在干旱或半干旱地区，土壤风蚀严重危害农业生产。垄状耕作通过改变地表微地形，增加地表粗糙度，影响近地表气流场结构和地表剪应力分布，一定程度上起到削弱风蚀的作用。本文风洞模拟研究了不同高度、不同间距垄状微地形条件下的表面流场结构、空气动力学特征和地表剪应力分布特征，以及垄状微地形对土壤风蚀速率的影响，旨在揭示不同结构的垄状微地形对土壤风蚀的影响机理，为构建土壤风蚀模型提供相关基础数据。风洞模拟实验中，垄的模型均为木质材料做成，因而本文研究对象为不可蚀垄状微地形。论文在以下几个方面取得进展： （1）垄状微地形对近地表流场的影响。垄间距较大时，垄周围近地表流场可划分迎风坡脚风速减弱低速区、迎风坡遇阻抬升区、顶部集流加速区、背风坡减速沉降区和消散恢复区共5个速度区；对于较小的垄间距，则不存在气流恢复区。第一行垄周围风速变化最大，之后风速变化逐渐减小，流场趋于稳定，流场的稳定距离随垄高和垄间距增加而增大。本文中的稳定流场定义为相邻两个垄间以及垄上各高度层的风速大小及变化趋势基本相同。实验风速和垄间距较小时，风速等值线比较平缓。随实验风速和垄间距增大，气流波动性增强，垄间逐渐出现低层气流的“U”型加速区和高层气流的倒“U”型减速区。垄间风速约以垄高为界出现明显分层，1H高度以下风速变化较1H高度以上更为剧烈，反映了1H高度以下贴地层气流和1H高度以上近地层气流截然不同的影响机制。 垄间风速廓线形态的变化进一步反映了垄的结构（包括垄高、垄间距）对气流的影响机制。垄间距较小时（≤5H），垄间地表风速廓线在半对数坐标系中可分为两段，其中贴地层内风速接近于0而且变化很小，以上高度风速廓线符合对数分布规律。此时垄周围流场功能分区不明显，表明均匀分布的垄以粗糙元的形式影响气流。垄间距较大时（>5H），垄间地表风速廓线在半对数坐标系中表现为上凸的曲线，完全偏离了对数律，风速垂向变化剧烈，而且垄周围流场形成不同的速度区，表明此时垄以风障的形式对气流产生影响。可见，较小和较大间距（以5H为界）的垄状微地形，对近地表气流的影响机制不同。 （2）垄间地表剪应力的分布特征。不同结构垄状微地形条件下地表剪应力的分布与贴地层气流具有相似的变化规律。第一行垄前地表剪应力沿风向逐渐减小，垄后逐渐增大。垄间距为5H和10H时，第一、二行垄间的地表剪应力明显高于下风向各行垄间；相邻两行垄之间，上风向垄后至下风向垄前地表剪应力先增大后减小的变化趋势大致相同，这种变化趋势反映了上风向垄后气流恢复变化和下风向垄前气流遇阻减弱的变化特征。垄间距为15H及其以上时，垄间地表剪应力变化过程可分为（上风向）垄后和（下风向）垄前两段，变化趋势均为先增大后减小，上风向垄后和下风向垄前各有一个峰值。第一段是垄后反向涡的作用而形成，第二段是垄间风速逐渐恢复和下风向垄的阻滞作用而形成。 在流场稳定区，平均地表剪应力随垄高、垄间距和摩阻风速的增加而增大，随垄密度（单位长度内垄的行数）的增大而减小。本文基于供试土壤样品的临界剪应力，计算了床面有效剪应力的分布，并建立了平均有效剪应力（ˉ(τ_eff )）与垄高（H）、垄间距（L）和摩阻风速（u*）的定量关系：ˉ(τ_eff )=aL√H (u_*-u_(*t) )^2。 （3）垄状微地形对风沙流和风蚀的影响。在平坦无垄床面，0~70 cm高度内风沙流结构遵循幂函数分布规律。横向垄的存在影响了风沙活动层内的气流能量分布，导致风沙流结构偏离这一规律。多数垄状结构下符合指数规律，部分垄状结构表现出和戈壁风沙流相似的“象鼻”结构。 近地表流场和地表剪应力的空间差异，导致了床面风蚀范围和风蚀强度分布的不均匀性，进而对床面总的风蚀速率或输沙率产生显著影响。垄间距和摩阻风速较小时，风蚀区主要位于第一、二行垄间。随摩阻风速和垄间距增大，流场稳定区也开始出现风蚀，且垄间风蚀区由一段增加到两段，分别是上风向垄后反向涡及其下风向的气流恢复的结果。垄高由5 cm增加到15 cm，垄后反向涡引起的风蚀区最大范围由0~0.2 m增加到0~0.75 m，而气流恢复形成的风蚀区范围则随垄间距和摩阻风速增大而增大。 根据现有输沙方程和风洞实验结果，综合考虑垄高、垄密度（n）和摩阻风速，本文推导得到垄状微地形条件下的输沙率模型：q=C√(d/D) 〖ρ/g u〗_*^2 (u_*-u_(*t) )exp?(-a √H/(u_*^3 ) n)。经过验证，该模型能够较好地反映不同结构垄状微地形对床面总输沙率的影响。结合平均有效剪应力计算结果，尝试建立了平均有效剪应力与风蚀速率的定量关系：Q_((τ) )=k√L ˉ(τ_eff )^2。

In arid or semi-arid areas, soil erosion by wind has seriously damaged agricultural production. Ridge tillage can influence the near-surface airflow field and the distribution of surface shear stress by changing the surface microtopography and increasing the surface roughness to achieve the effect of weakening soil wind erosion. With wind tunnel simulation experiments, this study measured the near surface airflow field and shear stress distribution on the surface covered by different ridges, and analyzed the influence of the transvers ridges on the wind erosion rate, thus revealing the influence mechanism of ridge-shaped microtopography on soil wind erosion and providing basic data for the construction of wind erosion model. Since the ridge models were made of wood materials in the experiments, the object of this study is limited to the unerodible ridge-shape microtopography. The paper has made progress in the following aspects: (1) The influence of ridge-shape microtopography on the near surface airflow field. When the ridge spacing is large, the near surface airflow field around the ridge can be divided into five sub-areas: a deceleration area upwind of the windward slope, an area with rising and accelerating airflow on the windward slope, an area of acceleration above the top of the ridge an area of sinking air and decelerating airflow on the leeward slope, and a downwind flow recovery area. For ridge spacing is small, there is no