Accurate runoff forecasting is of great significance for the optimization of water resource management and regulation. Given such a challenge, a novel compound approach combining time-varying filtering-based empirical mode decomposition (TVFEMD), sample entropy (SE)-based subseries recombination, and the newly developed deep sequential structure incorporating convolutional neural network (CNN) into a gated recurrent unit network (GRU) is proposed for monthly runoff forecasting. Firstly, the runoff series is disintegrated into a collection of subseries adopting TVFEMD, considering the volatility of runoff series caused by complex environmental and human factors. The subseries recombination strategy based on SE and recombination criterion is employed to reconstruct the subseries possessing the approximate complexity. Subsequently, the newly developed deep sequential structure based on CNN and GRU (CNNGRU) is applied to predict all the preprocessed subseries. Eventually, the predicted values obtained above are aggregated to deduce the ultimate prediction results. To testify to the efficiency and effectiveness of the proposed approach, eight relevant contrastive models were applied to the monthly runoff series collected from Baishan reservoir, where the experimental results demonstrated that the evaluation metrics obtained by the proposed model achieved an average index decrease of 44.35% compared with all the contrast models.

To construct an accurate and stable approach for water inflow forecasting, a series of advanced and effective techniques, such as variational mode decomposition (VMD), outlier robust extreme learning machine (ORELM) and multi-objective grey wolf optimizer (MOGWO), are appropriately integrated into this study. Considering that the influence of the mode number on the VMD decomposition effectiveness, such an argument is determined by observing the converged centre frequency distribution among the components. Then the characteristic items of water inflow series are extracted by VMD, thus obtaining a series of sub-components. Afterwards, ORELM is applied to predict each component, where the parameters of ORELM are optimized by MOGWO with multi-objective functions including forecasting accuracy and stability. Correspondingly, the aggregation of all components' prediction values is considered as the final results. The experimental results obtained by performing eight various models on real-time data demonstrate that the supplementary modules achieve positive effects on the improvement of prediction accuracy, where the proposed model implements an average performance promotion of 48.43% compared with the contrastive models.

Revealing the water use pattern of plants influenced by coal-mining-caused land subsidence is crucial to understand plant–water interactions and guide ecological restoration. However, available information on herbaceous plants, the dominant species in most arid and semi-arid regions with abundant coal resources, remains inadequate. We investigated the water use patterns of Stipa bungeana Trin. by measuring soil water content, root distribution, and stable isotopes of hydrogen (δ²H) and oxygen (δ¹⁸O) of soil water and plant stem water both before and after a rainfall event. The results revealed that prior to rainfall, both areas exhibited a low soil water content with no discernible difference in soil drought. However, the soil waters δ²H and δ¹⁸O were found to be more enriched at varying depths within the subsidence area, indicating a heightened level of soil evaporation. Both soil water content and soil water isotopic composition responded sensitively to rainfall, with rainfall primarily replenishing the shallow layer (0–20 cm), thereby reflecting an infiltration mode dominated by piston flow. More water seeped into deeper soil layers in the subsidence area compared to the non-subsidence area, with more preferential flow. Before rainfall, the sources of plant water uptake were consistent both at shallow and deep soil layers, implying that the proportion of water uptake gradually decreased with increasing depth. After rainfall, the sources of plant water uptake differed slightly between the two soil layers. The plants in non-subsidence and subsidence areas dominantly extracted soil water at depths greater than 10 cm and 20 cm, respectively. The root system in the subsidence area was more developed than that in the non-subsidence area. Plant water uptake was primarily influenced by the spatial distribution of roots, as well as the post-rainfall water distribution, regardless of whether they were in the subsidence area or not. Although land subsidence affected soil water transport, the water uptake pattern of Stipa bungeana Trin. was similar before and after rainfall, indicating the adaptive growth of plants through their roots in the subsidence area. The high adaptability of herbs such as Stipa bungeana Trin. makes them a viable option for vegetation restoration in subsidence areas. This study has significant implications for evaluating plant–water relationships in subsidence areas due to coal mining, thereby providing a fundamental basis and valuable reference for ecological restoration and management strategies within such affected regions.

Coal gangue is a fine-grained mineral with nutrient content, which can be used as a potential soil amendment. Nevertheless, current research on using coal gangue to improve soil water and support plant growth is still insufficient. In this study, coal gangue powder (CGP) was added to aeolian sandy soil. We compared the soil hydraulic properties and plant growth of original aeolian sandy soil (CK) and different CGP application rates (10% and 20%). The results indicated that the application of CGP transformed the soil texture from sandy to loamy, significantly reduced soil bulk density and saturated hydraulic conductivity (Ks) values, altered the soil water characteristic curve, enhanced soil water-holding capacity, and increased plant-available water. Compared with the CK group, the emergence rate of alfalfa seeds increased from approximately 50% to over 70% after CGP application. During the growth process, CGP application significantly elevated the net photosynthetic rate, transpiration rate, and stomatal conductance of alfalfa leaves. Rapid fluorescence kinetics monitoring of leaves demonstrated that alfalfa treated with CGP had a higher efficiency in light energy utilization. However, the photosynthetic capacity of leaves did not improve as the CGP application rate increased from 10% to 20%, suggesting that excessive CGP addition did not continuously benefit plant gas exchange. In conclusion, CGP application can improve the soil hydraulic properties of aeolian sandy soil and support plant growth and development, which is conducive to reducing the accumulated amount of coal gangue, alleviating plant water stress, and promoting ecological restoration in arid mining areas. We recommend a 10% addition of coal gangue powder as the optimal amount for similar soils.

Understanding the impacts of land use and land cover changes on ecosystem service values (ESVs) is crucial for effective ecosystem management; however, the intricate relationship between these factors in coal mining regions remains underexplored. In particular, the influence of coal mining activities on these dynamics is insufficiently understood, leaving a gap in the literature that hinders the development of robust management strategies. To address this gap, we investigated the interplay between land use change and the ESV at the interface of Yang Coal Mine No. 2 and the Shanxi Yalinji Guanshan Provincial Nature Reserve in Yangquan City, Shanxi Province. Using Landsat 8 remote sensing data from 2013 to 2021, our approach incorporated analyses using the Google Earth Engine (GEE) platform. We employed a random forest algorithm to classify land use patterns and calculated key indices—including the normalized difference vegetation index (NDVI), normalized difference built-up index (NDBI), enhanced vegetation index (EVI), and bare soil index (BSI)—which were combined with topographic features. Land use change dynamics were quantified via a transfer matrix, while changes in the ESV were evaluated using the ecosystem sensitivity index and ecological contribution rate. Our results revealed notable fluctuations: forestland increased from 2013 to 2018 before declining sharply from 2019 to 2021; grassland displayed similar variability; and constructed land experienced a continual expansion. Correspondingly, the overall ESV increased by 28.6% from 2013 to 2019, followed by a 19.5% decline in 2020 and 2021, with forest and grassland’s ESVs exhibiting similar trends. These findings demonstrate that land use changes, particularly those that are driven by human activities such as coal mining, have a significant impact on ecosystem service values in mining regions. By unraveling the nuanced relationship between land use dynamics and ESVs, our study not only fills the gap in the literature but also provides valuable insights for developing more effective ecosystem management strategies, ultimately advancing our understanding of ecosystem dynamics in human-impacted landscapes.