budget; by increasing the iron concentration of the global ocean, dust deposition can increase plankton productivity and thus decrease CO2 concentration in the atmosphere [4,45].

#### **4. Dust Emission and SOC Dynamics**

*4.1. Loss of SOC Due to Dust Emission*

Generally, SOC storage represents the net long-term balance between photosynthesis and respiration in terrestrial ecosystems [29,119]. The global SOC storage is estimated to be approximately 1550 Pg of carbon; this accounts for nearly 54% of the terrestrial carbon pool and is twice the magnitude of the atmospheric carbon pool (760 Pg) [48]. Soil erosion by wind, and the transport and deposition of the eroded material, redistribute SOC across landscapes and regions [20,120]. These physical processes substantially affect the biological mediation of carbon mineralisation in the soil system. Erosion and mobilisation of mineralised carbon could result in a net release of carbon from the soil system to the atmosphere, which may offset carbon sinks in vegetation [120–127]. The fraction of soil carbonates in SOC entering the atmosphere may reduce the intensity of terrestrial carbon sequestration and further increase the CO2 concentration in the atmosphere, which has a positive feedback effect on climate warming [123,124,128].

Dust emission affects SOC by selectively removing fine particles from the soil surface. In this way, the soil evolves toward a coarser texture [119]. Fine soil particles have a high content of stable SOC [122], which directly affects plant growth and soil biological activities, soil air CO2 concentration, soil water regimes, temperature, and respiration, and, therefore, carbon flux to the atmosphere [119,128]. Studies have shown that the SOC loss caused by wind erosion is mainly the active component of SOC and the organic carbon components combined with the soil fine-grained particles [48,54,119,129–131]. Dust emissions can affect soil reflectivity, and thus, soil moisture and temperature, thereby accelerating the in-situ mineralisation of residual SOC [48]. Soil desertification and dust emission reduce the soil's water-holding capacity, root depth, and the efficiency of water and nutrient uptake by plants, thus reducing soil productivity, the amount of organic matter returning to the soil, and the rate of POC formation [24,132]. Moreover, severe dust emission removes the topsoil and exposes the calcium carbonate-rich subsurface soil horizon. This can result in increased emission of CO2 into the atmosphere due to carbon oxidation [48,128].

Despite the significance of dust in the global carbon cycle, wind erosion-induced carbon emissions remain a poorly understood, unquantified component of the global carbon budget. The SOC erosion associated with dust emission in major regions of the world is presented in Table 4 [24,45,120,131]. The difference in an order of magnitude in total min-max dust emission, wind-eroded area, and total SOC erosion across different regions is shown in Table 3. Although several studies have attempted to estimate SOC losses due to dust emission in specific regions, such as China (75 Tg C yr−<sup>1</sup> [7]), Australia (1.59 Tg C yr−1) [46], the United States (34 Tg C yr−1), a small arable catchment in Germany (4.4 g C m−<sup>2</sup> yr−1) [133], and a dryland farming system in Western Australia (3.6 t C ha−<sup>1</sup> yr<sup>−</sup>1) [134], there is significant inconsistency among these results.

#### *4.2. Fate of SOC in Dust*

Soil losses due to wind erosion do not amount to a net loss of SOC; it is a process of SOC migration, in other words, a non-source and non-sink process [135,136]. The fate of the SOC involved in dust dynamics is determined by a series of complex interactions. As these interactions constitute a dynamic process, it is difficult to accurately estimate the ultimate fate of wind-eroded SOC. In general, the fate of the SOC is mobilised, as dust may include [134]: (1) proximal deposition, from creep and saltation, in the range of tens of meters; (2) deposition in lakes and rivers; (3) transport, in the form of dust, to a distant system; (4) release to the atmosphere by oxidation; and (5) variation in SOC with dust size.

