Background & Summary

Eolian landforms are the products of wind movement of sand, and sediments are the material basis for the formation and development of these landforms. Movement of sand by wind is influenced by single-particle characteristics such as particle size, shape, and density, and the characteristics of particle clusters, including grain-size distribution, aligned stacking form, porosity, and cohesion1. The action of wind or water can continually change the grain shape and mechanical composition of sediments, meaning that these sediments can contain partial or even critical information about the formation and evolution of landforms2. Desert sediments have been extensively studied, and most landscape-scale research has focused on the stability of typical dune surfaces3, sediment discrimination4, ecological restoration of sandy areas5, and movement of sand by wind6. Research on surface sediments has mainly considered the composition, distribution, and regional variation of grain size7; geology and sedimentary processes8; and geochemical composition9.

Dune sediments are an important record of the depositional environment during dune accretion and formation10. Wind is the main externally applied force during the evolution of deserts or dunes. In these settings, surface sediments are transported and deposited by wind, resulting in the formation of a variety of dune morphologies; conversely, dunes can influence local wind speed and direction11. Grain-size distribution is the most readily available information for eolian geomorphologic studies and is often used to elucidate the origin of sediments. Sediment grain size is an important parameter reflecting the formation and evolution of eolian landscapes, soil erosion by wind, dust storms, and desertification processes12, and is crucial in identifying depositional environments and determining the mode of particle transport and dynamic conditions13. Previous studies of the world’s deserts have extensively considered the grain-size distribution of sediments and regional variations of desert surface sediments, and have obtained information on the material composition of different deserts from their grain-size characteristics14,15,16. Discussing the grain-size characteristics of desert surface sediments is important for reconstructing eolian environments and exploring the sources of desert sand.

The Kubuqi Desert is the seventh largest desert in China, with a large proportion of flowing sand dunes. As a result of the combined influence of topography, wind, and rivers, the Kubuqi Desert varies from east to west in terms of the landform types, thickness of accumulations, constituent materials, and moisture conditions, which result in large differences in the morphology, height, and degree of activity of sand dunes.

Recent research on Kubuqi Desert sediments has focused on changes in the Yellow River channel17, Yellow River sediment sources18, and geologic dating and paleoenvironmental studies19. Because of the large geographical extent of the Kubuqi Desert, the complexity of the topography of the rivers and mountains around the desert, there are very few explanations for the origin of the Kubuqi Desert and the source of the sand material. There is currently no consensus on the material sources and depositional environments of other deserts in China, although many studies have been conducted on the sources of desert or sandy sediments influenced by the geomorphology of the Qinghai–Tibetan Plateau, the Loess Plateau, and the Yellow River20,21,22. These works can help to understand the evolution and sediment source of the Kubuqi Desert. The Kubuqi Desert sediments have been studied by means of isotope tracing, constructing models and combining them with geo-historical data. The oscillation of the Yellow River channel during the historical period was an important factor affecting the formation of the Kubuqi Desert. Sediment erosion and siltation could have provided material to the Kubuqi Desert during the evolution of the desert’s eastern channel23, and the pattern and degree of contribution of transport would have varied across different geomorphic sites24. Eolian transport and the discharge of large quantities of saltation component in the north-central part of the desert would also have been influential25. Yang et al.19 suggested that the present-day mobile dunes in the Kubuqi Desert may have been formed by landscape degradation and anthropogenic disturbance in the Yellow River Basin during the Han and Tang dynasties, with distal sources possibly from the Qilian Mountains on the northeastern margin of the Tibetan Plateau, and proximal direct sources from fluvial deposits in the Ningxia–Neimeng section of the Yellow River26. Yang et al.27 speculated that the dunes within the Kubuqi Desert may have had markedly different sediment sources, and that changes in flow following the diversion of the Yellow River for irrigation prompted sediments from the Hetao Plain region to become a source of the more stable eolian sediments required for the expansion of the Kubuqi Desert.

