An Evaluation of the Dust Emission Characteristics of Typical Underlying Surfaces in an Aeolian Region in the Middle Reaches of the Yarlung Zangbo River on the Qinghai–Tibet Plateau

Abstract

Some of the most severe aeolian damage occurs along the middle reaches of the Yarlung Zangbo River in Tibet. Dust emission amounts (DEAs) are often used to assess aeolian damage; however, the research on DEAs in this area is currently almost blank. This article uses field-measured wind speed data from 2021 to 2022 in the Shannan wide valley area, combined with the Gillette dust emission estimation model to quantitatively determine the contributions of three surface types (riverbank quicksand area, foothill sand dunes, and the river floodplain vegetation area) to DEAs in the research area. The influence of surface characteristics on DEAs is analyzed and discussed. The results show the following: (1) The threshold friction velocity ($u_{*t}$) in the riverbank quicksand area, foothill sand dunes, and the river floodplain vegetation area is 30.6 cm/s, 71.2 cm/s, and 85.6 cm/s, respectively, the threshold velocity ($u_t$) is 6.1 m/s, 7.0 m/s, and 7.5 m/s, respectively, and the vegetation area is 2.8 times and 1.3 times that of the quicksand area, respectively. (2) The DEAs were in the following order: the riverbank quicksand area (652.9 t/km²) > foothill sand dunes (326.5 t/km²) > the river floodplain vegetation area (107.8 t/km²), the riverbank quicksand area is about 6.1 times that of the river floodplain vegetation area, and DEAs are a significant seasonal distribution: winter (44.7%) > spring (28.3%) > autumn (15.7%) > summer (11.3%). (3) The DEAs from the dusty weather were in the following order: blowing sand (60.2%) > sandstorms (28.6%) > gusty winds (11.2%). (4) The DEAs increase with the increase in the average wind speed greater than 6.1 m/s, but the increase rate is obviously different, which showed that Changguo and Azha are greater than Sangyesi, Duopazhang, Sangri, and Senburi. At approximately the same average wind speed greater than 6.1 m/s, the DEAs in the quicksand area are much greater than in the vegetation area.

Highlights:

What are the main findings?

• The contributions of sandstorms, blowing sand, and gusty winds to dust emissions were determined;
• Equations were established for different amounts of surface dust emissions and wind speeds >6.1 m/s.

1. Introduction

Desertification is one of the most prominent ecological and environmental problems in Tibet. It not only causes serious damage and harm to the ecological environment in Tibet and people’s production and life, but it also has an important impact on the ecological environment in China, Southeast Asia, and even the whole world [1]. DEAs are the main content of desertification, sand landforms, and sand control [2,3,4,5,6].

Wind is an important driving force of dust emission. Dust emission only occurs when the wind speed (u) or the friction velocity ($u_{*}$) reaches a critical value, known as the threshold velocity ($u_t$) or the threshold friction velocity ($u_{*t}$). In this study, these are collectively referred to as emission thresholds. The emission threshold is not only a key indicator for determining whether dust emission can occur, but also a core parameter in soil erosion and sandstorm forecasting models, determining the accuracy of dust emission intensity and dust flux calculations [2,7]. Laboratory experiments by Ho et al. [8] and numerical simulations by Kamath et al. [9] have shown that aeolian sand flux over hard surfaces and poorly erodible soils scales differently with wind shear velocity from the scaling over fully erodible soils (mobile sand bed). When the ground is non-erodible, the saltation layer expands, the coupling between particle trajectories and the wind weakens, and a cubic scaling between the flux and the wind shear velocity is obtained, which contrasts to the quadratic scaling over fully mobile beds and low $u_{*}$ [10]. However, the scaling of dust emission flux is still poorly understood and depends on many natural factors, such as vegetation, etc.

