Comparison of Surface Wind Speed and Wind Speed Proles in the Taklimakan Desert

: 13 Near-surface (10 m) wind speed (NWS) plays a crucial role in many areas, including the hydrological 14 cycle, wind energy production, and the dispersion of air pollution. Based on wind speed data from 15 Tazhong and the northern margins of the Taklimakan Desert in Xiaotang in spring, summer, autumn, and 16 winter of 2014 and 2015, statistical methods were applied to determine the characteristics of the diurnal 17 changes in wind speed near the ground and the differences in the wind speed profiles between the two 18 sites. The average wind speed on a sunny day increased slowly with height during the day and rapidly 19 at night. At heights below 4 m the wind speed during the day was higher than at night, whereas at 10 m 20 the wind speed was lower during the day than at night. The semi-empirical theory and Monin-Obukhov 21 (M-O) similarity theory were used to fit the NWS profile in the hinterland of the Tazhong Desert. A 22 logarithmic law was applied to the neutral stratification wind speed profile, and an exponential fitting 23 correlation was used for non-neutral stratification. The more unstable the stratification, the smaller the n. Using M-O similarity theory, the “ linear to tens of ” law was applied to the near-neutral stratification. According to the measured data, the distribution of M φ with stability was obtained. The m γ was obtained 26 when the near-surface stratum was stable in the hinterland of Tazhong Desert and the m β was obtained 27 when it was unstable. In summer, m γ and m β were 5.84 and 15.1, respectively, while in winter, m γ and 28 m β were 1.9 and 27.1, respectively.


31
Near-surface (10 m) wind speed (NWS) and temperature are important parameters for studying 32 atmospheric dynamics and climate change. Research on wind speed and temperature changes will 33 improve our understanding of atmospheric circulation, leading to better climate analyses and 34 predictions. The NWS is one of the key variables in climate research. Changes in the NWS have 35 significant implications for human society and the natural environment (Pryor et al., 2006). The 36 intensification of NWS may aggravate soil erosion, resulting in more severe sandstorms (Alizadeh-37 Choobari wind speed above and below the boundary layer suggest many uncertainties behind the stationary 44 phenomenon. As a product of the atmospheric boundary layer, near-surface wind is transient in 45 nature and is affected by topography and boundary layer processes (Mahrt, 2009; Beluˇ si´ c and 46 Güttler, 2010; Güttler and Beluˇsi´ c, 2012). 47 The stability of the near-surface layer has a significant relationship with the near-surface wind. 48 Previous studies have shown that near-surface stability was comprehensively affected by surface 49 and boundary layer processes. In a stable stratified boundary layer environment, the net downward 50 flow of warm air leads to an increase in surface temperature. In the daytime, when the rate of decline 51 is negative, the net downward transport of cold air will cause a ground cooling effect. researchers have attempted to develop a similar similarity theory for the whole boundary layer. The 57 atmospheric boundary layer is also the main characteristic quantity for assessing turbulent mixing, 58 vertical disturbance, convective transmission, cloud belts, atmospheric pollutant diffusion, and 59 analyzing atmospheric environmental capacity (Therry and Lacarrere, 1983; Holtslag and Boville, 60 1993; Hong and Pan, 1996; Beyrich, 1997;Collier et al., 2005). Earlier studies of the boundary 61 layer mainly focused on the near-surface layer. The development of turbulence theory and 62 technological developments in measuring atmospheric processes have promoted research on the 63 atmospheric boundary layer (Monin andObukhov, 1954；Johnson, B.D.2021). The structure of the 64 atmospheric boundary layer and its evolution have significant diurnal characteristics. After sunrise, 65 solar radiation heats the ground. The increase in the heat flux generated by the near-surface layer 66 strengthens turbulent mixing, the height of the atmospheric boundary layer increases, and the heat 67 content of the boundary layer also increases. It mixes uniformly and the wind speed, temperature, 68 and specific humidity change little with height. This is referred to as the mixed layer. After sunset, 69 long-wave radiation is emitted from the ground, which cools down, resulting in a weakening of 70 turbulent transport. The near-surface layer forms a stable boundary layer, and the upper mixed layer 71 rises from the ground. During this uplift, the atmospheric turbulence characteristics are significantly 72 weakened, but the distribution of meteorological elements in the mixed layer during the day are still 73 maintained. This layer is obviously different from the upper free atmosphere, and is called the 74 residual layer. Because there is often an inversion layer on the top of the atmospheric boundary layer, 75 the upward development of turbulent mixing is inhibited, and therefore a boundary is formed 76 between the atmospheric boundary layer and the free atmosphere. Due to the differences in the 77 thermal properties of the topography and underlying surface, the height of the atmospheric boundary 78 layer presents obvious spatial variation characteristics. Although the Taklimakan Desert in China 79 and the Pearl River Delta have similar frequencies of boundary layer occurrence, the average 80 atmospheric boundary layer height in the Taklimakan Desert is significantly higher (Zhang, 2017). 81 Therefore, in different locations and under different weather backgrounds, the height of the 82 boundary layer can be very different.

