Field observations of wind profiles and sand fluxes above the windward slope of a sand dune before and after the establishment of semi-buried straw checkerboard barriers

Aeolian Research 20 (2016) 59–70

Contents lists available at ScienceDirect

Aeolian Research journal homepage: www.elsevier.com/locate/aeolia

Field observations of wind profiles and sand fluxes above the windward slope of a sand dune before and after the establishment of semi-buried straw checkerboard barriers Chunlai Zhang a,⇑, Qing Li a, Na Zhou a, Jiaqiong Zhang b, Liqiang Kang a, Yaping Shen a, Wenru Jia a a State Key Laboratory of Earth Surface Processes and Resource Ecology, MOE Engineering Research Center of Desertification and Blown-sand Control, Beijing Normal University, Beijing 100875, China b State Key Laboratory of Soil Erosion and Dryland Farming on the Loess Plateau, Institute of Soil and Water Conservation, Northwest A&F University, Yangling, Shaanxi 712100, China

a r t i c l e

i n f o

Article history: Received 20 June 2015 Revised 1 November 2015 Accepted 9 November 2015

Keywords: Wind profile Aerodynamic roughness length Sand flux Straw checkerboard barriers

a b s t r a c t Straw checkerboard barriers are effective and widely used measures to control near-surface sand flow. The present study measured the wind profiles and sand mass flux above the windward slope of a transverse dune before and after the establishment of semi-buried straw checkerboards. The 0.2 m high checkerboards enhanced the aerodynamic roughness length to larger than 0.02 m, which was two to three orders of magnitude higher than that of the bare sand. The modified Charnock model predicted the roughness length of the sand bed during saltation well, with Cm = 0.138 ± 0.003. For the checkerboards, z0 increased slowly to a level around 0.037 m with increasing wind velocity and the rate of increase tended to slow down in strong wind. The barriers reduced sand flux and altered its vertical distribution. The total height-integrated dimensionless mass flux of saltating particles (q0) above bare sand followed the relationship ln q0 = a + b(u⁄t/u⁄) + c(u⁄t/u⁄)2, with a peak at u⁄/u⁄t
2, whereas a possible peak appeared at u⁄/u⁄t
1.5 above 1 m  1 m straw checkerboards. The vertical distribution of mass flux above these barriers resembled an ‘‘elephant trunk”, with maximum mass flux at 0.05–0.2 m above the bed, in contrast with the continuously and rapidly decreasing mass flux with increasing height above the bare sand. The influences of the barriers on the wind and sand flow prevent dune movement and alter the evolution of dune morphology. Ó 2015 Elsevier B.V. All rights reserved.

1. Introduction Straw checkerboard barriers are an inexpensive, convenient, effective, and widely used measure to control near-surface sand flow (Zheng, 2009). The barriers have been successfully used in the protective system of the Baotou–Lanzhou railway line where it crosses the Tengger Desert and in the protective system of the Taklamakan Desert Highway in western China, as well as in other desert regions of China. Checkerboard barriers can be constructed from the straw of wheat, rice, reeds, or other plants. In practice, the straw is arranged in the shape of a checkerboard and half is buried in the sand and the other half is left exposed (Fig. 1). The barriers are generally installed at a size of 1 m  1 m or a little larger and typically have 0.2 m of straw height above the surface. The barriers disrupt the flow of wind and thereby change the struc⇑ Corresponding author at: State Key Laboratory of Earth Surface Processes and Resource Ecology, Beijing Normal University, 19 Xinjiekouwai Street, Haidian District, Beijing 100875, China. Tel./fax: +86 10 62207162. E-mail address: [email protected] (C. Zhang). http://dx.doi.org/10.1016/j.aeolia.2015.11.003 1875-9637/Ó 2015 Elsevier B.V. All rights reserved.

tures, direction, and intensity of the near-surface wind–sand flow above mobile sands, reducing erosion and facilitating surface stability. This plays an important role in fixing sand by preventing surface sand from encroachment and causing sand in transport to settle in place. This technique has now been introduced in Ghana, Egypt, and Iran (Zheng, 2009). Much research has been done on understanding the mechanisms by which the straw checkerboard barriers control sand flow, including studies of the principles that govern the control of blowing sand, the size of the barriers, and their effectiveness (Buckley, 1987; Wolfe and Nickling, 1993; Arens et al., 2001; Qu et al., 2007). It is generally accepted that the barriers increase the roughness of the underlying surface, reduce the wind velocity and sand transport intensity (Gao et al., 2004; Qiu et al., 2004), and create a stable, concave surface inside the grid created by the barriers (Ling, 1991; Wang and Zheng, 2002). Work has also been done on the remarkable ecosystem effects of the barriers, which can promote the formation of biological soil crusts and the restoration of habitat and species diversity (Li et al., 2004, 2006b). Previous research

60

C. Zhang et al. / Aeolian Research 20 (2016) 59–70

Fig. 1. Semi-buried straw checkerboard barriers.

focused mainly on wind tunnel simulations (Liu, 1988) or theoretical deductions (Wang and Zheng, 2002), and there have been few field observations (e.g., Qu et al., 2007). To improve our comprehension of how the barriers affect the near-surface wind and sand flow, field observations of the near-bed wind behavior and particle transport are needed. The near-bed behavior of wind above bare mobile sand in the presence of a particle flow has been well studied (Owen, 1964; Williams, 1964; Anderson and Haff, 1988, 1991; Sørensen, 1991, 2004; White and Mounla, 1991; Rasmussen et al., 1996; McKenna Neuman and Maljaars, 1997; Iversen and Rasmussen, 1999; Namikas, 2003). Owen (1964) first described the apparent saltation-induced aerodynamic roughness length, which depends on the shear velocity:

z00 ¼ Cu2 =2g

ð1Þ

z00

where is the apparent saltation-induced aerodynamic roughness length (m), C is a proportionality constant, u* is the shear velocity (m s1), and g is the acceleration due to gravity (m s2). Since then, several models have been proposed to relate the aerodynamic roughness length during saltation to the shear velocity (Raupach, 1991; Sherman, 1992; Butterfield, 1993; Farrell, 1999; Sherman and Farrell, 2008). However, the most physically plausible relationship is the modified Charnock relationship (Sherman, 1992; Sherman and Farrell, 2008):

gðz00  z0 Þ ¼ C m ðu  ut Þ2

ð2Þ

where z0 is the aerodynamic roughness length in the absence of saltation (m), Cm is the modified Charnock constant, and u*t is the impact threshold (m s1). Sherman and Farrell (2008) used a compilation of 137 wind profiles from field measurements and determined the value of the modified Charnock constant to be 0.132 ± 0.080. The best-fit value of the modified Charnock constant from a comprehensive numerical model of steady-state saltation (COMSALT) by Kok and Renno (2009) was 0.118, which is close to the empirical value obtained by Sherman and Farrell (2008). Sand flow caused by wind above sand dunes represents a fundamental geomorphological process. Bagnold (1936) was the first researcher to relate the sand transport rate to u⁄, but in the past century, many aeolian sand-transport models have been developed (Bagnold, 1937; Kawamura, 1951; Zingg, 1953; Owen, 1964; Kadib, 1965; Hsu, 1971; Lettau and Lettau, 1978; White, 1979; Sørensen, 2004; Kok and Renno, 2009). These models are able to predict the observed maximum total mass flux to different degrees of accuracy (Sherman and Li, 2012; Sterk et al., 2012). This important progress has reiterated the necessity to obtain reliable predictions of mass flux for verification of sediment transport models and the calibration of theoretically derived flux equations.