A significant amount of research has been conducted on the dust emission amounts [2,4], emission thresholds [11], sand landform formation and evolution process [12], the cause of desertification [13], and reasonable layout of the sand prevention project [14] in low-altitude deserts such as the Gobi and desertification-affected areas. The middle reaches of the Yarlung Zangbo River is the most important political, economic, and cultural core area in Tibet, with its climatic conditions during wind and dry periods [1], widespread loose sedimentary deposits [15,16], exposed river islands and floodplains along the river during the dry seasons of winter and spring, and human-induced damage to surface vegetation and soil structure [17], causing very severe sandstorm disasters in this area [1,18]. However, the research on the sandstorm problems in this area is relatively weak [15]. At present, the research on the sandstorm disasters in the middle reaches of the Yarlung Zangbo River mainly focuses on the following aspects: (1) Sediment source. The study concluded that the bottom of the valley (river floodplain) is the source of sandstorm disasters [16,17,18,19,20]. (2) Sediment nutrients. Different types of sediments have significant spatial heterogeneity in regions, with the highest river soil nutrient content and the lowest dune nutrient content [21]. (3) Aerodynamic roughness length. The vegetation areas $u_{*}$ and $z_0$ are about two times and 150 times that of the quicksand area, which is helpful for aeolian disaster assessment and sandstorm forecast and early warning in this area [22]. (4) Sand landforms. The sand landforms are complex and diverse [18], but the most typical ones are climbing dunes.

However, the above studies cannot well clarify the mechanism of aeolian motion on different surfaces; the main reason is that there are few comparative studies on the emission threshold and dust emission amounts of different surfaces, which leads to the lack of theoretical basis for the reasonable layout of sand and desertification control measures, and the effect of aeolian disaster preventions is not obvious. In view of this, this study selected the area from Gongga County to Zedang Town in the middle reaches of the Yarlung Zangbo River basin, where aeolian disasters are most severe, and selected three main surface types riverbank quicksand area (without any vegetation, Figure 1A), foothill sand dunes (vegetation coverage <35%, Figure 1B), river floodplain vegetation area (vegetation cover >60%, Figure 1C), and six field observation sites for the measured study of emission threshold and dust emission amounts of different surface types. Through research, we aim to plateau different surface type aeolian movement processes and mechanisms to provide theoretical basis, at the same time, according to the trend of increasing dusty weather in the area in recent years and the field investigation, and for the change of the Tibet autonomous region for poverty alleviation and the needs of engineering construction, to strengthen the Yarlung Zangbo River basin aeolian disaster prevention and control of ideas.

2. Data and Methods

2.1. Study Area

The study area was located in the middle reaches of the Yarlung Zangbo River (Figure 1). This region serves as the political, economic, and cultural core of Tibet and is one of the areas where sandstorm disasters are most severe. The sand landform is superimposed on the river landform, and the formation process is influenced by the action of wind and water [1], with completely different characteristics of sandstorm activity and geomorphic formation and an evolution process different from the plain area. Wind power is the most important dynamic condition affecting wind and sandstorm activities. The low vegetation coverage rate causes the surface material to be easily eroded [23,24]. During the winter and spring seasons, the water level of the Yarlung Zangbo River decreases, exposing a large number of river islands and floodplains, creating new sand sources and resulting in severe aeolian disasters [15].

The study area has a semi-arid plateau temperate climate, with cool and rainy summers and dry, cold, and windy winters, tending towards warm and humid conditions [25]. The prevailing wind direction in the study area is westerly, with an average annual wind-blown sand day count of 14.9 to 54.9 d. Strong winds are the primary driving force behind blowing sand and sandstorm weather in the Yarlung Zangbo River basin [26]. The annual statistics are as follows: the average altitude is 3650 m, the average temperature is 8.7 °C, the average precipitation is 378 mm, and the annual variation curve shows a unimodal pattern, with distinct dry (winter and spring) and rainy (summer) seasons. The average wind speed is approximately 2.6 m/s, with stronger winds during winter and spring, and weaker winds during summer and autumn [15].