Changes in the sand-carrying capacity of air or water currents, and topographic or scrub vegetation factors, can cause changes in sediment movement and sorting28,29,30. The cause of the rich geomorphologic types and landscapes of the Kubuqi Desert is still the transportation or accumulation of erodible materials on the surface by wind and water from the sediment source. Some scholars have found that the transport of wind-formed sand material is controlled by the vertical structure of the sediment surface layer. IJzendoorn et al.31 found that the formation and propagation of localized wind-formed layers can lead to sediment thinning, while identifying high-resolution vertical grain size variation and how it is affected by oceanic and wind-formed processes. Jung Hoi Soo, et al.32 delineated coastal environments in Korea by common sediment characteristics and found that the influence of water is the main constraint on the structural and geochemical characteristics of coastal sediments. Xi et al.33 studied sediments in the Qixing Lakes area of the Kubuqi Desert and found that fine-grained material on the dune surfaces was transported so that the dunes were not the main contributing surface type to the suspended material in the area. Although previous studies have provided some hypotheses for the origin of Kubuqi Desert sediments, the work done has been limited to localized areas of the desert or to certain dune types. There have been few comparative studies on the grain-size characteristics of different regions and types of dunes in the Kubuqi Desert, especially in the context of the complex geomorphological variability across the eastern, central, and western sections of the desert. In addition, the classification and localization of potential sand sources for the Kubuqi Desert have not yet been documented in detail.

Therefore, we systematically collected surface sediment samples from all different types of dunes throughout the Kubuqi Desert, described the regional variations in dune sediment characteristics, depositional processes and environmental conditions at the regional scale, and constructed a complete grain-size database of different types of dune sediments in the Kubuqi Desert. This complete open data set will provide basic data support for the study of the formation and evolution of dunes and deserts, the source of sand materials, and the restoration and reconstruction of wind-formed environments.

Study area

The Kubuqi Desert is located in Ordos, Inner Mongolia, China, on the northern edge of the Ordos Plateau and south of the Hetao Plain in the middle reaches of the Yellow River. The desert covers an area of 17,300 km2 in a long, narrow band measuring approximately 370 km from east to west and ca. 15–50 km from north to south (Fig. 1). The sand dunes of the Kubuqi Desert almost entirely cover the Yellow River terrace on the northern edge of the Ordos Plateau. The average annual wind speed is 3.5 m/s, which can increase to 4.9 m/s in individual areas in spring, and there are 25–35 windy days per year, half of which occur in spring. The prevailing wind directions are west, north, and northwest.

Fig. 1
figure 1

Geomorphologic background of the study area.

Full size image

The altitude of the desert is 1,000–1,400 m, and the terrain rises in a step-like manner from north to south. The Kubuqi Desert is dominated by mobile dunes, which account for 61% of the whole desert area; 12.5% of the area consists of semi-fixed dunes and 26.5% is occupied by fixed dunes. To reflect differences in sediment sorting by different wind-dynamic conditions during the formation and development of different types of dunes, the dynamic conditions of dune formation must be taken into account when determining sampling points. We applied the Bagnold, McKee, Lancaster, and Thomas classification systems for dunes34,35,36,37, to summarize the formation and development processes of the main dune types in the Kubuqi Desert. We identified five dune types: parabolic dunes; ridge–honeycomb dunes; crescent dunes and chains; reticulate dune chains; and compound dunes. We also obtained samples from shrub-coppice dunes and inter-dune lowlands that did not reflect the dominant wind direction. Compound dunes occur mainly in the central desert, and their height gradually decreases from 50–100 m in the west to 30–50 m in the east. Crescent dunes and chains are widely distributed in the western, south-central and eastern parts of the desert, with heights decreasing eastward from 10–30 m to 5–10 m. Reticulate dune chains are mainly found in the western and central parts of the desert, with heights of 10–30 m. Parabolic dunes are found only in the western part of the desert. Shrub-coppice dunes and ridge–honeycomb dunes are more centrally distributed in the central, eastern, and marginal parts of the desert.

In terms of natural zones, Dalate County–Zhungeer County in the eastern section of the Kubuqi Desert is a semi-arid zone with annual precipitation of 250–400 mm. The southeastern part of the desert is the dry erosive environment of the northern Ordos Plateau, where there is a large area of arsenical sandstone with serious soil erosion. Ten rivers, including the Hantaichuan, Xiliugou, and Heilaigou, originate from the plateau, flow through the desert from south to the north, and join with the Yellow River: these rivers are generally known as the “Ten Tributaries”. In fact, they are large gullies characterized by summer floods, winter dryness, and high sand content. In the middle section of the desert, the line between Hangjin County and Duguitala Town marks a transitional area. The Qixing Lakes are located in the north of the central part of the desert. In the south, there is high terrain formed of scrub hills and dry eroded highlands. The western part of the desert is an arid and semi-desert zone. West of Hangjin County–Duguitala Town is the arid zone with annual precipitation of only 150–250 mm. In this area there are very few rivers, only the Mo Lin river that disappears into the desert to the northwest. Southwest of the desert and east of the Yellow River lie the Zhuozi Mountains. The northern boundary of the Kubuqi Desert is the Yellow River, the lowest point of the entire desert. The northern bank of the Yellow River in the northwestern part of the desert is the Hetao Plain, and the Lang Mountains are to the plain’s far northwest. To the east of the Hetao Plain is Wuliangsuhai Lake, formed by years of irrigation recession, and to its northeast lies Wula Mountain.