downwind airflow recovery area. When airflow passes through the bed, wind velocity changes around the first row of ridge drastically and then the amplitude of change decreases downwind gradually untill a relative steady airflow is reached which is characterized by regular fluctuation over uniformly spaced ridges. The distance from the first row of ridge to the zone of stable airflow increases with the enhancement of ridge height and spacing. In this paper the steady airflow field is defined as wind velocity vibration along wind direction are basically the same between any two adjacent ridges at any height above the ground surface. When the wind velocity and ridge spacing are relatively small, the contour lines of the wind velocity are relatively straight. With the enhancement of the fluctuation of wind velocity and the increase of ridge spacing, a U-shap acceleration zone of the airflow in the lower layer and an inverted U-shap deceleration zone of the airflow in the higher layer gradually appear between two adjacent ridges. The airflow is obviously layered at the ridge height. The change of wind velocities below 1H height is more intense than that of wind velocities above 1H height, which reflects the different influence mechanism between the airflow below and above 1H height. The morphological change characteristics of wind velocity profile near surface between two adjacent ridges further reflects the different influence mechanism of ridge structure (including ridge height and spacing) to airflow. When the ridge spacing is less than or equal to 5H, the near surface wind profile can be divided into two segments in the semi-logarithmic coordinate system, the wind velocity of the lower segment is close to 0 and the change is very small, whereas the higher segment obey the logarithmic distribution law. There is no clear sub-area division of airflow field around ridges. In this situation, the uniformly distributed ridges affect the airflow in the form of rough elements. In contrast, when the ridge spacing is larger than 5H, the near surface wind profile between two adjacent ridges shows a convex curve in the semi-logarithmic coordinate system, which is completely deviated from the logarithmic law. The wind velocity varies remarkably with height and there are obvious sub-areas of airflow field around ridges. In this situation, the ridges affect the airflow as wind barriers. It can be seen that the small and large spacing (with 5H as the boundary) of the ridge-shape microtopography have different influence mechanism on the near surface airflow. (2) The distribution of surface shear stress between two adjacent ridges. The distribution of surface shear stress is similar to that of the wind velocity below 1H height under the conditions of different ridge height and spacing. The surface shear stress decreases gradually in front of the first row of ridge and then increases gradually. For the ridge spacings of 5H and 10H, the surface shear stress between the first and second row of ridges is larger than that between downwind two adjacent ridges. From the leeward slope of the upwind ridge to the windward slope of the adjacent downwind ridge, the change trend of surface shear stress between two adjacent ridges is approximately the same. This reflects the effects of airflow recovery behind the upwind ridge and the effects of airflow