2.2. Meteorological Observations

Riverbank quicksand areas, foothill sand dunes, and river floodplain vegetation areas are the main sand source areas for aeolian disasters in the middle reaches of the Yarlung Zangbo River. Therefore, this study selected three surface types and six observation points (Changguo, Azha, Sangyesi, Duopozhang, Sangri, and Senburi) for the field measurements of DEAs. Changguo and Azha are riverbank quicksand areas (Figure 1A); Sangyesi and Duopozhang are foothill sand dunes (Figure 1B); Sangri and Senburi are river floodplain vegetation areas (Figure 1C). The topographies of the three measurement points were relatively flat. A two-dimensional ultrasonic anemometer (Gill Instruments, UK; wind speed range: 0–30 m/s with a resolution of 0.01 m/s; wind direction range: 0°–359° with a resolution of 1°) was used to measure wind speeds at different heights. Wind-speed sensors were installed at heights of 20, 40, 100, 200, and 400 cm. The soil moisture probe was buried at 2.5 cm below the surface. The observation period is from 2021 to 2022, and all the data in this paper are minute data.

2.3. Calculation of DEAs by Dusty Weather

The DEAs during dusty weather (sandstorm, blowing sand, and gusty wind) are the sum of vertical dust fluxes during various periods with occurrence of dusty weather. The amount was determined using the calculation scheme proposed by Gillette et al. [27]:

$$F=\left\{\begin{array}{l}{C}_2{u}_{*}^4(1-{\displaystyle \frac{u_{*t}}{u_{*}}})\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}{u}_{*}>u_{*t}\\ 0\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}{u}_{*}<u_{*t}\end{array}\right.,$$

(1)

where F is the dust emission amounts (DEAs) (g/cm²/s), C₂ is a constant (1.4 × 10⁻¹⁵ kg/m³/s), and $u_{*}$ and $u_{*t}$ are the friction speed (cm/s) and critical friction speed (cm/s), respectively.

2.4. Aerodynamic Roughness Length

The aerodynamic roughness length ($z_0$) is the height at which the near-surface wind speed decreases to zero [28], and its value is closely related to the characteristics of the underlying surface. Using the wind speed data of 0.4, 1.0, 2.0, and 4.0 m at a gradient tower from 2010 to 2022, the aerodynamic roughness length ($z_0$) of the study area was calculated using the wind-speed profile equation under neutral atmospheric conditions (Equation (2), Figure 1).

$$u_z=\left({\displaystyle \frac{u_{*}}{k}}\right)ln\left(z\right)-\left({\displaystyle \frac{u_{*}}{k}}\right)ln\left(z_0\right).$$

(2)

where z is the measurement height (cm); $u_z$ is the horizontal component of the wind speed at the measurement height (cm/s); $z_0$ is the aerodynamic roughness length (cm); k is the von Kármán constant (0.4). Wind speed data were selected only when the wind speed at a height of 2.0 m was ≥4.0 m/s, at which point the influence of temperature on the wind speed profile can be neglected, and the atmospheric conditions can be approximated as neutral [29].

By using the aforementioned formulas, this study calculated values for $z_0$ in Changguo, Azha, Sangyesi, Duopozhang, Sangri, and Senburi, which are 0.047 cm, 0.11 cm, 3.5 cm, 4.2 cm, 5.6 cm, and 6.3 cm, respectively. For riverbank quicksand areas (Changguo, Azha), the average value of $z_0$ is 0.078 cm; for foothill sand dunes (Duopozhang, Sangyesi), the average value of $z_0$ is 3.85 cm; for river floodplain vegetation areas (Sangri, Senburi), the average value of $z_0$ is 5.95 cm (Figure 2).

This calculated values for $u_{*}$ in Changguo, Azha, Sangyesi, Duopozhang, Sangri, and Senburi are 54.6 cm/s, 63.2 cm/s, 77.3 cm/s, 82.4 cm/s, 91.4 cm/s, and 102.4 cm/s, respectively. For riverbank quicksand areas (Changguo, Azha), the average value of $u_{* }$ is 58.9 cm/s; for foothill sand dunes (Duopozhang, Sangyesi), the average value of $u_{* }$ is 79.8 cm/s; for river floodplain vegetation areas (Sangri, Senburi), the average value of $u_{* }$ is 96.9 cm/s.