Methods

Step 1: Obtaining samples of different types of dune sediments

The surface material of dunes is the most direct reflection of the modern wind conditions; thus, we mainly collected surface sediment samples. The sampling method was similar to the sampling criteria used by Lancaster38 in the Namib Desert: a sand surface of approximately 0.2 m × 0.2 m area was selected and sand grains were collected uniformly from the surface at a depth of 0–10 cm, with a sample weight of ca. 1000 g. To facilitate comparative analysis, we adopted a unified sampling specification: samples from crescent dunes and chains and parabolic dunes were obtained from the middle of the windward and leeward slopes and the top of the dune; samples of reticulate dune chains, compound dunes, and ridge–honeycomb dunes were collected in the middle of the different slopes; samples from shrub-coppice dunes were taken from the top of the dune; and inter-dune lowlands samples were obtained from the relatively gently sloping areas between the dunes. The sample sites basically cover the whole territory of the Kubuqi Desert, and extend to the peripheral area of the Kubuqi Desert inland rivers to analyze the inheritance relationships of sediment transport and accumulation. For each of the seven dune types, an area of approximately 1 km × 1 km was set up for sampling. In total, 396 sampling areas were delineated (Fig. 2), and five typical dunes or surfaces were selected within each area, meaning that a total of 1,980 surface sediment samples was collected. This sampling method allows us to discuss the mechanical composition of sediments from different dune types under different dynamic conditions, and to compare the same dune type in different regions of the desert. The results allow us to explore the role of different material sources and different dynamic conditions on the sorting of sand materials.

Fig. 2
figure 2

Geographic location of the study area and sampling points.

Full size image

Step 2: Determination of the base particle size data of the sample in the laboratory

Measurements of sediment particle size were obtained in the Key Laboratory of Aeolian Sand Physics and Sand Control Engineering, College of Desert Science and Engineering, Inner Mongolia Agricultural University. The volume fraction of the sample size was measured directly with a Mastersizer 3000 laser particle size tester (Malvern Company, Shanghai, China) with measurement range 0.01–3500 μm. For each sample, measurements were repeated three times to obtain the average value.

Step 3: Calculate particle size parameters

The particle size was classified using the Udden–Wentworth criterion, and transformed using the Krumbein equation39:

$$\varphi =-\,\log 2D$$
(1)

where D is particle diameter (mm).

The formula proposed by Folk and Ward was used to calculate the particle-size characteristic parameters40. The parameter \({\rm{\varphi }}\)5 represents the value of \({\rm{\varphi }}\) at a value of 5% on the cumulative particle-size curve, and \({\rm{\varphi }}\)16, \({\rm{\varphi }}\)25, \({\rm{\varphi }}\)50, \({\rm{\varphi }}\)75, \({\rm{\varphi }}\)84, and \({\rm{\varphi }}\)95 are the values of \({\rm{\varphi }}\) at values of 16%, 25%, 50%, 75%, 84%, and 95% on the cumulative particle-size curve, respectively.

The mean particle size (MZ) reflects the average kinetic energy of sand flow.

$$MZ=(\varphi 16+\varphi 50+\varphi 84)/3$$
(2)

The sorting coefficient (\(\sigma \)), also known as the standard deviation, represents the degree of dispersion of sand particle-size distribution.

$$\sigma =(\varphi 84-\varphi 16)/4+(\varphi 95-\varphi 5)/6.6$$
(3)

Skewness (SKG) represents the degree of symmetry of the particle-size curve, and thus reflects the symmetry of the sand-particle distribution.

$$S{K}_{G}=\frac{\varphi 16+\varphi 84-2\varphi 50}{2(\varphi 84-\varphi 16)}+\frac{\varphi 5+\varphi 95-2\varphi 50}{2(\varphi 95-\varphi 5)}$$
(4)