83
With the completion of the ecological protection barrier along the highway of the Taklimakan 84 Desert and the restored green spaces of the oil base in the hinterland of the Taklimakan Desert, the 85 nature of the regional underlying surface has changed, resulting in changes in surface wind speed. 86 In the context of climate change, the Taklimakan Desert climate is essentially a complex of basin 87 and desert climates, resulting in an extreme arid continental climate. 88 This study used wind speed data collected by the gradient system of the Taklimakan  hinterland-desert-oasis transition zone. The underlying surface is flat sandy land, with ancient river 110 bed in some areas, and there is no vegetation. The station is located on the south bank of an ancient 111 river bed. An area of mobile dunes is located about 250 m to the south, and mainly consists of 112 crescent dunes and compound crescent dune chains. The dune landform belongs to the ancient Tarim 113 River alluvial-flood plain, and is the intersection of two desert areas. The area has an inland warm 114 temperate desert climate, which is subject to drought and little rain. The annual average wind speed 115 is 2.5 m·s -1 , with a maximum in spring and summer, and a minimum in winter. The change of air 116 temperature at Xiaotang station is similar to that of wind speed, with the maximum in June-July 117 and the minimum in December-January. The change of wind speed and air temperature has an 118 obvious synchronization. Most (80%) sandstorms occur in spring and summer, with the lowest 119 frequency in winter. The annual average wind speed is 2.8 m/s, and the average temperature is 120 11.2°C. Tazhong is located in the center of the hinterland of the Taklimakan Desert (83°39 ' E, 38°58 ' 121 N), with the highest temperature in summer of 46.0°C, and the lowest temperature in winter of -122 25.0°C. The average temperature of the four seasons is 13.6°C, the average annual precipitation is 123 about 25.9 mm, and the evaporation is about 3,812.3 mm. The regional climate is abnormal, the 124 vegetation coverage is extremely low, and only cold and drought-resistant shrubs (e.g., Haloxylon 125 ammodendron) can survive. The peak period of dust storms is from March to August every year. 126 The frequency of dust storms at 12 m above the ground is about 500 per year. The normal average 127 wind speed is 2.5 m·s -1 , and the sand activity intensity index reaches a maximum value of about 128 8000 each year. The prevailing wind direction is northeast and northwest. 129 130 Figure 1. Map of the study area.

132
In this study, a windsonic two-dimensional ultrasonic wind speed and direction sensor (1590-133 PK-020, Campbell Scientific, Logan, UT, USA) and temperature and humidity sensor (1590-PK-134 020, Campbell Scientific) installed on meteorological observation towers were used to conduct 135 parallel comparative experimental observations of microclimate elements in Tazhong and Xiaotang  136 from January 1, 2014 to December 31, 2015 (Fig. 2). The wind speed and direction sensor had a 137 starting wind speed of 0.01 m/s; precision of wind speed ± 2%; range of 0-60 m/s and 0-359°; and 138 resolution of 0.01 m/s and 1°. The range of the temperature sensor was -80-60 °C , the accuracy was 139 ± 0.17°C, and the resolution was 0.1°C. The wind speed, temperature, and relative humidity data 140 used in the study were obtained at the height of 10 m. The data used were subjected to quality control, 141 including the synchronous calibration of the two observation points, the logical extremum of 142 observation data, and a time consistency check. In this study, the four seasons were classed as spring 143 (March to May), summer (June to August), autumn (September to November), and winter 144 (December to February). January, April, July, and October were the representative months of winter, 145 spring, summer, and autumn, respectively. 146 autumn, winter, and spring. Using the ground meteorological data, typical sunny days in each 152 seasonally representative month were selected to determine an average wind speed per hour. The 153 sunny days in summer were August 27, 28, and 31, 2014, the sunny days in autumn were October 154 3, 10, and 11, 2014, the sunny days in winter were January 4, 6, and 7, 2015, and the sunny days in 155 spring were April 12, 13, and 15, 2015. Based on the average daily variation of the surface layer 156 height in the four seasons, it was found that the daily variation of wind speed in Tazhong and 157 Xiaotang on sunny days had two characteristics. First, there were two peaks and two low values in 158 the daily variation of wind speed. The two peaks occurred at night and in the daytime, and the two 159 low values occurred in the morning and evening, respectively. However, the magnitude and 160 occurrence of wind speed were different in the different seasons. Second, there were two distinct 161 forms of change in the lower and upper layers. In the lower layer, the daytime wind speed was 162 greater than in the night, in the upper layer the nighttime wind speed was greater than during the 163 daytime, and the middle layer was stable, but there were differences in each season 164 3.

