In this study, Logistics Regression (LR), Support Vector Machine (SVM), Random Forest (RF), Gradient Boosting Machine (GBM), and Multilayer Perceptron (MLP) machine learning algorithms are combined with GIS techniques to map landslide susceptibility in Shexian County, China. By using satellite images and various topographic and geological maps, 16 landslide susceptibility factor maps of Shexian County were initially constructed. In total, 502 landslide and random safety points were then using the “Extract Multivalues To Points” tool in ArcGIS, parameters for the 16 factors were extracted and imported into models for the five algorithms, of which 70% of samples were used for training and 30% of samples were used for verification, which makes sense for date symmetry. The Shexian grid was converted into 260130 vector points and imported into the five models, and the natural breakpoint method was used to divide the grid into four levels: low, moderate, high, and very high. Finally, by using column results gained using Area Under Curve (AUC) analysis and a grid chart, susceptibility results for mapping landslide prediction in Shexian County was compared using the five methods. Results indicate that the ratio of landslide points of high or very high levels from LR, SVM, RF, GBM, and MLP was 1.52, 1.77, 1.95, 1.83, and 1.64, and the ratio of very high landslide points to grade area was 1.92, 2.20, 2.98, 2.62, and 2.14, respectively. The success rate of training samples for the five methods was 0.781, 0.824, 0.853, 0.828, and 0.811, and prediction accuracy was 0.772, 0.803, 0.821, 0.815, and 0.803, respectively; the order of accuracy of the five algorithms was RF > SVM > MLP > GBM > LR. Our results indicate that the five machine learning algorithms have good effect on landslide susceptibility evaluation in Shexian area, with Random Forest having the best effect.

Fully mechanized top-coal caving mining with high mining height, hard roofs and strong mining pressure are popular in the Shendong mining area, China. The occurrence of dynamic disasters, such as rock burst, coal and gas outburst, mine earthquakes and goaf hurricanes during the coal exploitation process under hard roof conditions, pose a threat to the safe production of mines. In this study, the characteristics of overburden fracture in fully mechanized top-coal caving with a hard roof and high mining height are studied, and the technology of advanced weakening by hard roof staged fracturing was proposed. The results show that the hard roof strata collapse in the form of large “cantilever beams”, and it is easy to release huge impact kinetic energy, forming impact disasters. After the implementation of advanced hydraulic fracturing, the periodic weighting length decreases by 32.16%, and the length of overhang is reasonably and effectively controlled. Ellipsoidal fracture networks in the mining direction of the vertical working face, horizontal fracture networks perpendicular to the direction of the working face, and near-linear fracture planes dominated by vertical fractures were observed, with the accumulated energy greatly reduced. The effectiveness of innovation technology is validated, and stress transfer, dissipation and dynamic roof disasters were effectively controlled.

The main purpose of this research is to apply the logistic regression (LR) model, the support vector machine (SVM) model based on radial basis function, the random forest (RF) model, and the coupled model of the whale optimization algorithm (WOA) and genetic algorithm (GA) with RF, to make landslide susceptibility mapping for the Ankang City of Shaanxi Province, China. To this end, a landslide inventory map consisting of 4278 identified landslides is randomly divided into training and test landslides in a ratio of 7 : 3. The 15 landslide influencing factors are selected as follows: slope aspect, slope degree, elevation, terrain curvature, plane curvature, profile curvature, surface roughness, distance to faults, distance to roads, landform, lithology, distance to rivers, rainfall, stream power index (SPI), and normalized difference vegetation index (NDVI), and the potential multicollinearity problem among these factors is detected by Pearson correlation coefficient (PCC), variance inflation factor (VIF), and tolerance (TOL). We evaluate the performance of the model separately by statistical training and test dataset metrics, including sensitivity, specificity, accuracy, kappa, mean absolute error (MSE), root mean square error (RMSE), and area under the receiver operating characteristic curve. The training success rates of LR, SVM, RF, WOA-RF, and GA-RF models are 0.7546, 0.8317, 0.8561, 0.8804, and 0.8957; the testing success rates are 0.7551, 0.8375, 0.8395, 0.8348, and 0.85007. The results show that the GA significantly improves the predictive power of the RF model. This study provides a scientific reference for disaster prevention and control in this area and its surrounding areas.