Studies of the near-bed behavior of wind within the saltation layer and of the movement of sediments caused by wind have clarified the properties of this air flow and sediment transport, though some of the theories still need verification with field observations. Above the sand surface covered by straw checkerboard barriers, the near-bed air flow and particle movement tend to be more complex than in their absence. Although wind tunnel observations and theoretical simulations have been carried out to explore these phenomena, there have been few field observations examining the difference in aeolian transport characteristics above these barriers compared with that above bare sand (e.g., Zhou et al., 2014). Given these limits of our knowledge, we designed a field study to provide additional information to increase our understanding of the effect of checkerboard roughness on saltation processes. Based on comparative field observations above the same windward slope of a sand dune, we (i) compared the vertical wind profiles above bare sand and two sizes of straw checkerboard barrier; (ii) explored the vertical distribution of sediment flow above the three surfaces; (iii) examined the applicability of the modified Charnock model (Sherman, 1992) for describing near-bed wind profiles; and (iv) analyzed the relationship between the sand transport rate and the wind shear velocity above the windward slope of sand dunes.

2. Materials and methods 2.1. Study site The Shapotou section of the Baotou–Lanzhou railway is located southeast of China’s Tengger Desert (37°270 N, 104°590 E), where network dunes, barchan dunes, and barchan chains are the main dune types (Fig. 2). The prevailing wind direction is from the NW and WNW; ENE is an important secondary wind direction. The railway protective system in the Shapotou area consists of four parts: an upright sand fence farthest upwind, a belt of straw checkerboards and vegetation without irrigation, an irrigated vegetation zone, and a gravel platform for sand transport with no settlement. The protective system has continued to work well since its establishment in 1956. The straw checkerboards and the vegetation belt without irrigation are the main parts of the protective system. At the upwind edge of the protective system, straw checkerboard barriers have been installed on the upper parts of the windward slopes of sand dunes to create the first barrier against blown sand, and have been crucial not only in fixing the mobile sand and retarding the migration of sand dunes, but also in altering the evolution of dune geomorphology and affecting the near-bed air flow in downwind areas (Zhang et al., 2007). Our experiments were carried out on the windward slope of a mobile transverse dune immediately upwind of the railway protective system (Fig. 2a). The topography of the dune was measured using a Trimble 4700 global positioning system (GPS) receiver with real-time differential correction that provided planimetric and altimetric precision of 0.01 m and 0.02 m, respectively (Fig. 2b). The dune was 4.8 m tall measured from the toe of the leeward slope and 13.7 m tall measured from the toe of the windward slope, and had a 68.8-m-long windward slope (Fig. 2c). The total windward slope was about 10.5°, but the slope was variable from the toe to the top. The fixed site of the wind and sand flux measurements was located on the upper windward slope of the dune (Fig. 2c), where the surface slope is 5.6°. The local sediments were predominantly well-sorted, fine to medium quartz sands; coarser surface lags were also present, particularly in association with larger ripples. Zhang et al. (2007) reported a mean grain size of 0.129 mm (2.95 U) and a sorting coefficient of 0.29 U calculated using the formulas of Folk and Ward (1957) at this site.

C. Zhang et al. / Aeolian Research 20 (2016) 59–70

61

Fig. 2. (a) Location of the study area, (b) topographic map of the study site in which the contour lines represent the elevation (m asl), and (c) cross-section of the dune in the direction of the dominant prevailing wind. The solid circle denotes the location on the dune where wind profiles and mass fluxes were measured.

2.2. Determination of the aerodynamic roughness length (z0) and shear velocity (u*) To calculate the aerodynamic roughness length of different surfaces, we continuously measured wind velocity profiles at a fixed observation site on the bare sand surface and on a sand surface covered by 1 m  1 m straw checkerboards, respectively. The velocity profiles were monitored using a mast of seven rotatingcup anemometers mounted at 0.1, 0.2, 0.4, 0.6, 1.0, 1.5, and 2.0 m. All anemometers were calibrated by Changchun Meteorological Instrument Factory, China. These anemometers recorded the average wind speed at 30-s intervals using a data logger thus the results represented average wind speeds, rather than instantaneous ones. A wind vane was mounted on top of the mast to record direction. We first measured the wind velocity profiles and sand fluxes above the bare sand surface from 13:05 to 14:05 h on 24 April. We then established 1 m  1 m straw checkerboards in the experimental area right after the measurement in that afternoon. The straw checkerboards were made of wheat straw and were installed so that the top was 0.2 m above the bed. After establishing the barriers, we smoothed the foot traffic and mounds of sand from burial so that the sand surface was almost flat within the open spaces in the checkerboard. The later four days were allotted for the surface to adjust to an approximately equilibrium condition that a concave surface inside the cells of the checkerboard formed by wind. The coverage of the checkboard barriers is 15% around measured with the photos of the checkboard barriers using image processing method. The fetch of checkerboards between the start and to where the instruments were located was 40 m (Fig. 2b),

which was 20 times of the wind profile measurement height (2 m) and was considered long enough to eliminating the fetch effect on the vertical wind speed profile. The wind velocity profiles and sand fluxes above 1 m  1 m straw checkerboards were measured from 08:35 to 09:35 h on 29 April. During the observation period, the wind only blew from the WNW direction. The location of the anemometers on the checkboard barriers is shown in Fig. 3. We obtained a total of 119 and 92 wind datasets above the bare sand and the sand covered by 1 m  1 m straw checkerboards, respectively. Every dataset included the 30-s mean wind velocities at seven heights. The greatest observed wind velocity at a height of 2.0 m was 14.68 m s1. Table 1 shows the number of occurrences of velocities in various velocity ranges measured at 2.0 m above the two surfaces. Fig. 4 shows the wind profiles for different wind velocities to a height of 2.0 m above the surface, for the mean wind velocity in each of the 1.0 m s1 intervals. For example, for the bare sand, we averaged the values from 18 datasets of observed wind data for the velocity interval between 5 and 6 m s1 to obtain one representative wind velocity profile for this group. Shear velocities (u⁄) and roughness lengths of the bed (z0) can be calculated from the average wind profiles using the following equations (Dong et al., 2001, 2003a):

u ¼ KB

ð3Þ

z0 ¼ eðA=BÞ

ð4Þ

where K is von Karman’s constant (0.4) and A and B are the intercept and slope, respectively, of the logarithmic function of the wind profiles, and are determined by means of least-squares curve fitting.

62

C. Zhang et al. / Aeolian Research 20 (2016) 59–70

Fig. 3. The checkboard barriers and the anemometers on the windward slope of the dune.

Table 1 Measured number of occurrences for each wind velocity interval (1.0 m s1) at a height of 2 m above different surfaces. Type of surface

Bare sand 1 m  1 m straw checkerboards

Wind velocity range (m s1) 3–4

4–5

5–6

6–7

7–8

8–9

9–10

10–11

11–12

12–13

13–14

14–15

1 1

4 1

18 18

32 17

23 24

17 12

4 9

6 9

3 1

6 –

3 –

3 –

Fig. 4. Wind profiles above the different surfaces. Only representative profiles of the mean for each velocity interval are shown, where uz¼2m represents the average wind velocity at a height of 2 m above the bed.

