Simulation performance and impact factors of soil surface roughness under different tillage practices based on the Monte Carlo algorithm

Soil surface roughness (SSR) is widely employed to quantify the spatial variability of soil surface height in farmland, which also plays a critical role in soil erosion. However, traditional methods for measuring SSR are often time-consuming and labor-intensive, with limitations in measurement scale and accuracy. In this study, simulation approach for random and oriented SSR under Gaussian and exponential distributions was developed using the Monte Carlo algorithm. The accuracy and effectiveness of simulation results were validated through field measurements during the soybean growth period under various tillage practices. Thereafter, variations in SSR and their influencing factors across different tillage practices were systematically analyzed. Results demonstrated that the simulation of soil surface height under ridge-based tillage outperformed no-tillage. Additionally, discrepancies were observed in the fitting performance of SSR parameters derived from measured and simulated soil surface heights. Specifically, root mean square height (RMSH) and mean deviation (R_a) exhibited the best fitting performance, followed by root mean square slope (σ_m), while correlation length (CL) showed the poorest agreement between measured and simulated values. Correlation analysis and analysis of variance further revealed that the correlation and stability between measured and simulated SSR parameters followed the order: no-tillage > reduced tillage > conventional tillage > rotary tillage > combined tillage. During the crop growth period, various factors such as field management, precipitation, wind erosion, and gravity influenced SSR under different tillage practices. The results of this study contribute to broadening the methodologies for obtaining SSR, and enhancing the understanding of farmland surface processes.