Enhancing the representation of human activities’ impact on surface processes to improve the model’s ability to simulate reality on global scale

Over the past decades, the World is suffering from a serious process of land degradation as a result of global climate change and the increasingly acute conflicts among population, resources and the environment. According to IGBPS (2018, 2023), the area of degraded soil worldwide is continuously increasing, and the global soil health situation is still deteriorating, with which soil erosion was regarded as the 1st threat to the planet soil. In order to reverse this trend towards land degradation, many regions and countries have carried out sustained and painstaking initiatives for soil and water conservation, whose results has been monitored using different methodologies, and providing efficient recommendation for local government. It is urgent to adopt state of the art technologies including the latest earth observation techniques to evaluate global soil erosion status and soil conservation benefit in a standardized way. Accurately achieving the status of global soil erosion and the distribution and types of soil and water conservation measures, will help to illustrate the difference in effectiveness of soil/water conservation practices, improve current technologies, promote soil/water conservation measures, eliminate interregional imbalances and promote the United Nations’ Sustainable Development Goals using solid science. Among numerous erosion models, only the USLE-family models are frequently employed in regional and global-scale soil erosion studies. Current research has established the distribution patterns of soil erosion at the global scale. However, significant challenge remains in balancing a model’s ability to represent real surface processes, its accuracy, and the target objectives f different levels of government. For global erosion surveys and mapping (GSERmap), we will draw upon experiences from China’s 2010 Soil and Water Conservation Census. By employing an unequal probability sampling units and investigation methods, combined with high-resolution remote sensing imagery, we aim to enhance the models’ simulation capability of real-world surface processes while maintaining a certain accuracy.