deceleration due to obstruction by the downwind ridge. For ridge spacing equal or larger than 15H, the change process of the surface shear stress between two adjacent ridges can be divided into two sections that locate at the behind of upwind ridge and in front of the downwind ridge, respectively. In both sections the surface shear stress increases first and then decreases. The first section formed by the action of the reverse vortex behind the upwind ridge, and the second section formed by the gradual recovery of the wind velocity. In the stable region of airflow, the average surface shear stress increases with the increase of ridge height, ridge spacing and wind friction velocity, but decreases with the increase of ridge density (number of ridges per unit length). Based on the threshold shear stress of the tested soil, this paper calculated the distribution of effective shear stress on the sand bed, and established the quantitative relationship among the average effective shear stress (ˉ(τ_eff )), ridge height (H), ridge spacing (L) and wind friction velocity (u*): ˉ(τ_eff )=aL√H (u_*-u_(*t) )^2。 (3) The influence of ridge-shape microtopography on the sand flow and wind erosion rate. In the flat sand bed with no ridges, the sand flow structure, namely the change of sand flux with height above the bed follows the power function. The existence of transverse ridges affects the distribution of airflow energy in the saltation layer, and leads the sand flow structure to deviate from the power function. The sand flow structure for most of the ridge structures follows the exponential decreasing law and some show "elephant nose" structure which is similar to the sand flow above gobi surface. The spatial differentiation of the near surface airflow field and surface shear stress leads to non-uniformity distribution of wind erosion, and then has an effect on the total wind erosion rate or sediment transport rate from the sand bed. When the ridge spacing and friction velocity are small, the wind erosion area is mainly located between the first and second rows of ridges. With the friction velocity and ridge spacing increasing, wind erosion begins to appear in the stable region of airflow field. The wind erosion area between two adjacent ridges increases from one to two segments, which is resulted from the reverse vortex behind the upwind ridge and the gradual recovery of the wind velocity between two adjacent ridges. When the ridge height increases from 5 cm to 15 cm, the range of wind erosion area behind the upwind ridge increases from 0~0.2 m to 0~0.75 m; meanwhile, the range of wind erosion area produced by the wind velocity recovery increases with the increasing of ridge spacing and friction velocity. Using wind tunnel simulation results as inputs for existing sediment transport equations, and taking the ridge height, number of ridges per unit length (n) and the friction velocity, we developed a model for predicting sediment transport over a sand bed that contained widely spaced ridges: q=C√(d/D) 〖ρ/g u〗_*^2 (u_*-u_(*t) )exp?(-a √H/(u_*^3 ) n). It is proved that the model can better reflect the influence of ridge-shape microtopography on the total sediment transport rate. In addition, this paper established a quantitative relationship between the average effective shear stress and the wind erosion rate: Q_((τ) )=k√L ˉ(τ_eff )^2.