Du Mingliang

Dr. Du Mingliang

orcid

Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
Verified
Journal Article
Open access
PDF file
SHShut Hu
Shut Hu
DMDu Mingliang
Du Mingliang
家杨家云 杨
家云 杨
YLYankun Liu
Yankun Liu
ZTZiyun Tuo
Ziyun Tuo
XMXiaofei Ma
Xiaofei Ma
Accurate forecasting of groundwater level dynamics poses a critical challenge for sustainable water management in arid regions. However, the strong spatiotemporal heterogeneity inherent in groundwater systems and their complex interactions between natural processes and human activities often limit the effectiveness of conventional prediction methods. To address this, a hybrid CNN-LSTM deep learning model is constructed. This model is designed to extract multivariate coupled features and capture temporal dependencies from multi-variable time series data, while simultaneously simulating the nonlinear and delayed responses of aquifers to groundwater abstraction. Specifically, the convolutional neural network (CNN) component extracts the multivariate coupled features of hydro-meteorological driving factors, and the long short-term memory (LSTM) network component models the temporal dependencies in groundwater level fluctuations. This integrated architecture comprehensively represents the combined effects of natural recharge–discharge processes and anthropogenic pumping on the groundwater system. Utilizing monitoring data from 2021 to 2024, the model was trained and tested using a rolling time-series validation strategy. Its performance was benchmarked against traditional models, including the autoregressive integrated moving average (ARIMA) model, recurrent neural network (RNN), and standalone LSTM. The results show that the CNN-LSTM model delivers superior performance across diverse hydrogeological conditions: at the upstream well AJC-7, which is dominated by natural recharge and discharge, the Nash–Sutcliffe efficiency (NSE) coefficient reached 0.922; at the downstream well AJC-21, which is subject to intensive pumping, the model maintained a robust NSE of 0.787, significantly outperforming the benchmark models. Further sensitivity analysis reveals an asymmetric response of the model’s predictions to uncertainties in pumping data, highlighting the role of key hydrogeological processes such as delayed drainage from the vadose zone. This study not only confirms the strong applicability of the hybrid deep learning model for groundwater level prediction in data-scarce arid regions but also provides a novel analytical pathway and mechanistic insight into the nonlinear behavior of aquifer systems under significant human influence.

23 September 2025 in Sustainability
Verified
Journal Article
Open access
PDF file
YLYankun Liu
Yankun Liu
DMDu Mingliang
Du Mingliang
XMXiaofei Ma
Xiaofei Ma
SHShuting Hu
Shuting Hu
ZTZiyun Tuo
Ziyun Tuo
Groundwater level (GWL) prediction in arid regions faces two fundamental challenges in conventional numerical modeling: (i) irreducible parameter uncertainty, which systematically reduces predictive accuracy; (ii) oversimplification of nonlinear process interactions, which leads to error propagation. Although machine learning (ML) methods demonstrate strong nonlinear mapping capabilities, their standalone applications often encounter prediction bias and face the accuracy–generalization trade-off. This study proposes a hybrid TCN–Transformer–LSTM (TTL) model designed to address three key challenges in groundwater prediction: high-frequency fluctuations, medium-range dependencies, and long-term memory effects. The TTL framework integrates TCN layers for short-term features, Transformer blocks to model cross-temporal dependencies, and LSTM to preserve long-term memory, with residual connections facilitating hierarchical feature fusion. The results indicate that (1) at the monthly scale, TTL reduced RMSE by 20.7% (p < 0.01) and increased R2 by 0.15 compared with the Groundwater Modeling System (GMS); (2) during abrupt hydrological events, TTL achieved superior performance (R2 = 0.96–0.98, MAE < 0.6 m); (3) PCA revealed site-specific responses, corroborating the adaptability and interpretability of TTL; (4) Grad-CAM analysis demonstrated that the model captures physically interpretable attention mechanisms—particularly evapotranspiration and rainfall—thereby providing clear cause–effect explanations and enhancing transparency beyond black-box models. This transferable framework supports groundwater forecasting, risk warning, and practical deployment in arid regions, thereby contributing to sustainable water resource management.