Kurtosis (KG) reflects the degree of concentration of sand particle size.

$${K}_{G}=\frac{\varphi 95-\varphi 5}{2.44(\varphi 75-\varphi 25)}$$
(5)

In general, there is a consistent pattern of variation of grain-size parameters in different parts of a dune surface; for example, the mean particle size of lateral dunes becomes finer from windward to leeward slopes, and the particle sizes of sand grains on both sides of linear dunes and on different slopes of pyramidal dunes are approximately similar, but coarser than those on the tops of the dunes29,38. The purpose of this study was to elucidate the grain-size differences and regional patterns of change between different types of surface sediments; therefore, the grain-size parameters of different geomorphologic parts of the same type of dune were averaged, and variation of the data within a group is reflected by the error bars (statistical standard deviation) in the figures.

Data Record

The dataset is available at https://doi.org/10.6084/m9.figshare.2881218841 in Figshare, and it is a comprehensive particle size database of surface sediments from different dune types in the Kubuqi Desert, China41. It consists of 2 data files, Excel DATA 1 and DATA 2, describing the grain size characteristics of surface sediments of different types of dunes in the Kubuqi Desert.

DATA 1

Characterization of particle size parameters of surface sediments from dunes in the Kubuqi Desert and database of depositional environments. It specifically contains the following data fields: Sheet 1: Basic information about the sample collection, including where the sample was collected (latitude and longitude), when it was collected, and an explanation of the acronyms for the different dune types. Sheet 2: (1) the name of the sample site; (2) the type of dune in the area where the sample is located; (3) the grain size corresponding to the cumulative percentage of sediment (μm); (4) the grain size corresponding to the cumulative percentage of sediment (mm); (5) the grain size corresponding to the cumulative percentage of transformed sediment (φ); (6) the four grain size parameters of the sediment; (7) the depositional environment of the sediment. This file provides the basic sediment composition data of different dunes and does the discrimination of the depositional environment.

DATA 2

Database of particle size parameter characteristics and composition of surface sediments of different types of dunes in the Kubuqi Desert. This file is subdivided into seven groups according to the type of dune, named by the abbreviation of the name of the dune type, and specifically contains the following data fields: (1) the number of the sample; (2) the type of dune; (3) the particle size parameters, including the mean particle size, sorting, skewness, and kurtosis; and (4) the percentage of the composition of the components of the sample. These metrics provide a measure of the internal composition of different types of dunes and potential modes of sand transport.

Technical Validation

The grain size data of different types of dune sediments in the Kubuqi Desert underwent a rigorous validation process to ensure their accuracy, reliability and representativeness. A series of technical steps were carefully implemented throughout the data collection, processing and analysis phases to ensure the high quality and reproducibility of the results. Sediment size is measured using a high-precision laser diffraction particle size analyzer (Mastersizer 3000), which accurately analyzes particle size in the range of 0.01–3500 microns. Particle size is determined by wet determination, particle size meter shading range of 5 ~ 15%, the working principle is based on the principle of laser scattering to test the size of the particle size, when the characteristics of the substance being detected or the physical behavior of a substance with the same to the size of the particles, the particles are defined as the equivalent size of the particles to be measured (particle size distribution). In order to ensure the reliability of the measurement results, in the field sampling, the same sample area we get 5 duplicate samples, fully mixed to ensure the accuracy and consistency of the analytical results; samples in the machine before the test is standardized and strict pre-treatment to ensure that the sample particles are fully dispersed; during the test process, between two samples test, the instrument will automatically wash the instrument with water thoroughly for 3 times, in order to prevent the previous test sample particles left behind. During the testing process, the instrument will automatically wash the instrument thoroughly with water 3 times between two samples to prevent particles of the previous test sample from remaining. Each data point is basic, transparent and reproducible, and the methods of sample testing, analysis and calculation are described in detail, enabling researchers to analyze this data directly, as well as expanding it to other desert regions.

Regional variations in particle-size characteristics

Although there are no large geomorphologic units within the Kubuqi Desert that alter the regional circulation structure and influence the local prevailing wind direction, local sand hills, depressions, and river movements have important influences on material sorting in different regions. Possible sources of material in the region are flood, alluvial, and lacustrine phase materials; as a result, there are pronounced regional differences in the grain-size characteristics of surface sediments in the desert.