Above a sand dune such as the one in this study, the logarithmic zone governed by u* is typically limited to the first 0.2 m, which means that u* and z0 derived from wind profiles at heights ranging from 0 to 2.0 m would deviate from the real values. However, due to the general limitations of the instrumentation that can be used at field measurement facilities, it is very difficult to measure wind profiles within the first 0.2 m. In order to minimize the errors, we used wind data from the lower five heights (0.1, 0.2, 0.4, 0.6, and 1.0 m)

to fit the logarithmic curves for the near-bed wind profile of the bare bed. On the sand ground covered by straw checkerboards, since the anemometer at 0.1 m was below the height of the roughness and therefore sheltered giving a low wind speed bias compared to those above the roughness, we used the wind data from the heights of 0.2, 0.4, 0.6, 1.0 and 1.5 m, instead of 0.1, 0.2, 0.4, 0.6, and 1.0 m to fit the logarithmic curves for the wind profiles above the checkerboards. We obtained high goodness of fit (R2 > 0.98 for

C. Zhang et al. / Aeolian Research 20 (2016) 59–70

most wind profiles in the logarithmic regression) (for details please see Appendix A). This suggests that our calculations of u* and z0 from the measured wind profiles were reliable. 2.3. Measurements of sand flux The sand flux in the near-surface layer to a height of 0.6 m was measured above the two surfaces using the vertical sand traps developed by Zhang et al. (2014). The traps consisted of a set of vertically stacked 30 square sampling heads with the same openings of 2 cm high and 2 cm wide. Maximum trap height above the bed was 0.6 m. Because the sand trap in our study is a nonpoint device, we must choose a representative elevation for each sampling head to characterize the vertical distribution. Ellis et al. (2009) suggested that where large numbers of relatively ‘short’ traps as 2 cm high are used, the choice of a representative elevation is less important because the potential differences in results from using alternative elevation representations are reduced. Present study used the geometric mean of the top and bottom elevations of each sampling head as used by Namikas (2003) and Rasmussen and Mikkelsen (1998) and recommended by Ellis et al. (2009). Corresponding wind data were collected during this experiment. Both the mast that held the anemometers and the vertical sand trap were set up at a fixed measurement location on the windward slope of the dune (Fig. 2a and c). Limited by the field measurement conditions, we were only able to obtain a few usable datasets for mass flux above the surface covered by the 1 m  1 m straw checkerboards at wind velocities of 5.12, 5.81, 7.35, 7.55, 8.40, 8.55, 8.79, and 9.77 m s1. Individual runs lasted from 60 to 300 s. Above the bare sand, we obtained 38 usable datasets for the mass flux at wind velocities from 6.15 to 14.68 m s1. The trapped sediments from each height were weighed separately with an electronic scale with a precision of 1 mg to obtain the vertical distribution of the mass flux. 3. Results and analysis 3.1. Wind profiles and the aerodynamic roughness length The roughness length above the bare sand generally increased with increasing wind velocity. Ranging from 5  105 to 5  104 m, z0 increased very slowly and seemed nearly stable for the stationary surfaces at low wind velocity. But in the case of the moving bed especially at wind velocities higher than 9 m s1, it increased rapidly to values of 102 m magnitude, being two to three orders of magnitude higher than the values at lower wind velocities. In this study, roughness length was calculated from wind velocity profiles, whether during transport or not. Rapidly increased roughness length during transport was considered resulting from the increasing mass flux in the saltation clouds at higher velocities. The roughness length of a bed that is undergoing saltation is the apparent saltation-induced roughness length, z00 , and is dependent on the shear velocity (Owen, 1964; Raupach, 1991; Sherman, 1992). To assess the ability of the Owen model (Owen, 1964) and the modified Charnock model (Sherman, 1992; Sherman and Farrell, 2008) in predicting the enhanced roughness lengths for the windward slope of the dune, we used linear regression analysis for the relationships between u2⁄ /2g and z00 and between (u⁄  u⁄t)2/g and (z00  z0) derived from the 103 wind profile datasets for which u⁄ was >u⁄t. In this analysis, we assigned the roughness length of the static sand surface in the presence of ripples an estimated value of 104 m based on the values observed for bare sand at a wind velocity lower than 6 m s1, which represents the estimated threshold wind velocity at a height of 2 m for typical mobile sand

63

surfaces in northern China (Wu, 1987). Fluid threshold shear velocity, u⁄t, was determined based on the mean grain size, and was calculated to be 0.17 m s1 using Bagnold’s (1941) equation rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi   ut ¼ A gd qsqq , where qs and q are the densities of the grain and the fluid respectively, d is the mean grain size and A is constant (=0.1). We calculated the proportionality constant C of the Owen model (Owen, 1964) to be 0.109 ± 0.002 (R2 = 0.955) using the 103 datasets using linear regression analysis, where C is the slope of the least-squares line (Fig. 5a). Similarly, we calculated the modified Charnock constant (Cm) (Sherman, 1992; Sherman and Farrell, 2008) to be 0.138 ± 0.003 (R2 = 0.951; Fig. 5b), which is very close to the results obtained by Sherman and Farrell (2008) and by Kok and Renno (2009), which were 0.132 ± 0.080 and 0.118, respectively. Though both the linear regressions obtained high R2 values, the physical meaning of Cm is more definitive (Sherman and Farrell, 2008), therefore the modified Charnock model appears to provide a better prediction of the enhanced roughness of the natural sand surface on the windward slope of sand dunes in our study area. For the surfaces covered by the straw checkerboards, it is not meaningful to discuss whether a Charnock or modified Charnock relationship exists between u⁄ and z0 due to the disturbance of the barriers on wind and particle movement. Instead, the changes of roughness length as a function of wind velocity (Fig. 6) implied different effects of wind velocity above the checkerboard-covered sand surface. z0 increased swiftly from 2.1  102 m at a mean velocity of 3.79 m s1 to 3.2  102 m at a mean velocity of 6.0 m s1 above the 1 m  1 m straw checkerboards. These values were two to three orders of magnitude higher than those above the natural sand surface. However, at wind velocities higher than 6 m s1, z0 increased with increasing wind velocity very slowly, from 3.2  102 m at a mean velocity of 6.0 m s1 to 3.6  102 m at a mean velocity of 11.56 m s1, greatly decreasing the difference in z0 between the two surfaces. Gillies et al. (2002) examined that grass-like roughness reconfigures itself with increasing wind speed, viz. higher wind speeds cause a simultaneous decrease in frontal area and optical porosity thus lowering its drag coefficient as well as bending in the wind. Therefore, for sand covered by straw checkerboard with a height of 0.2 m, there may be a critical value of z0 that would change little with increasing wind velocity, and that value appears to be 3.7  102 m (Fig. 6). However, since our data series were not sufficiently long (the largest mean wind velocities were only 11.56 m s1 for the 1 m  1 m straw checkerboards), this hypothesis will require further fieldwork to verify. The effect of sand flow on z0 of the bare sand strengthens when wind velocity increases due to the more intense mass flux and the greater height reached by the saltating particle cloud. In contrast, because the sediments in the straw checkerboards are protected against the wind and because the straw is so flexible that it tends to be bent by the wind, thereby reducing exposure of the surface between the straw squares, z0 of the checkerboard-covered sand did not increase as quickly as that of the bare sand. Though sediment flow and its influence on z0 for the 1 m  1 m straw checkerboards cannot be neglected, and strengthened with increasing wind velocity, the influence of the particle cloud on z0 could not offset the bending effect of the straw. Differences in magnitude of the roughness lengths of the two surfaces suggest that the aerodynamic roughness due to the vegetation cover overwhelms that due to the saltating sand (Owen, 1964). Many field observations of wind profile above vegetation showed that roughness length of vegetated surface decreased with wind velocity ascribing to the bending of stems of plants in the wind (Sutton, 1953), which differs a little with our results that z0