2.5. Calculation of Emission Threshold

The emission threshold includes the threshold velocity ($u_t$) and the threshold friction velocity ($u_{*t}$), which are key parameters for determining whether dust emission occurs. These two parameters can be calculated using Equation (3)

$$u_t={\displaystyle \frac{u_{*t}}{k}}ln\left({\displaystyle \frac{z}{z_0}}\right)$$

(3)

where z is the measurement height (cm). The wind speed data used in this paper are 2 m high. Variations in the emission threshold are closely related to atmospheric and soil conditions, especially soil conditions, such as vegetation, soil moisture, and soil particle size [1,30,31]. The parameterization scheme developed by Shao and Lu [32] is widely used in sandstorm models. This scheme is expressed by the following equation:

$$u_{*t}(d, \lambda , \omega )=u_{*ts}(d\left)f_{\lambda }\right(\lambda \left)f_{\omega }\right(\omega )$$

(4)

where d (cm) is the sand particle size (

Table 1. The introduction of six sand transport stations.

Observation Point	Subfloor Characteristics	Particle Size (µm)
Changguo	The observation station is located about 10 m away from the riverbank. There are no plants between the station and riverbank. The ground consists of loose flowing sand.	236
Azha	About 150 m away from the river channel, there are sporadic artificial protective forests between the observation station and riverbank. On the west side, there is a crescent-shaped sand dune chain of about 20 m in width, which consists of mobile sand dunes.	184
Sangyesi	On the mountain foothills approximately 1.2 km from the river channel, the observation station is surrounded by natural vegetation with plant height <0.8 m and canopy coverage <30%. The area is a mobile sandy terrain.	180
Duopozhang	On the river floodplain, approximately 1.0 km from the river channel, there is an observation station surrounded by artificial protective forests. The vegetation has a height of 2.0 m and a spacing of 5 m. The area is characterized by mobile sandy terrain.	150
Sangri	The observation station is located about 15 m away from the river channel. There are no plants within 1.5 m around the station. On the eastern side, there is a crescent-shaped sand dune that is 1.0 m high. The surface of the ground is alluvial sediment and contains physical crust.	112
Senburi	The riverbed is composed of loose sediment. Within a 10 m radius of the observation station, there is scattered vegetation with a height of around 1.0 m, and the vegetation coverage is over 60%.	82

). Changguo, Azha, Sangyesi, Duopozhang, Sangri, and Senburi are 236 µm, 184 µm, 180 µm, 150 µm, 112 µm, and 82 µm, respectively. For riverbank quicksand areas (Changguo, Azha), the average value of d is 210 µm; for foothill sand dunes (Duopozhang, Sangyesi), the average value of d is 165 µm; for river floodplain vegetation areas (Sangri, Senburi), the average value of d is 97.0 µm.

${ u}_{*ts}$ (cm/s) is the threshold friction velocity of sand particles of size under ideal conditions; $\lambda$ and $\omega$ (m³ /m³) are the roughness density of the roughness elements and soil moisture, respectively; and $f_{\lambda }(\lambda$) and $f_{\omega }(\omega$) are the correction functions for surface roughness elements and soil moisture, respectively [33]. All correction functions were ≥1. Shao and Lu [31] derived the following simple expression for an ideal threshold friction velocity ($u_{*ts}$):

$$u_{*ts}\left(d\right)={\left[A_N({\displaystyle \frac{\sigma gd}{\rho }}+{\displaystyle \frac{\epsilon }{\rho d}})\right]}^{0.5}$$

(5)

where $\sigma$ is the sand particle density (2560 kg/m³); $\rho$ is the air density (0.85 kg/m³); $g$ is the gravitational acceleration (9.81 m/s²); and $A_N$ and $\epsilon$ are constants ($A_N$ = 0.0123, $\epsilon$ = 0.165 g/s²), which account for the magnitude of the interparticle cohesive forces and are empirically determined by a wind tunnel experiment [34].

The surface correction equation for roughness elements is as follows [35]:

$$f_{\lambda }\left(\lambda \right)=\sqrt{(1-{\sigma }_r{m}_r\lambda )(1+{\beta }_r{m}_r\lambda )},$$

(6)

where $m_r$ ($m_r$ = 0.5) is the parameter to account for nonuniformity in the surface stress, ${\sigma }_r$ (${\sigma }_r$ = 1) is the ratio of the basal to frontal area of the roughness element, ${\beta }_r$ (${\beta }_r$ = 90) is the ratio of the drag coefficient for a single roughness element to that of the surface without roughness elements, $\lambda$ is the frontal area index and $\lambda$ = −0.35 ln (1 − a), and a is the vegetated fraction (the value for riverbank quicksand areas was about 0%, the value for foothill sand dunes was about 12%, and the value for river floodplain vegetation areas was about 63% during the study period).