The net change in SOC stocks reflects the balance between carbon sequestration and soil carbon emission. Some studies have indicated that the main losses in the process of dust emission are mainly the active organic carbon of SOC and the organic carbon combined with soil fine-grained components [137–140]. Soil active organic matter components are the habitat and survival matrix of soil microorganisms. Therefore, the loss of SOC caused by dust emission can significantly reduce soil biological activity. The decrease in soil biological activity and the change in soil structure and water-holding capacity caused by wind erosion can significantly change the biological process of carbon mineralisation and result in the net release of carbon from the soil system to the atmosphere. Therefore, from the perspective of the global carbon balance, more attention should be paid to the loss of mineralised SOC due to wind erosion. The mechanisms of carbon mineralisation during the migration and deposition of wind-eroded material are yet to be determined. This raises the question of how to estimate the effect of dust emissions on the global carbon balance. The current estimates of SOC loss usually ignore the redistribution of SOC generated by dust emissions; consequently, they overestimate the contribution of SOC erosion to atmospheric CO2. The fate of wind-eroded SOC is still discussed in merely qualitative terms. Quantitative analysis is limited to smaller space-time scales. In-depth study and quantification of SOC in dust, especially the fate of wind-eroded SOC in the global dust cycle, is essential to quantify the release of CO2 from SOC dust to the atmosphere, the contribution of SOC deposition to downwind carbon sinks, and the effect of dust processes on the global carbon balance.

#### **5. Conclusions**

Advances in dust modelling in the past five decades have changed the requirements for input data, and increased model complexity and the availability of model outputs. Owing to the diversity of the required inputs, hybrid observation methods (integrating multiple observation methods) should be adopted to provide dust models with input data. Although the development of dust models has progressed considerably over the past 30 years, the model simulation results are still replete with uncertainties. Dust models developed in a specific region require careful calibration when used to other regions. It is only possible to simulate dust processes in an area after the model's parameters have been localised with the use of observation data. There are no universally accepted parameters for dust models in different regions/countries. Therefore, it is necessary to develop a set of parameters for different regions. It is recognised that anthropogenic activities can also induce dust emissions; as such, they are non-negligible contributors to global dust concentrations [140–145]. However, all models reviewed in this study simulated 'natural' or indirect anthropogenic (e.g., cropland and pastureland) dust processes, neglecting the contribution of direct anthropogenic dust (e.g., city construction and transportation). This leads to considerable uncertainties in estimating dust emissions. Therefore, to improve the accuracy of dust emission simulations, the consideration of anthropogenic dust emissions is imperative [146].

SOC loss due to wind erosion is a key component of the global carbon cycle. A better understanding of the role of dust processes in the global SOC flux and carbon budget is needed. Although it is recognised that SOC is transported and redistributed by dust processes, SOC cycling schemes used in land surface models (LSMs) typically only consider the effects of net primary production and heterotrophic respiration. Current estimates of SOC loss results in significant underestimations due to the omission of the effects of dust emission. Moreover, the dust emission flux observation does not include the measurement of SOC concentrations; there is a lack of SOC concentration in different dust sizes, and how dust emission is directly linked to SOC erosion is not well represented. It is necessary to explore the various effects of dust processes on SOC pools, mineralisation rates, and SOC emission to the atmosphere in dust source regions, and on the enrichment of SOC in deposition regions. Currently, although some Earth System Models have the ability to simulate the effects of mineral dust deposition on biogeochemistry [78,147,148], most dust models are limited to estimating dust emission and deposition and do not consider the effects of dust on the global carbon cycle. Similarly, the current carbon cycle models ignore the effects of SOC movement caused by dust processes. Therefore, representing the linkages between dust processes and the carbon cycle in both dust and carbon cycle models is essential.

**Author Contributions:** Conceptualisation, methodology, and software; W.C., H.M., H.S. and H.Z.; formal analysis, investigation, and resources; W.C. and H.M.; data curation and writing—original draft preparation; H.Z. and W.C.; writing—review and editing and visualisation; H.S., H.M. and W.C.; supervision, project administration, and funding acquisition; H.S. and H.Z. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the National Natural Science Foundation of China, grant numbers 41401107 and 32130066; the Research Program of Henan University, China, grant number 2015YBZH001; and the Foundation of Henan University, grant number 2015YBZR019.

**Acknowledgments:** We thank Webb and McGowan for providing a meaningful basis for this review. We also thank Yang Zhang from Northeastern University for providing insightful suggestions for this study.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


38. Woodruff, N.P.; Siddoway, F. A wind erosion equation. *Soil Sci. Soc. Am. J.* **1965**, *29*, 602–608. [CrossRef]