64

C. Zhang et al. / Aeolian Research 20 (2016) 59–70

Fig. 5. Variation in the aerodynamic roughness length (z0) as a function of wind velocity at a height of 2 m above the bare sand and sand covered by 1 m  1 m straw checkerboards. The lines are regression curves for two surfaces and the equations are z0 = 2.58003  107  u4.27326 (R2 = 0.792) and z0 = 0.03822  0.21926/u2 (R2 = 0.431), respectively.

Fig. 6. Regression analysis to test (a) the Charnock relationship (Owen, 1964) and (b) the modified Charnock relationship (Sherman, 1992). Here, z00 is the apparent saltation-induced roughness length, z0 is the roughness length of the static sand surface, u⁄ is the shear velocity, and g is the acceleration due to gravity.

increased quickly at low wind velocity but tended to be stable when wind velocity increases further. Another opinion from wind tunnel results of the standing vegetation by Dong et al. (2001) is that the effect on z0 is highly variable. Their work divided the wind velocity profile over the standing vegetation into three sections, each of which can be described by logarithmic function. Bo et al. (2015) reached the similar conclusion by computational fluid dynamics. For field measurement on the ground with straw checkerboard barriers, it is difficult to measure the wind profile below 20 cm above the ground with the current wind speed observation equipment. This limitation prevented us from measuring and discussing the wind profile in the ground layer below 20 cm in the field condition of the straw checkerboard. No matter the surface type, z0 derived from the wind profiles showed considerable noise (Fig. 5). This resulted mainly from instability of the nearbed air flow, especially in the presence of the straw checkerboard barriers. 3.2. Sand flux above different surfaces In general, the amount of sediment traveling in saltation decreases rapidly with increasing height. The most commonly cited relationship for mass flux takes an exponential form (Williams, 1964; Anderson and Hallet, 1986; Greeley et al., 1996;

Dong et al., 2003b; Namikas, 2003; Kok and Lacks, 2009; Farrell et al., 2012). However, a power function (Zingg, 1953) and a logarithmic function (Rasmussen and Mikkelsen, 1998) have also been proposed. In the present study, the vertical distribution of the mass flux within 0.6 m above the bare sand followed an exponential function model fairly well for all wind velocity categories (R2 = 0.89–0.99). Above the sand covered by the 1 m  1 m straw checkerboards, mass flux decreased rapidly with increasing height at low wind velocities (5.12 m s1 and 5.81 m s1), which was similar to the results above the bare sand. However, the mass flux increased to a peak at a height of 0.05–0.2 m and then decreased with increasing height at higher wind velocities, which was quite different from the continuously and rapidly decreasing flux as a function of height above the bare sand (Fig. 7). This kind of curve form was described as an ‘‘elephant trunk” by Qu et al. (2005) in their study of blown sand on a gobi surface. The total height-integrated mass flux of saltating particles is a key parameter for studies of dune formation (Sauermann et al., 2001). Much work has been done to relate the aeolian mass flux to the shear velocity of the wind flow (Bagnold, 1941; Owen, 1964; Lettau and Lettau, 1978; Sørensen, 2004). By graphing the dimensionless mass flux, q0 (where q0 = qg/qu3⁄ , q is the total height-integrated mass flux, and q is air density) as a function of the dimensionless shear velocity (u⁄/u⁄t), researchers have increasingly discovered that there is a peak in the curve, followed by a decrease at higher shear velocities; this research includes field experiments (Gillette et al., 1996), wind tunnel studies (McKenna Neuman and Maljaars, 1997), and numerical models by Kok and Renno (2009). Some theoretical work explains the peaked shape of this curve (Sørensen and McEwan, 1996; Sørensen, 2004). Sørensen’s (2004) observed that the peak of the dimensionless mass flux appears at u⁄/u⁄t
1.5 for homogeneous sand with a mean diameter of 0.242 mm. In our measurements, the peak in the curve for mass flux above the bare sand occurred at u⁄/u⁄t
2 (Fig. 8). Therefore, sand flow above the dune’s windward slope is very similar to that above flat sand. The mass flux curve can be expressed as follows:

ln q0 ¼ a þ bðut =u Þ þ cðut =u Þ2

ðR2 ¼ 0:728Þ

ð5Þ

where a, b, and c are regression coefficients with the values of 1.92408, 10.30813, and 10.91743, respectively. In the field condition, it is very difficult to measure the threshold on the spot. On the straw covered sand, we made use of the measured data of sand flux under different shear velocities to fit an empirical relation between shear velocity and sand flux. Predicted from the empirical relation (q = a + bu⁄c, a = 28123.86,

C. Zhang et al. / Aeolian Research 20 (2016) 59–70

65

Fig. 7. The vertical distribution of the sand flux (q) as a function of height (z) above (a) bare sand on the natural windward slope of the dune and (b) the slope covered by 1 m  1 m straw checkerboard barriers. Note the very different x-axis scales in the two graphs.

Fig. 8. Changes in the relationship between the dimensionless transport rate q0 (q0 = qg/qu3⁄ ) as a function of the dimensionless shear velocity (u⁄/u⁄t) for two surfaces.

b = 29104.89, c = 0.0396, R2 = 0.868), threshold shear velocity for the straw covered sand is about 0.421 m s1. Then we plotted the dimensionless mass flux (q0) to the dimensionless shear velocity (u⁄/u⁄t) in the same graph for the bare sand (Fig. 8). It showed a similar relationship between q0 and u⁄/u⁄t with that on the bare sand based on our limited measurement data, but sand transport above such a surface is much lower. It seems that a peak of q0 exists at u⁄/u⁄t
1.5, but it is difficult to fit a curve due to small measurement data As a result of remarkably reduced sand transport, the dunes where the straw checkerboard barriers were installed was prevented from migrating and the evolution of the dune topography was also altered by the barriers (Li et al., 2006a). For the dunes that were completely covered by straw checkerboard barriers, horizontal movement ceased and the dunes began to grow upwards due to sand accumulation. 4. Conclusions In this study, we analyzed wind profiles and particle mass fluxes above the windward slope of a sand dune before and after the establishment of semi-buried straw checkerboard barriers. On the basis of our findings, we can draw the following conclusions:

