1 INTRODUCTION
Soil aggregates are the basic units of structure in soil. The breakdown and dispersion of soil aggregates caused by raindrop splashing, initiate soil erosion (Bronick and Lal, 2005; Legout et al., 2005; Xiao et al., 2017). Soil aggregate stability plays a key role in affecting surface runoff and soil erosion; furthermore, it has been widely used to evaluate soil structural stability and soil quality (Six et al., 2000; Ghadir et al., 2007; Anderson et al., 2019). Numerous studies have been conducted on aggregates in soil erosion (Fu et al., 2017, 2019), and these studies mostly analyzed the particle size distribution of split aggregates after rainfall. During a rainfall event, the fragmentation and separation of aggregates was mainly caused by large-diameter raindrops; as the raindrop diameter increased, the degree of fragmentation in the aggregates became stronger (Fu et al., 2017; Li et al., 2018). Additionally, Fu et al. (2019) used a simulated rainfall test to analyze the extent of soil splash erosion and the distribution of particle sizes and determined the effects of different rainfall energy levels on the breakdown of soil aggregates. Furthermore, Fu et al. (2020) assessed the effects of secondary raindrop splash erosion on aggregate fragmentation and fragment size distribution and found that secondary raindrop splash erosion would cause soil aggregates to break again, leading to a reduction in soil fertility and productivity. Obviously, previous studies have ignored the impacts of raindrop splash on the microstructure of aggregates.
In recent years, the development of X-ray computer scanning (CT) technology has allowed for its application in the study of soil structure and soil aggregates (Peth et al., 2008; Munkholm et al., 2012; Garbout et al., 2013; Gao et al., 2019). Through synchrotron-based X-ray microcomputed tomography (SR-μCT) scanning, high-resolution images of soil can be obtained, and three-dimensional qualitative and quantitative analysis of soil aggregates can be performed, thereby analyzing the internal microstructure of aggregates in a nondestructive and comprehensive manner (Taina et al., 2008; Ma et al., 2015; Han et al., 2019). Some research has utilized CT scanning technology to quantify the effects of different land use types, fertilization conditions, and types of vegetation restoration and succession on soil pore structure and aggregates (Luo et al., 2010; Zhou et al., 2012; Dal Ferro et al., 2013; Zhao et al., 2017). Hu et al. (2016) and Li et al. (2019) applied CT scanning technology to analyze soil macropores and root structure from alpine vegetation and different ecosystems in the Qinghai Lake Basin of China in an effort to explore the impact of roots on soil macropore network characteristics. In addition, work from Ma et al. (2015) based on SR-μCT scanning technology investigated the influences of pore characteristics on water stability and the tensile strength of aggregates under wetting and drying cycles. Additional research has used high-resolution CT to characterize the development of crust over time, and the results showed that the cumulative porosity varied according to soil depth (Lee et al., 2008). With the help of simulation rainfall experiments and SR-μCT scanning, Li et al. (2018) analyzed the effects of raindrop splashing on the fragmentation mechanism of aggregate microstructure from a two-dimensional perspective.
As we all know, the Loess Plateau is a region among those with the most serious soil erosion in the world, and one of its main causes of erosion is water erosion. The fragmentation and dispersion of topsoil aggregates caused by raindrops is the first step of soil erosion, but it is unclear how rainfall causes microstructural changes in soil aggregates. Although SR-μCT scanning has been widely used in soil research, this does not hold true in the field of soil erosion; to date, CT scanning has rarely been employed to analyze the impact of rainfall on the microstructure of aggregates. However, the microstructure of soil aggregates determines the soil stability and quality characteristics (Zhou et al., 2012). Therefore, utilizing SR-μCT and image analysis, the aims of this research were (1) to compare the differences in the three-dimensional microstructure of soil aggregates under raindrops and (2) to analyze the relationship between the rainfall intensity and microscopic characteristics of the soil aggregates. The results presented here may be beneficial by revealing the mechanisms of aggregate fragmentation, pore clogging and the formation of a surface crust during soil erosion.