Diurnal variations of NWS in summer 165
The diurnal variations of surface wind speed in Tazhong and Xiaotang in summer are shown 166 in Fig. 3 (a). The wind speed at different altitudes from 0.5 to 10 m displayed an increasing trend. 167 The maximum wind speeds in each layer in Tazhong  in Xiaotang. At 0.5-2 m, the situation between the two sites at 17:30-23:30 was also the opposite. 174 Comparing the diurnal variation curves of wind speed at different altitudes, it was found that the 175 wind speed increased continuously from 0.5 to 1 m. The wind speed at 0.5 m was below 5 m/s, 176 while the wind speed at 2 m was always above 4 m/s. 177 After sunrise (07.00) in the morning, due to the increasing solar altitude angle and the 178 strengthening of turbulence exchange, the wind speed in each height layer increased rapidly. The 179 average increase of wind speed in each height layer from 0.5 to 10 m was about 0.9 m/s per hour, 180 until 10.00. However, the wind speed at Xiaotang and Tazhong in the 0.5 and 1 m layers displayed 181 a sharp downward trend and then increased slightly until 16:30. At 2, 4, and 10 m, the wind speed 182 increased relatively quickly until 19:00, at which point the wind speed began to decline sharply, 183 leading to a diurnal difference in wind speed of about 6 m/s. After 18.00, with the decreasing solar 184 altitude angle, the ground radiation balance decreased rapidly, the turbulence weakened, and the 185 wind speed decreased rapidly. The decreasing trend in NWS was most obvious below 4 m, with the 186 average wind speed decreasing by l m/s per hour. Due to the gradual formation of a near-surface 187 inversion, the turbulence was further weakened, and the wind speed below 10 m continued to 188 decrease, reaching a minimum at 21.00, with an average wind speed of less than 1 m/s from 21.00 189 to 22.00. This may be because the valley between the large sand ridges alongside the tower was 190 associated with downhill winds at this time. The highest wind speeds at 10 m in Xiaotang occurred at 01:00 and 08:30 pm in the daytime, with 204 values of 5.81 and 6.9 m/s, respectively. The lowest wind speeds occurred at 03:00 and 23:30, with 205 values of 2.1 and 1.8 m/s, respectively. The average daily maximum wind speed was only 2.8 m/s 206 larger than the minimum wind speed, and the daily variation of the low-level wind speed was much 207 reduced compared with that in the higher layer. 208 At 08:00, the NWS within 10 m in Tazhong and Xiaotang decreased rapidly, and at 15:30, the 209 wind speed in each altitude layer decreased significantly. The decrease in Tazhong was greater than 210 that in Xiaotang, with a difference of about 4 m/s, which was different from the pattern observed in 211 the summer. In Tazhong, the wind speeds remained high but the wind speed curves at all five levels 212 in the autumn daytime were more closely arranged than those in Xiaotang, indicating that the wind 213 speed difference in the surface layer within 10 m during the autumn daytime was smaller than that 214 in summer. The wind speed difference between the highest and lowest layers in autumn was 1-2.2 215 m/s, which was about 0.5 m/s smaller than that in summer. 216

Diurnal variations of NWS in winter 217
As shown in Fig.3 (c), the diurnal variation of wind speed in the 10 m layer of Tazhong and 218 Xiaotang in winter was generally small, and the wind speed in the whole 10 m layer was less than 219 3 m/s. In the Xiaotang area, below 4 m from 00:30 to 09:00 the wind speed was maintained at about 220 0.2 m/s. At 09:30, the wind speed began to increase, until at 23:30 it reached about 1.5 m/s. The 221 wind speed trends in Tazhong and Xiaotang followed the completely opposite trend. From 00:30 to 222 09:00 the wind speed at 0.5 m was greater than in the other layers, with a maximum of about 1.8 223 m/s. In winter, the average wind speed in the daytime was larger than that in the nighttime above 2 224 m. The highest wind speeds in Tazhong and Xiaotang generally occurred at 13:00 and 19:30 in the 225 daytime, and the lowest wind speeds occurred at 21:00 and 06:00, respectively. The highest wind 226 speeds at 10 m were 2.5 and 2.9 m/s, and the lowest wind speeds were 0.2 and 0.02 m/s, respectively. 227 The average wind speeds at night in Tazhong and Xiaotang from 0.5 to 1 m were higher than in the 228 daytime. The difference between the maximum and minimum wind speed at 10 m was only 1.3 m/s, 229 which was smaller than in summer and autumn. After 10:30 in winter, the wind speed at each altitude 230 increased slowly. This increase was small and occurred about 2 h later than in summer. After 10.00, 231 the wind speed began to increase at 2 m, and then began to decrease until 16:30. This feature also 232 occurred at 09:00 in autumn at 2 m. and at 10 m it increased from 0.5 to 6.1 m /s. After 18:30, the wind speed at all altitudes showed a 247 downward trend, although this was less obvious than that in summer and autumn. From midnight 248 to sunrise, the NWS within 10 m was much smaller than that in summer, and was similar to that in 249 winter.

