Ensemble machine learning methods have been widely used for modeling landslide susceptibility, but there has been no uniform ensemble method for this problem. The main objective of this study is to compare popular ensemble machine learning-based models and apply them to landslides susceptibility mapping. The selected models include the random forest (RF), which is a typical bagging ensemble model, and three advanced boosting models, namely, adaptive boosting (AB), gradient boosting decision trees (GBDT), and extreme gradient boosting (XGBoost). This study considers 94 landslide points and 12 affecting factors. The data are divided into a training dataset consisting of 70% of the overall data, and a validation dataset, containing the remaining 30% of the data. The models are evaluated using the area under the receiver operating characteristic curve (AUC) and three common performance metrics: sensitivity, specificity, and accuracy. The results indicate that the four ensemble models have an AUC of more than 0.8, suggesting that they can appropriately and accurately predict landslide susceptibility maps. In particular, the XGBoost model achieves the best performance among all models, having a sensitivity of 92.86, specificity of 90.00, and accuracy of 91.38. Furthermore, the bagging model has a sensitivity of 89.29, specificity of 86.67, and accuracy of 87.93, and it is superior to the GBDT, which achieves a sensitivity of 86.21, specificity of 86.21, and accuracy of 86.21, and the AB, reaching a sensitivity of 82.14, specificity of 80.00, and accuracy of 81.03. The results presented in this study indicate that the advanced ensemble model, the XGBoost model, could be a promising tool for the selection of ensemble models for predicting landslide susceptibility mapping.

To clarify the long-term dynamics of groundwater and its response to precipitation in the Heilonggang region, China, variation trends of the monitored groundwater were studied according to the Mann–Kendall statistics. Based on observations from four monitoring wells, the persistence and periodicities of the loose porous aquifers, and the interrelation between precipitation and groundwater levels was examined based on a number of tools including autocorrelation, cross-correlation, spectral analysis, and wavelet transform. The results show that the groundwater levels at W1 and W2 present a downward trend while those at W3 and W4 show an upward trend. The dominant time period increases from 2.1 years (upstream) to 3.7 years (downstream). The time lags between aquifers and rainfall at W1, W2, W3 and W4 are 139.14 ± 59.76 days (2008–2020), 23.27 ± 12.03 days (2005–2014), 145.01 ± 68.00 days (2007–2020), and 59.22 ± 26.14 days (2005–2019), respectively. The lags for the pumping years are 1.2~2.1 times of those during the years without pumping. The recharge ratio overestimated by the water table fluctuation method decreases from 0.32 at W2 to 0.17 at W4, suggesting that the site W2 has a good potentiality for groundwater recharge. This research helps us better understand the local groundwater circulation and provide references for groundwater management.

Determining the recharge sources of adjacent old-working water outlets and their hydraulic connections is essential for zoned pollution control in closed coal mines. This study examined two outlets, S1 and S2, 174 m apart in the Chunjingwa closed coal mine area, Shanxi Province, China. Discharge dynamics, hydrochemistry, hydrogen–oxygen stable isotopes, and a goaf pumping test were combined. S2 showed 9.7 times the discharge variation of S1 and responded more rapidly to rainfall. Most contamination indicators had higher median exceedance levels at S2; isotopically, S1 overlapped with goaf boreholes ZK2 and ZK3, whereas S2 was heavier. During pumping, S1 discharge fell by about 91% and partially recovered, whereas S2 remained stable. The evidence identifies two distinct recharge–discharge systems on opposite sides of F2 and indicates limited cross-fault hydraulic communication at the tested scale and duration. F2 is interpreted as a structural divide, not a uniformly impermeable fault. S1 is regulated mainly by goaf-water storage to the south, whereas S2 is dominated by shallow-catchment, rapid-infiltration recharge to the north. The framework provides a practical alternative to artificial tracer tests for outlet-scale source identification and zoned remediation.

The serious shortages and destruction of groundwater resources caused by coal mining, especially group mining, is a major hurdle for the sustainable development of the mining economy, resource management, and preserving the environment that urgently needs to be addressed. Large efforts have been made to protect the groundwater resources affected by coal mining but they have had little effect. The difficulty and uncertainty in investigating the structural changes in roof strata and the relationship between the roof rock fissure formation rate and hydraulic conductivity are the main obstacles hindering the investigation, assessment and management of groundwater resources in roof strata. In this study, we tackled this problem by performing water injection tests in ground surface boreholes and underground boreholes in the roof strata, and pumping tests in wells over the goaf. Clear differences in hydraulic conductivity changes in the collapsed roof strata were observed. These observations provide valuable information on changes in groundwater flow and the distribution of collapsed and fissured zones. The pumping tests also revealed an increase in hydraulic conductivity in the subsidence basin of Quaternary deposits over the goaf. This increased hydraulic conductivity is useful for the assessment of groundwater resources and the prediction of single-well water yields.

Determining the recharge sources of adjacent old-working water outlets and their hydraulic connections is essential for zoned pollution control in closed coal mines. This study examined two outlets, S1 and S2, 174 m apart in the Chunjingwa closed coal mine area, Shanxi Province, China. Discharge dynamics, hydrochemistry, hydrogen–oxygen stable isotopes, and a goaf pumping test were combined. During natural monitoring, discharges at S1 and S2 ranged from 1.172–1.958 and 1.012–8.630 m³/h, respectively, with the variation at S2 (7.618 m³/h) being 9.7 times that at S1 (0.786 m³/h). Most contamination indicators had higher median exceedance levels at S2; for SO₄²⁻, the median was 14.6 at S2 versus 8.4 at S1. Isotopically, S1 overlapped with ZK2 and ZK3, whereas S2 was distinct. During the 27-day pumping test, S1 discharge fell by about 91%, whereas S2 showed no clear response. The evidence identifies two distinct recharge–discharge systems on opposite sides of F2 at the tested scale. F2 is interpreted as a structural divide, not a uniformly impermeable fault. S1 is regulated mainly by goaf-water storage to the south, whereas S2 is dominated by shallow-catchment, rapid-infiltration recharge to the north. The framework provides a practical alternative to artificial tracer tests for outlet-scale source identification and zoned remediation.

Karst piedmont fault overflow springs are widely developed in structurally controlled, basin-margin settings, where basin-bounding faults obstruct regional groundwater flow. However, the coupled fault blocking, fault conduction, and caprock sealing mechanisms governing their genesis remain insufficiently quantified. This study investigates the Shentou Spring system in Shanxi Province, North China—a typical piedmont fault overflow spring—to develop a three-dimensional genetic model characterised by coupled blocking, conduction, and overflow processes. Integrating borehole datasets, multi-year groundwater level monitoring records, hydrochemical and isotopic measurements, and detailed structural mapping, this study identifies three key controlling mechanisms. First, spatial variations in the throw of the Mayi Fault control fault blocking efficiency, partitioning the fault zone into complete barrier and semi-permeable segments, which underpins the incomplete drainage behaviour of the spring system. Second, the Gengzhuang Fault intersects the high-permeability Qilihe and Yuanzihe groundwater flow zones, acting as the primary conduit that transports groundwater from distant recharge areas to the discharge zone. Third, the Quaternary caprock in the discharge area features a critical thickness threshold of approximately 30 m; confined karst groundwater breaches the overlying caprock and forms spring outlets where caprock thickness falls below this threshold. The proposed tripartite coupled model provides a semi-quantitative framework for interpreting the genesis of piedmont fault overflow springs. In practical terms, it supports the delineation of fault-conduit protection zones and the design of long-term water-quality monitoring networks along fault-controlled flow paths.