$f_{\omega }\left(\omega \right)$ is the correction equation for soil moisture [32]:

$$f_{\omega }(\omega )=\left\{\begin{array}{l}{e}^{22.7\omega }\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\omega <0.03\\ e^{95.3\omega -2.03}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\omega >0.03\end{array}\right.$$

(7)

where $\omega$ (m³/m³) is the volumetric soil moisture. The volumetric soil moisture for the riverbank quicksand areas, foothill sand dunes, and river floodplain vegetation areas is 0.25 m³/m³, 0.009 m³/m³, and 0.14 m³/m³, respectively.

By using the aforementioned formulas, this study’s calculated values for $u_{*t}$ in Changguo, Azha, Sangyesi, Duopozhang, Sangri, and Senburi are 28.4 cm/s, 32.8 cm/s, 67.3 cm/s, 75.1 cm/s, 82.8 cm/s, and 88.4 cm/s, respectively. For the riverbank quicksand areas (Changguo, Azha), the average value of $u_{*t}$ is 30.6 cm/s; for foothill sand dunes (Sangyesi, Duopozhang), the average value of $u_{*t}$ is 71.2 cm/s; for the river floodplain vegetation areas (Sangri, Senburi), the average value of $u_{*t}$ is 85.6 cm/s.

This study’s calculated values for $u_t$ in Changguo, Azha, Sangyesi, Duopozhang, Sangri, and Senburi are 5.9 m/s, 6.2 m/s, 6.8 m/s, 7.2 m/s, 7.3 m/s, and 7.6 m/s, respectively. For the riverbank quicksand areas (Changguo, Azha), the average value of $u_t$ is 6.1 m/s; for foothill sand dunes (Sangyesi, Duopozhang), the average value of $u_t$ is 7.0 m/s; for the river floodplain vegetation areas (Sangri, Senburi), the average value of $u_t$ is 7.5 m/s.

3. Results and Discussion

3.1. Characteristics of Total DEAs Variations

The characteristics of the total annual DEAs from 2021 to 2022 are shown in Figure 3. From 2021 to 2022, the total annual DEAs showed a significant decreasing trend at all observation points except for the river floodplain vegetation area (Senburi). The foothill sand dunes (Sangyesi, Duopozhang) had the largest average annual decrease rate of 23.5%, followed by the riverbank quicksand areas (Changguo, Aza), with a decrease rate of 14.1%. In contrast, the river floodplain vegetation area (Senburi) showed an increasing trend in annual total DEAs, with a growth rate of only 5.6%.

In addition, the annual average DEAs were highest in the riverbank quicksand areas (Changguo: 688.4 t/km²; Azha: 617.5 t/km²), followed by the foothill sand dunes (Sangyesi: 396.5 t/km²; Duopozhang: 256.4 t/km²), and the river floodplain vegetation areas had the lowest values (Sangri, 147.8 t/km²; Senburi, 67.9 t/km²). Overall, the annual average DEAs exhibit the following trend: Changguo > Azha > Sangyesi > Duopozhang > Sangri > Senburi. The reason for these differences is that the riverbank quicksand areas (Changguo, Aza) have loose wind-blown sediment on the surface with sufficient sand sources, making them highly susceptible to erosion, whereas the river floodplain vegetation areas (especially in Senburi, where the surface is covered by vegetation, and Sangri, where there is a physical crust) have limited sand sources. This reduces the near-surface wind speed and makes it difficult to release sand and dust particles. This conclusion is consistent with the findings of Luo et al. [36].