Among the 396 sampling areas in the Kubuqi Desert, the dominant mean particle size was clay, very fine sand, fine sand, medium sand, and coarse sand in 20, 41, 293, 41, and 1 area(s), respectively. Overall, the desert was dominated by very fine sand, fine sand, and medium sand. The northwestern edge of the Kubuqi Desert, the Qixing Lakes region were dominated by coarse material, and Dalate County by fine material (Fig. 3). The explanation for this regional difference is that the area of huhemudu as well as the Qixing Lake used to be an oscillating area of the Yellow River channel, where the flowing water carried fine-grained material and more coarse-grained material remained on the surface, as a result of riverine-phase deposition. The area around Dalate county, which is the east-central region of the northern boundary of the Kubuqi Desert, where a large number of channels converge into the Yellow River or break off the flow, where fine-grained material is deposited, and northwest of Hangjin Banner, where sandy material that has been transported from the upwind area is carried by the northwesterly winds, and offloaded for deposition in the barrier of the plateau, which is also the downstream zone of channels in the northwestern part of the Ordos Plateau, making the grain size of the sediments in this region finer. Fine sand was the most widely distributed grain size, found in all surface types in the desert hinterland from west to east. Very fine sand occurred in the lower reaches of the Molin River in the southwestern part of the desert, the northwestern part of the salar, and the Heilaigou and the Hashilachuan rivers, and was concentrated on the sides of ditches and rivers. Medium sand was concentrated at the northwestern edge of the desert, in the northern part of the center, and in the area around the Qixing Lakes. Very fine, fine, and medium sand mostly occurred in crescent dunes and chains, reticulate dune chains, compound dunes, and ridge–honeycomb dunes. In the northwestern region, where the parabolic dunes are concentrated, medium sand was abundant. Silt was mainly found in the interior of the Qixing Lakes area, on the north-central border of the desert, and along the route from Duguitala Town eastward to Dalate County. These areas are oasis transition zones, alluvial plains, ditches, or the lowermost reaches of rivers in the periphery of the desert with better water conditions. Coarse sand was detected in the more heavily eroded upstream arsenical sandstone area. This sand represents a remnant of the coarser debris left over from ditch flooding removal, which occurs in only one location in the sampling area.

Fig. 3
figure 3

Regional variations in the dominant grain size of the Kubuqi Desert.

Full size image

Depositional environments reflected in particle-size parameters

The regional variations in grain-size parameters reflect the material origin of a portion of the Kubuqi Desert and the depositional environments that the material has experienced. This information on depositional environments reflects, to some extent, the sediment transport history and sorting processes. Sahu42 introduced sediment grain-size analysis using the linear multivariate discriminant method and developed four different discriminant functions to distinguish different depositional environments. The discriminant functions are as follows:

$${Y}_{1}=-\,3.5688Mz+3.7016{\sigma }^{2}-2.0766S{K}_{G}+3.1135{K}_{G}$$
(6)

Y1 < −2.7411 indicates eolian deposition. Otherwise, proceed to Eq. 7.

$${Y}_{2}=15.6534Mz+65.7091{\sigma }^{2}+18.1071S{K}_{G}+18.5043{K}_{G}$$
(7)

Y2 < 65.3650 indicates deposition in a beach environment. Otherwise, proceed to Eq. 8.

$${Y}_{3}=0.2825Mz-8.7604{\sigma }^{2}+4.8932S{K}_{G}-0.0482{K}_{G}$$
(8)

Y3 > −7.4190 indicates deposition in a shallow-marine environment. Otherwise, proceed to Eq. 9.

$${Y}_{4}=0.7215M{\rm{z}}-0.4030{\sigma }_{1}^{2}+6.7322S{K}_{G}+5.2927{K}_{G}$$
(9)

Y4 < 9.8433 indicates turbidity-current deposition; otherwise, it is fluvial deposition.

Depositional environments were identified by applying the discriminant functions (Eqs. 6–9) to the samples collected in this study. The depositional environments on the surface of the Kubuqi Desert exhibit obvious regional characteristics (Fig. 4).

Fig. 4
figure 4

Sedimentary environments in the Kubuqi Desert.

Full size image

According to the discriminant functions, the Kubuqi Desert is dominated by eolian deposition (231 areas), with smaller occurrences of shallow-marine deposition (63 areas), turbidity-current deposition (52 areas), fluvial deposition (45 areas), and beach deposition (5 areas) (The identified depositional environments represent the sample points and are used to distinguish between the samples). Eolian deposition occurs mainly in various areas that contain mobile dunes, such as in Suruitu village in the west; on the south bank of the Yellow River on the north side of Qixing Lakes in the central part of the desert; in the Duguitala to Hangeer gou watersheds; in the desert hinterland from Engebei to south of Zhandanzhao; and south of Dalate County to the Dagou drainage basin in the east of the desert.