Straw checkerboards increase the aerodynamic roughness length (z0). Above the sand covered by the 0.2-m-tall straw checkerboards, z0 values were on the order of 2.1  102–4.2  102 m, which are two to three orders of magnitude higher than the value of the bare sand bed at wind velocity <6 m s1. Due to the plastic properties of the straw (i.e., it bent under strong winds), z0 was observed to increase little as wind velocity exceeded 6 m s1. For the bare sand surface on the windward slope of the dune, z0 increased with increasing wind velocity, supporting previous suggestions that the modified Charnock model did a reasonable job of predicting the roughness length during saltation above a bed; the modified Charnock constant (Cm) was 0.138 ± 0.003. Sand entrainment and transportation in the belt of straw checkerboards differed greatly from the bare sand case. The total height-integrated dimensionless mass flux of saltating particles (q0) above the bare sand was strongly related to the dimensionless shear velocity (u⁄/u⁄t). A peak in the curve for this relationship appeared at u⁄/u⁄t
2 and the curve followed the form ln q0 = a + b(u⁄t/u⁄) + c(u⁄t/u⁄)2. Similarly, a peak appeared at u⁄/u⁄t
1.5 for sand flow at the same position on the dune, but covered by 1 m  1 m straw checkerboards, but it was much weaker than that above the bare sand. More data will be required to define the precise form of the relationship between q0 and u⁄/u⁄t for the straw covered sand. The vertical distribution of the mass flux above the sand covered by the 1 m  1 m straw checkerboard barriers changed as the wind velocity increased. At low wind velocities, the mass flux decreased rapidly with increasing height. At high wind velocities, it typically showed an initial increase with increasing height that reached a maximum at heights ranging from 0.05 to 0.2 m then subsequently declined with increasing height beyond 0.05–0.2 m. The shape of the resulting curve was consequently reconfigured and resembled an elephant trunk, which is quite different from the continuously, rapidly decreasing mass flux with increasing height that occurred above the bare sand. The difference in the response of the vertical distribution of sand flux to wind velocity appears to be related to changes in particle movement within the straw checkerboards between different wind velocities. Differences in both the near-surface airflow and the sand flow between bare dune slopes and dunes covered by straw checkerboards are the main causes for stabilization of dunes in the study area. Besides, semi-buried straw checkerboards have been considered leading to the variation of dune morphology. Further field observations and more detailed research will be required to clarify the dynamic relationships among the straw checkerboard barriers, sand movement, and dune development.

66

C. Zhang et al. / Aeolian Research 20 (2016) 59–70

Beijing Normal University for their assistance during the field measurements.

Acknowledgments This work was supported by the Natural Science Foundation of China (Grants Nos. 41171004 and 41330746). The authors fully acknowledge three anonymous reviewers and the associate editor who gave us invaluable comments to improve the manuscript. We are grateful for the support they received from the staff of the Shapotou Desert Research & Experiment Station, Chinese Academy of Sciences. We also thank Y.G. Liu, X.J. Ma, and H.Z. Wang of

Type of surface

No.

Bare sand

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50

u (m s1) 3.49 4.17 4.75 4.78 4.96 5.07 5.17 5.20 5.21 5.27 5.57 5.60 5.63 5.64 5.67 5.69 5.70 5.77 5.77 5.78 5.80 5.94 5.98 6.02 6.05 6.07 6.07 6.12 6.13 6.13 6.14 6.23 6.25 6.28 6.32 6.46 6.55 6.55 6.58 6.61 6.62 6.67 6.69 6.71 6.71 6.71 6.73 6.73 6.83 6.88

Appendix A Regression of the relationships between wind velocity uz and height z for all velocity profiles. Fitted function is uz = A + B lnz, where uz and z are in m s1 and m, respectively. u in this table is the wind velocity at 2 m above the ground.

A 3.2257 3.9405 4.5004 4.1692 4.8663 4.3802 4.6699 4.2525 4.3550 4.5172 4.2393 5.2067 4.9670 4.7626 4.4201 5.0022 4.3113 4.8636 5.6442 4.4934 5.0884 5.3701 4.4863 5.8889 4.8781 5.6462 5.5663 5.9156 5.8960 5.2040 5.5326 5.1557 6.0405 4.3621 4.8954 4.9625 6.3052 6.4610 6.3652 5.6002 5.8291 6.2580 5.3897 5.0717 6.1131 6.0453 5.8030 5.7241 5.8161 5.8374

B

R2

u⁄ (m s1)

z0 (m)

0.3331 0.4481 0.5135 0.4749 0.5378 0.5059 0.4832 0.4349 0.4748 0.5050 0.4713 0.6131 0.6003 0.5475 0.5245 0.5564 0.4919 0.5353 0.6089 0.5164 0.5847 0.6219 0.5172 0.6826 0.5230 0.60325 0.6652 0.6889 0.6555 0.6106 0.6149 0.5935 0.6241 0.5170 0.5835 0.5511 0.6567 0.6977 0.7222 0.6576 0.6041 0.6743 0.6421 0.5513 0.7203 0.6924 0.6715 0.6781 0.7081 0.6906

0.959 0.946 0.995 0.98 0.975 0.984 0.976 0.991 0.959 0.988 0.99 0.958 0.988 0.991 0.974 0.969 0.939 0.991 0.989 0.972 0.974 0.928 0.953 0.976 0.965 0.975 0.963 0.960 0.975 0.982 0.980 0.977 0.987 0.989 0.989 0.971 0.986 0.989 0.986 0.991 0.957 0.982 0.957 0.961 0.975 0.979 0.972 0.981 0.985 0.966

0.133 0.179 0.205 0.190 0.215 0.202 0.193 0.174 0.190 0.202 0.188 0.245 0.240 0.219 0.210 0.223 0.197 0.214 0.244 0.207 0.234 0.249 0.207 0.273 0.209 0.241 0.266 0.276 0.262 0.244 0.246 0.237 0.249 0.207 0.233 0.220 0.263 0.279 0.289 0.263 0.242 0.270 0.257 0.220 0.288 0.277 0.269 0.271 0.283 0.276

0.00006 0.00015 0.00016 0.00015 0.00012 0.00017 0.00006 0.00006 0.00010 0.00013 0.00012 0.00020 0.00026 0.00017 0.00022 0.00012 0.00016 0.00011 0.00009 0.00017 0.00017 0.00018 0.00017 0.00018 0.00009 0.00009 0.00023 0.00019 0.00012 0.00020 0.00012 0.00017 0.00006 0.00022 0.00023 0.00012 0.00007 0.00010 0.00015 0.00020 0.00006 0.00009 0.00023 0.00010 0.00021 0.00016 0.00018 0.00022 0.00027 0.00021

67

C. Zhang et al. / Aeolian Research 20 (2016) 59–70 Appendix A (continued)

Type of surface

No.

u (m s1)

A

B

R2

u⁄ (m s1)

z0 (m)

51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111

6.89 6.89 6.94 6.98 6.98 7.01 7.12 7.22 7.24 7.25 7.27 7.32 7.33 7.36 7.40 7.46 7.46 7.48 7.49 7.49 7.56 7.60 7.61 7.64 7.73 7.90 7.91 7.93 8.14 8.14 8.17 8.24 8.38 8.40 8.42 8.44 8.46 8.52 8.63 8.70 8.72 8.78 8.80 8.90 8.98 9.33 9.49 9.54 9.78 10.13 10.22 10.42 10.55 10.83 10.91 11.18 11.61 12.18 12.33 12.65 12.69