The monthly variations in the DEAs are shown in Figure 4. As shown in Figure 4, the riverbank quicksand areas (Changguo and Aza), foothill sand dunes (Sangyesi and Duopozhang), and river floodplain vegetation areas (Sangri and Senburi) exhibited significant seasonal variations. For the riverbank quicksand areas, the highest dust emissions occurred in winter, with a value of 260.5 t/km², followed by spring, with a rate of 180.4 t/km². The DEAs in summer and autumn are similar, at 118.9 t/km² and 93.1 t/km², respectively. In general, the DEAs were in the order of winter (39.9%) > spring (27.6%) > autumn (18.2%) > summer (14.3%). For foothill sand dunes, the highest DEAs occurred in winter (134.5 t/km²), followed by spring (105.6 t/km²). The DEAs in summer and autumn are similar, at 54.9 t/km² and 31.4 t/km², respectively. In general, the DEAs were in the order of winter (41.2%) > spring (32.3%) > summer (16.8%) > autumn (9.6%). For river floodplain vegetation areas, the highest DEAs occurred in winter, with a value of 48.2 t/km², followed by spring, with rate of 30.5 t/km². The DEAs in summer and autumn are similar, at 16.9 t/km² and 12.2 t/km², respectively. In general, the DEAs were in the order of winter (44.7%) > spring (28.3%) > autumn (15.7%) > summer (11.3%).

These findings differ from those of the Taklimakan and Gobi deserts [11,37,38]. The latter are influenced by the unique weather system in the Taklamakan Desert, known as the “Eastern Irrigation Weather”, which leads to frequent cold air activities and strong winds and sandstorms in spring [39], resulting in the highest dust emission amounts during that season. In summer, the amount of dust emission was slightly lower because of the well-developed summer monsoon winds. In autumn and winter, the temperature decreased and the occurrence of dusty weather was significantly reduced, resulting in the lowest dust emission amounts. In contrast, the study area is located in the middle reaches of the Yarlung Zangbo River, influenced by seasonal runoff. During winter and spring, when the water level is low, extensively exposed river bars and floodplains have dry and loose surface sediments [15,16], which are easily transported northward towards the piedmont and slopes under strong airflow, leading to higher dust emission amounts in winter and spring. In summer, precipitation increases, water levels rise, and the sand source decreases, resulting in the lowest dust emissions.

The four-season daily total DEAs variations in this study are shown in Figure 5. Although the daily total DEAs in the four seasons were similar throughout the year, there were some differences. In winter, spring, summer, and autumn, the dust emissions during the daytime (08:00–20:00) accounted for 82.4%, 76.3%, 76.9%, and 70.3% of the total daily dust emissions, respectively, which was higher than the dust emissions during the nighttime (20:00–08:00). The maximum DEAs values are 2.6 t/km², 1.2 t/km², 0.8 t/km², and 0.5 t/km² for winter, spring, summer and autumn, respectively, with a difference of approximately 5.2 times between the maximum and minimum values. The occurrence times of the maximum DEAs values are 14:10, 16:20, 16:25, and 15:20. This result is consistent with the findings in the Taklimakan Desert and Inner Mongolia Desert [40,41].

3.2. Characteristics of DEA Variations by Dusty Weather

The characteristics of daily DEAs variations due to dusty weather (blowing sand, sandstorms, and gusty winds) in this study are shown in Figure 6. For the riverbank quicksand areas, the daily average DEAs of blowing sand, sandstorms, and gusty winds are 365.6 t/km², 180.2 t/km², and 78.5 t/km², respectively, accounting for 58.6%, 28.9%, and 12.5%, respectively, of the total DEAs throughout the year. For foothill sand dunes, the daily average DEAs of blowing sand, sandstorms, and gusty winds are 189.4 t/km², 87.8 t/km², and 36.9 t/km², respectively, accounting for 60.3%, 27.9%, and 11.7%, respectively, of the total DEAs throughout the year. For the river floodplain vegetation areas, the daily average DEAs of blowing sand, sandstorms, and gusty winds are 65.8 t/km², 30.8 t/km², and 10.0 t/km², respectively, accounting for 61.7%, 28.9%, and 9.3%, respectively, of the total DEAs throughout the year.

In general, the observed trend was blowing sand (60.2%) > sandstorms (28.6%) > gusty winds (11.2%), indicating that the research area was mainly characterized by blowing sand.

