Agricultural irrigation of the South Jingyang tableland in Shaanxi Province, China has led to a continuous rise of the groundwater level and has triggered a series of loess landslides, thereby seriously affecting the life and property safety of local residents. Research shows that the major cause of the landslide in the loess layer of the South Jingyang tableland is the rising groundwater level. Therefore, the research on the formation mechanism of landslide in this area should include the investigation of the stratigraphic structure and groundwater level distribution characteristics. On this basis, a series of approaches, such as electrical resistivity tomography (ERT), borehole, and laboratory tests, was carried out on the South Jingyang tableland, and the groundwater level distribution and stratigraphic structure in the study area were determined. The qualitative relationship between resistivity value and water content at different depths was detected using the inversion results of ERT and borehole data. Through laboratory tests, the quantitative relationship between resistivity values under different water contents was established. The precise depth of the groundwater level was inferred by connecting the qualitative relationship with the quantitative relationship, and then a detailed 3D geological model was established by linking the inversion results of ERT with the field borehole lithology data and geological survey data. The detection results show that when the qualitative and quantitative analyses of the ERT inversion results were combined, the distribution of the groundwater level was accurately judged. The ERT is effective in reflecting the stratigraphic structure and hydrological characteristics of the Loess Plateau, and its potential as a supplementary technology for detecting the groundwater level is reasonable. This study addresses the limitation and inaccuracy in determining the stratum structure and groundwater level by solely relying on borehole information or ERT. The established 3D geological model not only provides a basis for the study of groundwater table fluctuation, but also a technical guidance for the stability evaluation of loess slope, landslide prediction, and early warning in the study area.

The appropriate selection of machine learning samples forms the foundation for utilizing machine learning models. However, in landslide susceptibility evaluation, discrepancies arise when non-landslide samples are positioned within areas prone to landslides or demonstrate spatial biases, leading to differences in model predictions. To address the impact of non-landslide sample selection on landslide susceptibility predictions, this study uses the western region of Henan Province as a case study. Utilizing historical data, remote sensing interpretation, and field surveys, a sample dataset comprising 834 landslide points is obtained. Ten environmental factors, including elevation, slope, aspect, profile curvature, land cover, lithology, topographic wetness index, distance from river, distance from faults, and distance from road, are chosen to establish an evaluation index system. Negative sample sampling areas are delineated based on the susceptibility assessment outcomes derived from the information value model. Two sampling strategies, whole-region random sampling (I) and partition-based random sampling (II), are employed. Random Forest (RF) and Back Propagation Neural Network (BPNN) models are used to forecast and delineate landslide susceptibility in the western region of Henan Province, with prediction accuracy evaluated. The model prediction accuracy is ranked as follows: II-BPNN (AUC = 0.9522) > II-RF (AUC = 0.9464) > I-RF (AUC = 0.8247) > I-BPNN (AUC = 0.8068). Under the Receiver Operating Characteristic (AUC) curve and accuracy, the II-RF and II-BPNN models exhibit increases in the region by 12.17% and 15.61%, respectively, compared to the I-RF and I-BPNN models. Moreover, the II-BPNN model shows improvements over the I-BPNN model with increases in AUC and accuracy by 14.54% and 16.52%, respectively. This indicates enhancements in model performance and predictive capability. In terms of recall and specificity, the II-RF and II-BPNN models demonstrate increases in recall by 15.09% and 17.47%, respectively, and in specificity by 15.80% and 14.99%, respectively. These findings suggest that the optimized models have better predictive capabilities for identifying landslide and non-landslide areas, effectively reducing the uncertainty introduced by point data in landslide risk prediction.

The Henan area of the middle and lower reaches of the Yellow River is situated within the third sedimentary loess area, positioned as the southeasternmost segment within the transitional belt connecting the Loess Plateau with the North China Plain. Addressing concerns related to loess collapse, landslides, and subgrade settlement across various regions attributable to the collapsible nature of Malan loess in western Henan, this study undertook collapsibility testing of undisturbed Malan loess in the province. The different mechanisms of loess collapsibility in different regions were explained from the microstructure by using the indoor immersion-compression test double-line method, scanning electron microscope (SEM), and particles and cracks analysis system (PCAS). The relationship between quantitative factors of microstructure and collapsibility of loess was analyzed by linear regression analysis. The findings indicate that under identical overburden pressure and immersion conditions, the collapsibility of Malan loess in western Henan diminishes progressively from west to east. Microstructural tests were conducted on various loess specimens using scanning electron microscopy, revealing that the distribution of loess particles is notably concentrated in the Xingyang and Gongyi areas, leading to a reduction in pore area compared to the Shanzhou and Mianchi areas. While the Mianchi and Shanzhou areas exhibit a loose arrangement of loess particles, those in Xingyang and Gongyi are comparatively denser. Analysis of microstructural images through the particles and cracks analysis system elucidated that the pore arrangement in the Gongyi and Xingyang areas is more stable than in the Mianchi and Shanzhou areas. Additionally, there is a gradual concentration in particle distribution, accompanied by an increase in agglomeration degree. According to the analysis and comparison of microstructure and quantitative parameters of four groups of loess samples before and after collapsibility, it is revealed that the change mechanism underlying loess collapsibility in various regions of western Henan primarily stems from the external factors influencing the microstructural alterations within the loess. The microstructural determinants contributing to collapsibility changes in different regions encompass three principal aspects: Firstly, modifications in the grain morphology of the Malan loess skeleton in western Henan are notable. Secondly, variations in the internal pore characteristics of loess microstructure are observed. Thirdly, disparities exist in the interconnections between soil particles. The findings of this research hold significant worth for improving construction safety and geological hazard prevention within the Loess region of western Henan.

A novel and highly efficient photocatalyst of a AgBiO₃/BiOCl heterojunction has been developed via a facile water bath and in situ precipitation method. The photocatalytic activities of the catalysts were investigated by the degradation of ciprofloxacin (CIP) under visible-light irradiation (>420 nm). The experiment results revealed that the photocatalytic performance of the optimized AgBiO₃/BiOCl heterojunction was much higher than pure AgBiO₃ and BiOCl. The degradation efficiency of the as-prepared AgBiO₃/BiOCl heterojunction (ABC-30) for CIP could reach 88% within 160 min, with 2.89 and 3.76 times higher activity than pure AgBiO₃ and BiOCl, respectively. The improved photocatalytic performance of AgBiO₃/BiOCl was attributed to the synergistic effect of the enhanced light absorption range and effective separation and transfer of the photo-induced charge carrier. The optimized heterojunction showed broad-spectrum catalytic activities towards various organic contaminants. The degradation efficiencies varied with the nature of the pollutant and decreased in the following order: Lanasol Red 5B (100%) > methyl orange (99%) > methylene blue (98%) > tetracycline (92%) > ciprofloxacin (88%) > ofloxacin (85%) > norfloxacin (78%) > rhodamine B (59%) > metronidazole (43%) > phenol (40%) > carbamazepine (20%). Furthermore, the trapping experiments and ESR indicated that superoxide radicals and holes were the main reactive species.