5.4429 6.8578 6.2788 6.9009 6.6384 5.0167 6.7404 5.8839 6.5676 7.1053 7.0726 7.1678 5.7571 7.1239 6.4214 6.0184 5.8970 7.3908 7.1931 7.2277 7.2831 7.5235 5.6128 7.5646 6.9886 6.2185 7.7875 7.5293 6.5254 7.9258 7.5499 8.0005 8.2203 8.0747 8.5094 8.0674 8.4501 8.2099 8.4717 8.5327 8.8114 8.2710 9.0254 9.0254 6.7435 8.9755 9.6866 10.0762 8.2514 10.2490 10.2332 10.4397 10.4916 10.0838 10.9333 11.3337 11.5263 12.0204 12.3416 12.4672 12.7531

0.6395 0.7228 0.6956 0.7837 0.7552 0.6038 0.7239 0.7098 0.7638 0.7242 0.8321 0.7939 0.7071 0.8615 0.7505 0.7346 0.6883 0.8305 0.7820 0.8089 0.8457 0.8761 0.6507 0.8452 0.8359 0.7459 0.8854 0.8122 0.7598 0.8895 0.8596 0.9894 0.9367 0.9503 0.9659 0.9435 1.0883 0.9832 0.9830 1.1307 1.5667 1.0388 1.0086 1.2719 0.6923 1.1496 1.5616 2.4369 0.9790 2.0016 2.3150 2.1012 2.4214 1.2959 2.3630 2.5151 2.7722 2.5719 3.3054 3.2563 2.8648

0.977 0.994 0.981 0.997 0.980 0.991 0.979 0.967 0.974 0.968 0.981 0.989 0.986 0.976 0.985 0.940 0.985 0.993 0.989 0.989 0.989 0.988 0.906 0.993 0.974 0.997 0.987 0.986 0.964 0.987 0.975 0.986 0.993 0.985 0.952 0.982 0.983 0.983 0.985 0.987 0.987 0.988 0.984 0.930 0.969 0.981 0.942 0.938 0.943 0.915 0.997 0.996 0.955 0.974 0.925 0.944 0.980 0.961 0.980 0.983 0.914

0.256 0.289 0.278 0.314 0.302 0.242 0.290 0.284 0.306 0.290 0.333 0.318 0.283 0.345 0.300 0.294 0.275 0.332 0.313 0.324 0.338 0.350 0.260 0.338 0.334 0.298 0.354 0.325 0.304 0.356 0.344 0.396 0.375 0.380 0.386 0.377 0.435 0.393 0.393 0.452 0.627 0.416 0.403 0.509 0.277 0.460 0.625 0.975 0.392 0.801 0.926 0.840 0.969 0.518 0.945 1.006 1.109 1.029 1.322 1.303 1.146

0.00020 0.00008 0.00012 0.00015 0.00015 0.00025 0.00009 0.00025 0.00018 0.00005 0.00020 0.00012 0.00029 0.00026 0.00019 0.00028 0.00019 0.00014 0.00010 0.00013 0.00018 0.00019 0.00018 0.00013 0.00023 0.00024 0.00015 0.00009 0.00019 0.00013 0.00015 0.00031 0.00015 0.00020 0.00015 0.00019 0.00042 0.00024 0.00018 0.00053 0.00361 0.00035 0.00013 0.00083 0.00006 0.00041 0.00202 0.01601 0.00022 0.00597 0.01203 0.00695 0.01313 0.00042 0.00979 0.01104 0.01564 0.00934 0.02390 0.02174 0.01166

(continued on next page)

68

C. Zhang et al. / Aeolian Research 20 (2016) 59–70

Appendix A (continued)

Type of surface

1 m  1 m straw checkerboard

No.

u (m s1)

A

B

R2

u⁄ (m s1)

z0 (m)

112 113 114 115 116 117 118 119 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54

12.89 12.97 13.00 13.10 13.19 14.59 14.64 14.74 3.79 4.77 5.05 5.06 5.08 5.44 5.44 5.53 5.54 5.56 5.61 5.63 5.63 5.64 5.81 5.83 5.83 5.93 5.93 5.96 6.01 6.13 6.36 6.36 6.54 6.55 6.62 6.70 6.73 6.75 6.78 6.82 6.86 6.86 6.87 6.91 6.92 7.01 7.04 7.11 7.55 7.19 7.22 7.26 7.35 7.38 7.40 7.43 7.52 7.60 7.62 7.65 7.67 7.74

12.8328 12.8444 13.1812 12.0661 13.3279 14.1158 14.4553 14.6895 2.9079 3.8834 3.6063 3.7628 3.6524 3.9127 4.1050 4.4339 4.1215 4.8550 4.2132 4.1951 4.2455 4.1420 4.9903 4.6742 4.2964 4.3725 4.4248 4.6365 5.1721 4.7719 4.5887 4.6595 5.7034 4.6354 4.7507 5.4680 4.9858 5.3054 4.9705 4.6955 5.2251 5.2757 4.7710 5.4424 5.6848 5.0625 5.5687 5.0671 6.8080 5.2559 5.0877 5.3922 6.0625 6.4477 5.8396 5.3271 5.2015 6.2281 5.4129 6.2653 5.6415 6.0462

3.1173 3.1619 3.0821 2.5042 3.4743 3.6614 3.8483 3.7267 0.7561 1.0813 1.0145 1.0234 1.0374 1.1185 1.1914 1.2545 1.2113 1.3791 1.2191 1.2225 1.2380 1.2051 1.4501 1.3468 1.2516 1.2862 1.2984 1.3058 1.5335 1.3998 1.3860 1.3876 1.6402 1.3647 1.3862 1.6369 1.4425 1.5861 1.4892 1.4109 1.5442 1.5720 1.4668 1.6530 1.6546 1.5310 1.5786 1.5210 2.0778 1.6214 1.5415 1.5732 1.7724 1.8857 1.6760 1.5717 1.5565 1.8068 1.6090 1.8583 1.6868 1.7571

0.940 0.972 0.914 0.989 0.918 0.968 0.981 0.993 0.986 0.989 0.973 0.996 0.98 0.981 0.992 0.987 0.997 0.973 0.978 0.989 0.988 0.978 0.97 0.978 0.992 0.995 0.989 0.98 0.974 0.998 0.988 0.994 0.974 0.992 0.999 0.988 0.984 0.995 0.981 0.976 0.992 0.993 0.981 0.993 0.987 0.994 0.992 0.989 0.98 0.994 0.976 0.997 0.974 0.971 0.994 0.978 0.987 0.988 0.994 0.99 0.982 0.995

1.247 1.265 1.233 1.002 1.390 1.465 1.539 1.491 0.302 0.433 0.406 0.409 0.415 0.447 0.477 0.502 0.485 0.552 0.488 0.489 0.495 0.482 0.580 0.539 0.501 0.514 0.519 0.522 0.613 0.560 0.554 0.555 0.656 0.546 0.554 0.655 0.577 0.634 0.596 0.564 0.618 0.629 0.587 0.661 0.662 0.612 0.631 0.608 0.831 0.649 0.617 0.629 0.709 0.754 0.670 0.629 0.623 0.723 0.644 0.743 0.675 0.703