The characteristics of the monthly DEAs variations due to dusty weather (sandstorms, blowing sand, and gusty winds) in the study area are shown in Figure 7. As shown in Figure 7, the riverbank quicksand areas, foothill sand dunes, and river floodplain vegetation areas exhibited significant seasonal variations in DEAs. The highest DEAs occur during winter, with values of 240.4 t/km², 177.4 t/km², and 57.6 t/km², respectively, accounting for 38.3%, 48.4%, and 44.9% of the total annual DEAs, respectively. The second highest DEAs occur during spring, with values of 194.7 t/km², 108.3 t/km², and 38.5 t/km², respectively, accounting for 31.0%, 29.6%, and 30.0% of the total annual DEAs, respectively. The DEAs are relatively lower during summer and autumn, with values of 104.1 t/km², 30.4 t/km², and 20.3 t/km², and 88.2 t/km², 50.1 t/km², and 11.8 t/km², respectively, accounting for 16.6%, 8.3%, and 15.8%, and 14.1%, 13.7%, and 9.2% of the total DEAs, respectively.

The spatial differences in the DEAs reflect the comprehensive effects of wind power, sand source, and the underlying surface on the DEAs. During the observation period, the monthly average wind speed differences at the riverbank quicksand areas, foothill sand dunes, and river floodplain vegetation areas varied slightly, ranging from 0.6–1.9 m/s, 0.7–1.6 m/s, and 0.3–2.0 m/s, respectively. However, significant differences were observed in average wind speeds >6.1 m/s, >7.0 m/s, and >7.5 m/s, as well as the cumulative wind speeds >6.1 m/s, >7.0 m/s, and >7.5 m/s at the riverbank quicksand areas, foothill sand dunes, and river floodplain vegetation areas, respectively (Figure 8 and Figure 9). This indicates that the average wind speed has limitations in characterizing the DEAs in the study area. The DEAs in the riverbank quicksand areas (Changguo, Azha), foothill sand dunes (Sangyesi, Duopozhang), and river floodplain vegetation areas (Sangri, Senburi) showed an increasing trend with increasing average wind speeds >6.1 m/s, >7.0 m/s, and >7.5 m/s, respectively. However, the rates of increase differed noticeably. Specifically, the riverbank quicksand areas showed a higher rate of increase than foothill sand dunes and river floodplain vegetation areas. When average wind speeds >6.1 m/s, >7.0 m/s, and >7.5 m/s are approximately equal, the DEAs in the riverbank quicksand areas are much higher than in foothill sand dunes and river floodplain vegetation areas, indicating that the differences in DEAs between the riverbank quicksand areas, foothill sand dunes, and river floodplain vegetation areas in relation to average wind speeds >6.1 m/s, >7.0 m/s, and >7.5 m/s, respectively, reflect the influence of vegetation cover and sand sources on DEAs.

4. Discussion

4.1. The Impact of Underlying Surface Characteristics on DEAs

The DEAs in the riverbank quicksand areas, foothill sand dunes, and river floodplain vegetation areas showed a good linear trend with increasing average threshold velocity >6.1 m/s, >7.0 m/s, and >7.5 m/s, respectively. However, the DEAs varied with changes in the average threshold velocity (Figure 8 and Figure 9). This indicates that the sand source and vegetation in the study area are important factors contributing to spatiotemporal variations in DEAs. The more abundant the sand source, the greater the DEAs. Conversely, the higher the vegetation cover, the smaller the DEAs. The Changguo, Sangri, and Senburi observation points are located on riverbank quicksand areas. The water level of the Yarlung Zangbo River has been decreasing since October, leading to exposure of the riverbed and an increase in the sand source [15]. However, when the water level decreased in Senburi and Sangri, the exposed surface was covered with river pebbles, and the sand source did not increase significantly; therefore, the DEAs did not show a significant increase. The sand sources of Changguo and Azha both increased, while the average wind speed also increased. The DEAs increased from 14.6 t/km² in October to 78.9 t/km² in November, with an increase of about 2.3 times. Sangyesi and Duopozhang are located on foothill sand dunes; therefore, during the observation period, the sand source remained unchanged. Therefore, the DEAs were mainly controlled by the threshold velocity and increased with increasing threshold velocity (Figure 8 and Figure 9).