Comparison of the different wind speed profiles 256
A curve fitting of the variation in wind speed with height was applied to enable an 257 understanding the near-surface meteorological characteristics in the hinterland of the Taklimakan 258 Desert, as well as an in-depth understanding of the desert wind-sand movement law, and the 259 development and utilization of wind energy resources. It is important to note that in the subsurface The relationship between shear stress and mixing length, and the average wind spee 274 d was substituted into l=kz : The pairwise (2) integral is: 281 A is an integral constant. Many experiments have found that the height at which the average 283 wind speed is zero typically occurs at a height z from the ground that is referred to as o z .

284
If boundary conditions are applied:  atmosphere was consistent with the transition period of the temperature profile. 296

Wind velocity profile under non-neutral stratification 297
Under non-neutral stratification, the wind speed profile will deviate from the logarithmic law, 298 and can be expressed by a simple exponential law in the form of:  Tazhong  306 on August 25, 2014 with the exponential law. The fitting results were very good, and the correlation 307 coefficients were all above 0.98. The stability parameters of each period were fitted. The stability 308 parameter n was smaller in the daytime than in the night, changing from 0.145 to 0.140 from 10:00 309 to 14:00 pm, indicating that the more unstable the stratification was, the smaller the n value was. 310 After 20:00, the atmosphere gradually became stable with an n value of 0.284. With the gradual 311 formation of a nighttime inversion layer, the n value increased, tending to 1. 312

The wind speed profile model established by M-O similarity theory 313
Under uniform and steady conditions, in the atmospheric layer where the wind direction was 314 not significantly deflected with height, according to M-O similarity theory, the dimensionless wind 315 speed gradient in the near-surface layer can be expressed as: 316 where k is the Kaman constant (generally 0.4), * u is the local friction velocity, and M φ is  According to Yamamot (1997), in the layer where the wind direction has no obvious monotonic 368 deflection with height, if the influence of Coriolis force is ignored, only the change of u with height 369 should be considered. The differential equation obtained by Eqs. (12) and (13) is as follows: 370 The m φ value for the measured data at the tower height of 10 m was calculated using Eq. (6),  Due to the influence of vegetation, the increase in ground moisture reduced the surface reflectance, 411 increased the absorption of surface radiation, increased the radiation balance value, and slowed 412 down the temperature change. In the growing season, restored green spaces reduced the surface 413 wind speed, maintain water and soil conditions, and reduced the extent of erosion caused by wind-414 sand disaster weather.

415
By analyzing the diurnal variations of the mean wind speed and temperature on sunny days 416 and the characteristics of different wind speed profiles in the middle of observation towers at 417 Tazhong and Xiaotang in summer, autumn, winter, and spring of 2014 and 2015, the following 418 conclusions were reached. 419 In the four seasons, the diurnal variation of average wind speed on sunny days at different 420 heights in the surface layer of Tazhong and Xiaotang had two characteristics. First, there were two 421 peaks and two low values in the diurnal variation of wind speed. The two peaks occurred at night 422 and in the daytime, and the two low values occurred in the morning and evening, respectively. 423 However, the wind speed varied and the time of occurrence of maximum values differed among the 424 seasons. Second, there were two distinct forms of change in the lower and upper layers. In the lower 425 layer, the daytime wind speed was greater than in the night, in the upper layer the nighttime wind 426 speed was greater than during the daytime, and the middle layer was stable, but there were 427 differences in each season the diurnal variation of wind speed in autumn and spring was larger in 428 Tazhong than in Xiaotang, whereas the variation was similar in summer and winter. 429 When a semi-empirical theory was applied, under neutral stratification there was a high 430 correlation coefficient when a logarithmic law was used for the fitting, and the timing of the 431 occurrence of a neutral atmosphere was consistent with the transition time of the temperature profile. 432 Under non-neutral stratification, the wind speed profile deviated from the logarithmic law, and the 433 fitting was better for an exponential law. For the M-O wind speed profile model, when the 434 atmosphere was close to a neutral stratification and when the general function