16 April 2026 in Sustainability
Verified
Journal Article
Open access
PDF file
YLYankun Liu
Yankun Liu
DMDu Mingliang
Du Mingliang
XMXiaofei Ma
Xiaofei Ma
LCLongyao Chen
Longyao Chen
The “Golden Triangle” area of the Kuitun River Basin is a typical arid irrigated region where limited surface-water supply and intensive groundwater abstraction have caused persistent groundwater decline and overexploitation. This study developed a one-way semi-loosely coupled GWAS–GMS/MODFLOW framework for allocation optimization, spatial projection, groundwater simulation, and effectiveness evaluation. An improved NSGA-II-S algorithm was used to optimize multi-source water allocation, and the resulting management-unit outputs were projected into MODFLOW source–sink terms using area-weighted transformation and mass conservation. GWAS satisfactorily reproduced monthly runoff at the Jiangjunmiao station, and GMS/MODFLOW reasonably matched observed groundwater heads at six national monitoring wells (R2 = 0.952–0.998; RMSE = 0.15–0.355 m; MAE = 0.12–0.315 m). Three 2030 scenarios were compared: baseline, optimized allocation, and combined pumping restriction with allocation adjustment. Under the baseline scenario, groundwater storage remained in deficit. The optimized allocation scenario expanded the quasi-stable zone (−0.1 to 0.1 m/a), whereas the combined scenario most effectively suppressed severe decline hotspots, reducing the proportion of the −3 to −2 m/a zone from 11.22% to 2.68%. These results indicate that recharge reallocation and pumping restriction play complementary roles in groundwater overexploitation control. The framework provides a quantitative basis for coordinated multi-source water management in arid irrigated basins.

23 February 2026 in Water
Verified
Journal Article
Open access
PDF file
ZTZiyun Tuo
Ziyun Tuo
DMDu Mingliang
Du Mingliang
BWBin Wu
Bin Wu
CZChangjiang Zou
Changjiang Zou
SHShuting Hu
Shuting Hu
XMXiaofei Ma
Xiaofei Ma
Land subsidence induced by excessive groundwater abstraction has emerged as a major geo-environmental hazard in arid oasis regions, calling for reproducible methods to quantitatively assess the abstraction-reduction–subsidence response and to support zoned management. This study integrates Sentinel-1A PS-InSAR deformation data with groundwater-level measurements to develop and calibrate a MODFLOW-SUB model that couples three-dimensional groundwater flow and one-dimensional skeletal compaction. The InSAR deformation field is used to constrain the conceptual model and key parameters. Four abstraction-reduction scenarios (20%, 40%, 60%, and 80%) are designed to characterize response curves using indicators such as maximum cumulative subsidence, annual subsidence rate, and the area exceeding specified thresholds. In addition, a multi-criteria composite index integrating driving forces, geological susceptibility, and exposure is applied for hazard zoning and scenario comparison. The results show that PS-InSAR constraints improve the spatial agreement of the simulations. The time-series RMSE between simulated and InSAR-derived deformation is approximately 20 mm, and the end-of-period cumulative subsidence error is within 10 mm. From 2019 to 2020, the maximum cumulative subsidence reached 166 mm, and the peak subsidence rate reached 101 mm/a. A clear lag between groundwater-level fluctuations and subsidence is observed, with the maximum correlation occurring at ~35 days for ACJ-1 and ~59–83 days for ACJ-2. This spatial variability is associated with the thickness and permeability of clay layers. Forecasts for 2021–2028 indicate that, under business-as-usual abstraction, maximum subsidence may reach 695 mm. Across scenarios, subsidence mitigation exhibits diminishing marginal returns with increasing abstraction reduction. Under the adopted model settings, a 20% reduction in abstraction yields substantial decreases in maximum subsidence and high-hazard area, representing a practical trade-off between mitigation benefits and water-use costs. Overall, the integrated workflow of monitoring, inversion, coupled modeling, scenario analysis, and zoning, together with the resulting zoned management recommendations, provides decision support for land-subsidence mitigation and water-allocation planning in arid oasis regions.