Beach and shallow-marine deposits are both concentrated in the southeastern part of Huhemudu and in the area of the Qixing Lakes from Jirigalangtu to Duguitala. Shallow-marine deposits are found in the lower reaches of Molinhe, near the salt lakes of Bayinwusu, along and downstream of the Ten Tributaries, and along the south bank of the Yellow River south of Tuoketuo County. However, marine deposition cannot at present occur in those areas, because of the long distance from the ocean. We explain this discrepancy as follows: the discriminant functions of Sahu42 do not distinguish between shallow-marine deposition and lacustrine deposition. Lake deposits are similar to shallow-marine deposits, with the main differences being the sediment mineral composition and carbonate content.

The fluvial deposition referred to in Sahu’s formula is actually estuarine deltaic deposition, because the samples considered by Sahu42 were predominantly deltaic sediments. The area of fluvial deposition in the Kubuqi Desert overlaps with the area of shallow-marine deposition; fluvial deposition also occurs in the upstream of Molihe, the middle part of the Kubuqi Desert, the channel from Duguitala to Hangjin County, and the downstream inlet of the Ten Tributaries. These areas are also where turbidity-current deposition was identified.

Factors affecting the particle size characteristics of the Kubuqi Desert are diverse and integrated. The patterns of sediment particle-size distribution are reflected by the mean particle size and sorting coefficient of the sediment, which in turn are controlled by the material composition of the source area, the history of desert formation, and the evolution of dune formation. Although there is some variation in the composition of the surface material for different deserts and for different individual dunes, in most deserts the sandy material has a particle size of fine to medium sand (mean particle size 1.60–2.65 φ), with excellent to medium sorting (sorting coefficient 0.26–0.55 φ)36. The major mobile deserts in China, such as the Badanjilin Desert and the Tengger Desert, also have mean particle sizes that fall within the above ranges. The exception is the Taklamakan Desert, the largest mobile desert in China, in which the mobile dune surface material is dominated by very fine and fine sand (mean particle size 2.0–4.0 φ)43.

We compared the grain-size parameters of the surface sediments of the major mobile dune types of the Kubuqi Desert with those of the same types of dunes in other deserts (Table 1). The surface sand grains of the crescent dunes and chains of the Kubuqi Desert are similar to those of the Kumtag Desert, finer than those of the Badanjilin and Namib deserts, and coarser than those of the Taklamakan Desert and the southeastern edge of the Tengger Desert. The grain size of the Kubuqi Desert lies in the middle of the grain-size range of the world’s deserts, although with slightly poorer sorting than other deserts, but better than that of the Kumtag Desert. With the exception of the Kumtag Desert, the surface material of linear dunes in other deserts is coarse and poorly sorted. The mean particle size of the sandy material on the surface of the Kubuqi Desert compound dunes is fine, but coarser than that of the Kumtag and Taklamakan deserts, and the sortability is only slightly better than that of the Taklamakan Desert.

Table 1 Grain-size parameters of dune sediments in different deserts and dune types.
Full size table

The poor sorting of shrub-coppice dunes in the Kubuqi Desert partly results from the fact that most of the samples are from riverine or historical alluvial fans, and material transport is stronger in the hydrodynamic scour of an alluvial environment. The reticulate dune chains and inter-dune lowlands in the Kubuqi Desert exhibit a negatively skewed distribution, because the coarse-grained fraction of the neighboring dunes becomes aggregated on the surfaces of the inter-dune lowlands. Negative skewness occurs in reticulate dune chains because on the leeward slopes of reticulate dunes, sand particles crossing the top of the dune are locally supersaturated by the deceleration of the airflow separation and the top-supporting effect of the airflow within the reverse vortex, which makes the eolian flow locally supersaturated. Thus, deposition and sliding occur on the leeward slopes, adding a coarser component to the finer-grained material on the leeward slopes, which manifests as a negative skewness. Similarly, the negatively skewed distribution of sediment particle size is reflected in the kurtosis. The thick tails in the particle-size distribution cause the frequency curve to be thin and narrow because of the shortening of the tails, which results in large kurtosis values.