0.01630 0.01721 0.01389 0.00808 0.02158 0.02117 0.02337 0.01942 0.02137 0.02756 0.02859 0.02530 0.02958 0.03025 0.03189 0.02918 0.03329 0.02959 0.03156 0.03234 0.03241 0.03216 0.03202 0.03109 0.03230 0.03339 0.03311 0.02870 0.03429 0.03307 0.03649 0.03481 0.03089 0.03349 0.03248 0.03542 0.03154 0.03526 0.03552 0.03586 0.03392 0.03488 0.03867 0.03717 0.03220 0.03664 0.02937 0.03575 0.03776 0.03910 0.03686 0.03246 0.03269 0.03274 0.03068 0.03373 0.03537 0.03184 0.03459 0.03433 0.03528 0.03203

69

C. Zhang et al. / Aeolian Research 20 (2016) 59–70 Appendix A (continued)

Type of surface

No.

u (m s1)

55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92

7.77 7.78 7.80 7.81 7.85 7.91 7.92 8.10 8.13 8.21 8.29 8.34 8.56 8.66 8.74 8.76 8.81 8.93 8.99 9.26 9.33 9.33 9.43 9.77 9.73 9.84 9.89 9.90 10.00 10.01 10.24 10.43 10.73 10.46 10.57 10.64 10.66 11.56

References Anderson, R.S., Haff, P.K., 1988. Simulation of eolian saltation. Science 241, 820–823. Anderson, R.S., Haff, P.K., 1991. Wind modification and bed response during saltation of sand in air. Acta Mech. 1 (Suppl. 1), 21–51. Anderson, R.S., Hallet, B., 1986. Sediment transport by wind: toward a general model. Geol. Soc. Am. Bull. 97, 523–535. Arens, S.M., Baas, A.C.W., Van Boxel, J.H., Kalkman, C., 2001. Influence of reed stem density of foredune development. Earth Surf. Processes Landforms 26, 1161–1176. Bagnold, R.A., 1936. The movement of desert sand. Proc. R. Soc. A 157, 594–620. Bagnold, R.A., 1937. The transport of sand by wind. Geogr. J. 89, 409–438. Bagnold, R.A., 1941. The Physics of Blown Sand and Desert Dunes. Methuen, London. Bo, T.L., Ma, P., Zheng, X.J., 2015. Numerical study on the effect of semi-buried straw checkerboard sand barriers belt on the wind speed. Aeolian Res. 16, 101–107. Buckley, R., 1987. The effect of sparse vegetation on the transport of dune sand by wind. Nature 325, 426–428. Butterfield, G.R., 1993. Sand transport response to fluctuating wind velocity. In: Clifford, N.J., French, J.R., Hardisty, J. (Eds.), Turbulence: Perspectives on Flow and Sediment Transport. John Wiley, New York, pp. 305–335. Dong, Z.B., Gao, S.Y., Fryrear, D.W., 2001. Drag coefficients, roughness length and zero-plane displacement height as disturbed by artificial standing vegetation. J. Arid Environ. 49, 485–505. Dong, Z.B., Liu, X.P., Wang, H.T., 2003a. The aerodynamic roughness with a blowing sand boundary layer (BSBL): a redefinition of the Owen effect. Geophys. Res. Lett. 30, 1047–1050. Dong, Z.B., Liu, X.P., Wang, H.T., Zhao, A.G., Wang, X.M., 2003b. The flux profile of a blowing sand cloud: a wind tunnel investigation. Geomorphology 49, 219–230. Ellis, J.T., Li, B., Farrell, E.J., Sherman, D.J., 2009. Protocols for characterizing aeolian mass-flux profiles. Aeolian Res. 1 (1), 19–26.

A 5.5323 6.4848 6.0430 6.4270 6.1637 6.3506 5.7067 5.8981 6.2839 6.2491 6.8924 6.6757 6.8400 7.3440 6.9842 6.8519 7.2317 6.4888 7.5620 7.6204 7.9680 7.5250 7.6004 8.9118 7.6277 8.1931 8.0136 7.9301 8.1196 8.0614 8.2341 8.2263 9.8960 8.5618 8.6070 8.9534 8.8990 9.5724

B

R2

u⁄ (m s1)

z0 (m)

1.6386 1.8771 1.7251 1.8889 1.8476 1.9295 1.7315 1.8328 1.8564 1.8170 2.0859 1.9406 2.0137 2.1512 2.0185 2.0250 2.1503 1.9957 2.2168 2.4099 2.3278 2.2331 2.3199 2.7578 2.3034 2.4554 2.3882 2.2945 2.3849 2.4341 2.4728 2.5951 3.0234 2.4969 2.5977 2.6488 2.6246 2.8727

0.992 0.986 0.995 0.996 0.994 0.995 0.968 0.985 0.993 0.999 0.987 0.994 0.993 0.977 0.987 0.994 0.992 0.977 0.985 0.978 0.984 0.987 0.997 0.977 0.989 0.983 0.994 0.99 0.989 0.987 0.989 0.99 0.968 0.992 0.998 0.994 0.992 0.975

0.655 0.751 0.690 0.756 0.739 0.772 0.693 0.733 0.743 0.727 0.834 0.776 0.805 0.860 0.807 0.810 0.860 0.798 0.887 0.964 0.931 0.893 0.928 1.103 0.921 0.982 0.955 0.918 0.954 0.974 0.989 1.038 1.209 0.999 1.039 1.060 1.050 1.149

0.03417 0.03160 0.03011 0.03329 0.03558 0.03720 0.03704 0.04003 0.03387 0.03209 0.03672 0.03206 0.03348 0.03291 0.03143 0.03392 0.03463 0.03872 0.03300 0.04234 0.03261 0.03440 0.03777 0.03950 0.03646 0.03555 0.03489 0.03155 0.03322 0.03645 0.03580 0.04201 0.03789 0.03242 0.03640 0.03404 0.03369 0.03571

Farrell, E.J., 1999. An investigation of surface-wind systems for aeolian saltation: field and laboratory experiments (M.S. thesis). Department of Geography, University of Southern California, Los Angeles. Farrell, E.J., Sherman, D.J., Ellis, J.T., Li, B., 2012. Vertical distribution of grain size for wind blown sand. Aeolian Res. 7, 51–61. Folk, R.L., Ward, W.C., 1957. Brazos River bar, a study in the significance of grainsize parameters. J. Sediment. Petrol. 27, 3–26. Gao, Y., Qiu, G.Y., Ding, G.D., Qing, S.Y., 2004. Effect of Salix psammophila checkerboard on reducing wind and stabilizing sand. J. Desert Res. 24 (3), 365–370 (in Chinese with English abstract). Gillette, D.A., Herbert, G., Stockton, P.H., Owen, P.R., 1996. Causes of the fetch effect in wind erosion. Earth Surf. Processes Landforms 21, 641–659. Gillies, J.A., Nickling, W.G., King, J., 2002. Drag coefficient and plant form-response to wind speed in three plant species: Burning Bush (Euonymus alatus), Colorado Blue Spruce (Picea pungens glauca.), and Fountain Grass (Pennisetum setaceum). J. Geophys. Res. 107 (D24), 10-1–10-15. Greeley, R., Blumberg, D.G., Williams, S.H., 1996. Field measurements of the flux and speed of wind-blown sand. Sedimentology 43, 41–52. Hsu, S., 1971. Wind stress criteria in eolian sand transport. J. Geophys. Res. 76, 8684–8686. Iversen, J.D., Rasmussen, K.R., 1999. The effect of wind speed and bed slope on sand transport. Sedimentology 46, 723–731. Kadib, A.A., 1965. A function for sand movement by wind. University of California Hydraulics Engineering Laboratory Report HEL 2–8. University of CaliforniaBerkeley, Berkeley, CA. Kawamura, R., 1951. Study of sand movement by wind. In: The Reports of the Institute of Science and Technology: vol 5. University of Tokyo, Japan, pp. 95– 112. Kok, J.F., Lacks, D.J., 2009. Electrification of granular system of identical insulators. Phys. Rev. E 79, 051304.