Vegetation cover was the main factor affecting the DEAs. Senburi and Sangri had the smallest DEAs (Figure 3), mainly because the Senburi observation point was dominated by annual herbaceous plants, and the vegetation coverage during the observation period was the highest (coverage > 60%). Changguo and Azha had no vegetation on their surfaces, resulting in the highest DEAs values. The Sangyesi observation point has a vegetation coverage of <30% and a height of <0.8 m, resulting in lower DEAs compared to the vegetation-less areas of Changguo and Azha. The Duopozhang observation point has an artificial protective forest with a vegetation height of approximately 2.0 m, which resulted in lower DEAs than that of Sangyesi.

4.2. The Impact of Key Meteorological Factors on DEAs

A precipitation weather process was selected in the study area on 17–18 April 2021, and the impact of precipitation on DEAs was analyzed (Figure 10). From 0:00 to 6:00 on April 17, there was no precipitation in the study area, and the wind speed increased from 4.6 m/s to 16.7 m/s, with the DEAs greater than zero. From 7:00 to 10:00 on April 17, with the decrease in wind speed, the dusty weather ended. Starting at 11:00 on the 17th, precipitation weather occurred in the study area, soil moisture gradually increased, reaching the highest value of 0.13 m³/m³ at 03:00 on the 18th, then with the interruption of precipitation weather, soil moisture gradually decreased, reaching the minimum value of 0.03 m³/m³ at 14:00 on the 18th. During the period of precipitation weather (from 11:00 on the 17th to 06:00 on the 18th), the soil moisture was relatively high, and the threshold velocity ($u_t$) was greater than the measured wind speed, so the DEAs were zero. This indicates that precipitation increases the emission thresholds by affecting soil moisture, significantly inhibiting the DEAs. From 20:00 to 23:00 on the 18th, strong wind weather occurred in the study area, with the wind speed increasing from 5.3 m/s to 14.1 m/s At this time, the threshold velocity ($u_t$) was lower than the measured wind speed, with the DEAs greater than zero. However, the DEAs were significantly lower than under equivalent wind conditions before precipitation. This indicates that the soil moisture caused by precipitation has a longer duration in inhibiting DEAs.

The above analysis showed that the increase in soil moisture caused by precipitation in the study area had a significant inhibitory effect on DEAs, and the inhibitory effect lasted for 14 h, lower than 24 h in the Taklimakan Desert [41]; still, this study only analyzed the effect of a typical precipitation process on DEAs. It is not clear about the difference of precipitation on DEAs, and further analysis is needed.

5. Conclusions

(1) In the study area, the threshold friction velocity ($u_{*t}$) in the riverbank quicksand areas, foothill sand dunes, and the river floodplain vegetation areas is 30.6 cm/s, 71.2 cm/s, and 85.6 cm/s, respectively, the threshold velocity ($u_t$) is 6.1 m/s, 7.0 m/s, and 7.5 m/s, respectively, and the vegetation area is 2.8 times and 1.3 times that of the quicksand area, respectively. Accordingly, the focus of sandstorm disaster prevention and control should be in the riverbank quicksand area.

(2) The largest DEAs in the riverbank quicksand areas are 652.9 t/km², followed by the foothill sand dunes, 326.5 t/km², and the river floodplain vegetation areas show the smallest, 107.8 t/km²; the DEAs showed significant seasonal distribution—winter (44.7%) > spring (28.3%) > autumn (15.7%) > summer (11.3%)—indicating the sandstorm activity in winter and spring.

(3) The DEAs from the dusty weather were in the following order: blowing sand (60.2%) > sandstorms (28.6%) > gusty winds (11.2%), indicating dusty weather in the research area is mainly blowing sand.

(4) The response mode of the river floodplain vegetation areas and the riverbank quicksand areas is different to the threshold velocity ($u_t$). The DEAs in the vegetation areas increase slowly with the threshold velocity, while the DEAs in the quicksand areas increase slowly at the low wind speed. When the wind speed increases to the threshold velocity (6.1 m/s), the DEAs increase rapidly.