where \(W_{f}\) is the total wind factor (\(m^{3}/s^{3}\)); \(\rho\) is the air density (\(kg/m\)); \(g\) is the acceleration due to gravity (\(m/s\)); SW is soil wetness (dimensionless); SDis the snow cover factor; Sa is the sand content (5.5%–93.6%); Si is the silt content (0.5%–69.5%);\(Sa/Cl\) is the sand to clay ratio (1.2%–53.0%); OM is organic matter (0.18%–4.79%); \(\text{CaCO}_{3}\) is calcium carbonate (0.0%–25.2%); Cl is clay content (0.32%–4.74%);\(\text{SLR}_{f}\) is the flat residues; \(\text{SLR}_{s}\) is the standing residues; and \(\text{SLR}_{c}\) is the crop canopy. In this study, the crop seedlings that provide partial canopy cover and protect the soil were calculated as the effects of vegetation cover on wind erosion.

2.3 Statistical methods

2.3.1 Trend analysis
Sen’s slope analysis method was used to calculate the soil wind erosion trend between 1991 and 2015. This method is a non-parametric test that only requires the independence of sample data and does not need to fit the normal distribution (Gilbert, 1987). The trend degree, \(\beta,\) is defined as follows: