Soil is the foundation of agriculture, and water erosion is one of the major ecological issues threatening land resources globally. In China, the phenomenon of water erosion is particularly prominent due to complex topography, diverse climates, and intense human activities—ranging from the myriad gullies of the Loess Plateau to the topsoil removal in the southern red soil regions. Water erosion not only leads to a decline in soil fertility but can also trigger secondary disasters such as debris flows and river siltation. To scientifically assess and predict water erosion risks, soil erosion models have become essential tools. These techniques, which simulate the water erosion process through mathematical formulas, help researchers and decision-makers quantify loss, identify high-risk areas, and provide a basis for soil and water conservation measures.
What has been the trajectory of water erosion model research in China? What are the most widely used models currently? What are the characteristics of research distribution across different regions? What shortcomings need urgent attention? Professor Qingfeng Zhang from Northwest A&F University, along with researchers from multiple institutions, systematically reviewed the research progress of water erosion models in China from 1982 to 2022 using a combination of bibliometric and statistical analysis methods, providing a panoramic perspective to answer these questions. Relevant study has been published in the journal Frontiers of Agricultural Science and Engineering (DOI: 10.15302/J-FASE-2024580).
The researchers screened 786 relevant papers from the China National Knowledge Infrastructure (CNKI) and Web of Science (WoS) databases, covering various aspects such as model application, improvement, development, and qualitative assessment. The analysis revealed a distinct “empirical model dominance” characteristic in China’s water erosion model research—nearly 75% of applications are concentrated on three major empirical models: the Universal Soil Loss Equation (USLE), the Revised Universal Soil Loss Equation (RUSLE), and the Chinese Soil Loss Equation (CSLE). These models establish mathematical relationships by statistically analyzing a large amount of observational data, making them relatively simple to operate, especially suitable for water erosion assessments at regional or large watershed scales. Consequently, they are most widely applied in southeastern and central regions (such as Shaanxi, Yunnan, and Sichuan).
Notably, the focus of research has significantly shifted over time: before 2006, studies were more concerned with describing the characteristics of water erosion, such as loss patterns under different topographies or land uses; after 2006, the focus gradually shifted to analyzing influencing factors (such as rainfall, vegetation, and human activities) and the spatiotemporal evolution of water erosion processes. This transition reflects a deepening of model applications from “identifying the status quo” to “analyzing mechanisms”.
However, the research also reveals multiple challenges in the current application of water erosion models. First is the “verification gap”—many studies directly apply models to evaluate water erosion amounts or distributions but lack comparison with field observational data, raising concerns about the reliability of the results. Second is the “regional limitation”—empirical models are usually developed based on observational data from specific areas; directly applying them to other geographical environments (such as from the Loess Plateau to the southern red soil regions) may lead to errors due to parameter mismatches. Furthermore, research on physical process models (which simulate erosion processes based on hydrodynamic principles) is relatively weak; these models can more accurately describe sediment transport mechanisms, but they are complex in terms of parameters and data requirements, limiting their practical application. Additionally, the mechanistic studies on key processes of water erosion (such as rill development and sediment transport) are still insufficient, constraining model optimization.
To address these issues, the authors suggest: enhancing observations of soil erosion processes and improving the reliability of model validation through long-term, systematic accumulation of field data; establishing methods for data or mathematical formula conversion to account for differences in geographical environments; and deepening research on the mechanisms of erosion processes, particularly exploring the mechanisms of key links, to provide a more solid theoretical foundation for model optimization.
DOI: 10.15302/J-FASE-2024580