70

C. Zhang et al. / Aeolian Research 20 (2016) 59–70

Kok, J.F., Renno, N.O., 2009. A comprehensive numerical model of steady state saltation (COMSALT). J. Geophys. Res. 114, D17204. http://dx.doi.org/10.1029/ 2009JD011702. Lettau, K., Lettau, H.H., 1978. Experimental and micro-meteorological field studies of dune migration. In: Lettau, K., Lettau, H.H. (Eds.), Exploring the World’s Driest Climate. IES Report 101. Institute of Environmental Studies, University of Wisconsin, Madison, pp. 110–147. Li, S.Y., Lei, J.Q., Xu, X.W., Wang, H.F., Zhou, Z.B., Zhou, H.W., Cheng, H.P., 2006a. The effects on the morphology and advance of isolated barchans on interdune corridor of complex ridges in the central Taklimakan Desert. Arid Land Geogr. 29 (5), 632–638 (in Chinese with English abstract). Li, X.R., Xiao, H.L., Zhang, J.G., Wang, X.P., 2004. Long-term ecosystem effects of sand-binding vegetation in Shapotou Region of Tengger Desert, Northern China. Restor. Ecol. 12, 376–390. Li, X.R., Xiao, H.L., He, M.Z., Zhang, J.G., 2006b. Sand barriers of straw checkerboard for habitat restoration in extremely arid desert region of China. Ecol. Eng. 28, 149–157. Ling, Y.Q., 1991. Study on protected effect of straw checkerboard. Res. Desert Control 2, 49–59 (in Chinese). Liu, X.W., 1988. Wind tunnel experiments of straw checkerboard barriers. In: Shapotou Desert Experimental Research Station. Study on shifting sand control in Shapotou region, Southeast of the Tengger Desert. Ningxia People’s Press, Yingchuan, pp. 326–334. Namikas, S.L., 2003. Field measurement and numerical modeling of aeolian mass flux distributions on a sandy beach. Sedimentology 50, 303–326. McKenna Neuman, C., Maljaars, M., 1997. Wind tunnel measurement of boundarylayer response to sediment transport. Boundary Layer Meteorol. 84, 67–83. Owen, P.R., 1964. Saltation of uniform grains in air. J. Fluid Mech. 20, 225–242. Qiu, G.Y., Lee, I.B., Shimizu, H., Gao, Y., Ding, G.D., 2004. Principles of sand dune fixation with straw checkerboard technology and its effects on the environment. J. Arid Environ. 56, 449–464. Qu, J.J., Huang, N., Ta, W.Q., Lei, J.Q., Dong, Z.B., Liu, X.W., Xue, X., Zu, R.P., Zhang, K.C., 2005. Structural characteristics of Gobi sand-drift and its significance. Adv. Earth Sci. 20 (1), 19–23. Qu, J.J., Zu, R.P., Zhang, K.C., Fang, H.Y., 2007. Field observations on the protective effect of semi-buried checkerboard sand barriers. Geomorphology 88, 193–200. Rasmussen, K.R., Iversen, J.D., Rautaheimo, P., 1996. Saltation and wind-flow interaction in a variable slope wind tunnel. Geomorphology 17, 19–28. Rasmussen, K.R., Mikkelsen, H.E., 1998. On the efficiency of vertical array aeolian field traps. Sedimentology 45, 789–799. Raupach, M.R., 1991. Saltation layers, vegetation canopies and roughness lengths. Acta Mech. 1 (Suppl. 1), 83–96.

Sauermann, G., Kroy, K., Herrmann, H.J., 2001. A continuum saltation model for sand dunes. Phys. Rev. E 64, 31305–31316. Sherman, D.J., 1992. An equilibrium relationship for shear velocity and roughness length in aeolian saltation. Geomorphology 5, 419–431. Sherman, D.J., Farrell, E.J., 2008. Aerodynamic roughness lengths over movable beds: comparison of wind tunnel and field data. J. Geophys. Res. 113, F02S08. http://dx.doi.org/10.1029/2007JF000784. Sherman, D.J., Li, B., 2012. Predicting aeolian sand transport rates: a reevaluation of models. Aeolian Res. 3, 371–378. Sørensen, M., 1991. An analytical model of windblown sand transport. In: Barndorff-Nielsen, O.E., Willetts, B.B., (Eds.), Aeolian Grain Transport. Acta Mech. 1(Suppl. 1), 67–81. Sørensen, M., 2004. On the rate of aeolian sand transport. Geomorphology 59, 53–62. Sørensen, M., McEwan, I.K., 1996. On the effect of mid-air collisions on aeolian saltation. Sedimentology 43, 65–76. Sterk, G., Parigiani, J., Cittadini, E., Peters, P., Scholberg, J., Peri, P., 2012. Aeolian sediment mass fluxes on a sandy soil in Central Patagonia. Catena 95, 112–123. Sutton, O.G., 1953. Micrometeorology. McGraw-Hill Book Company, New York (p. 333). Wang, Z.T., Zheng, X.J., 2002. A simple model for calculating measurement of straw checkerboard barriers. J. Desert Res. 22 (3), 229–232 (in Chinese with English abstract). White, B.R., 1979. Soil transport by winds on Mars. J. Geophys. Res. 84, 4643–4651. White, B.R., Mounla, H., 1991. An experimental study of Froude number effect on wind-tunnel saltation. Acta Mech. 1 (Suppl. 1), 145–157. Williams, G., 1964. Some aspects of the eolian saltation load. Sedimentology 3, 257–287. Wolfe, S.A., Nickling, W.G., 1993. The protective role of sparse vegetation in wind erosion. Progr. Phys. Geogr. 17, 50–68. Wu, Z. (Ed.), 1987. Aeolian Geomorphology. Science Press, Beijing. Zhang, C.L., Zhou, N., Zhang, J.Q., 2014. Sand flux and wind profiles in the saltation layer above a rounded dune top. Sci. China: Earth Sci. 57 (3), 523–533. Zhang, C.L., Zou, X.Y., Pan, X.H., Yang, S., Wang, H.T., 2007. Near-surface airflow field and aerodynamic characteristics of the railway-protection system in the Shapotou region and their significance. J. Arid Environ. 71, 169–187. Zheng, X.J., 2009. Mechanics of Wind-blown Sand Movement. Springer-Verlag, Berlin. Zhou, N., Zhang, C.L., Kang, L.Q., Tian, J.L., 2014. Flow field controlling the concave surface of the semi-buried checkerboards and its characterization by grain sizes of sediments. Geogr. Res. 33 (11), 2145–2156 (in Chinese with English abstract). Zingg, A.W., 1953. Wind tunnel studies of the movement of sedimentary material. In: Proceedings of 5th Hydraulic Conference. Institute of Hydraulics, Iowa City, pp. 111–135.