Landslides caused countless economic and casualty losses in China, especially in mountainous and hilly areas. Landslide susceptibility mapping is an important approach and tool for landslide disaster prevention and control. This study presents a landslide susceptibility assessment using frequency ratio (FR) and index of entropy (IOE) models within a geographical information system for She County in the mountainous region of South Anhui, China. First, the landslide locations were ascertained in the study area using historical landslide records, aerial photographs, and multiple field surveys. In all, 502 landslides were identified and randomly divided into two groups as training (70%) and validation (30%) datasets. Additionally, the landslide-influencing factors, including slope angle, slope aspect, curvature, landform, lithology, distance to faults, distance to roads, distance to rivers, rainfall, and normalized difference vegetation index, were selected and their relative importance and weights were determined by FR and IOE models. The results show that the very high and high susceptibility classes cover nearly 50% of the study area. Finally, the comprehensive performance of the two models was validated and compared using receiver operating characteristic curves. The results demonstrated that the IOE model with the area under the curve (AUC) of 0.802, which is slightly better in prediction than the FR model (AUC = 0.786). The interpretation of the susceptibility map indicated that landform, slope degree, and distance to rivers plays a major role in landslide occurrence and distribution. The research results can be used for preliminary land use planning and hazard mitigation purposes.

Water inrush disaster is one of the major disasters affecting the production safety of coal mines following roof caving, fire, gas outburst, and dust explosion disasters. It is urgent to reveal the water inrush mechanism and take effective measures to prevent the disasters. More than 80% of water inrush accidents occur around geological structural zones such as faults and karst collapse columns (KCCs). The water inrush events from KCCs caused huge economic losses and heavy casualties, and the water inrush process often shows certain hysteresis characteristics. Taking the water inrush disaster from a KCC during roadway excavation in PanEr Coal Mine of Huainan Mining Area as the case study, the delayed inrush mechanism of KCC was analyzed from the aspects of floor failure, KCC activation, seepage transition, and water inrush development characteristics. The results show that the rock mechanical properties and the excavation depth are the main factors affecting the floor failure characteristics. The seepage transformation from pore flow to fracture flow and pipeline flow, with the change in internal composition structure, is the internal mechanism of the delayed water inrush from KCC. The research is of great significance for the prediction and prevention of water inrush disasters from KCCs.

Mining activities interfere into the natural groundwater chemical environment, which may lead to hydrogeochemical changes of aquifers and mine water inrush disasters. The study of hydrogeochemical evolution processes of underground aquifers is helpful to the prevention and control of mine water inrush. The results show that the study area is mainly impacted by four hydrogeochemical processes: dissolution, cation exchange, desulfurization and reduction, and pyrite oxidation. The Cenozoic aquifers are dominated by carbonate dissolution and desulfurization. The Permian aquifers are impacted mainly by cation exchange and sulfate dissolution, followed by pyrite oxidation. The Carboniferous aquifers are mainly impacted by dissolving sulfate, followed by pyrite oxidation and cation exchange. The hydrogeochemical evolution of the aquifers was controlled by mining activities and tectonic changes, and a certain regularity in space. For the Cenozoic aquifers, sulfate dissolution and cation exchange increase from west to east, and desulfurization weakens. For the Permian aquifers, cation exchange and sulfate dissolution are stronger near synclines and faults, pyrite oxidation is enhanced, and desulfurization decreases from the middle to the east of the mining area. For the Carboniferous aquifers, there is a higher dissolution of rock salt, pyrite oxidation, and cation exchange from west to east, and the desulfurization effect weakens.