Water

Hydrodynamic models of river networks are commonly used for flood disaster simulation, and the accuracy of model parameter settings directly affects the reliability of simulation results. Among these, Manning’s roughness coefficient is the core parameter for calibrating one-dimensional(1D) hydrodynamic models, as it is the most sensitive and frequently adjusted parameter. Taking the Yunxi Area of Huai’an City as a case study, this paper proposes an integrated workflow using orthogonal experiments and successive approximation for calibrating Manning’s roughness coefficients in river networks. In this workflow, 13 river reaches (from six major rivers) serve as experimental factors. The Manning’s roughness coefficients for the main channel and floodplains are assigned different values as experimental levels. Model performance is evaluated using the Nash–Sutcliffe Efficiency (NSE) and Root Mean Square Error (RMSE). A multi-factor and multi-level orthogonal table L₂₇(3¹³) of main channel or floodplains roughness is alternately selected to design 27 sets of experiments. Through HEC-RAS simulation and orthogonal analysis, the roughness coefficients of the main channel and floodplains are alternately screened and successively approximated to the target values. Finally, the roughness coefficients of the main channel and floodplains for each river reach meeting the accuracy requirements are obtained, with corresponding values of NSE = 0.93 and RMSE = 0.04 m. The results show that orthogonal experimental design significantly reduces the number of simulation tests and effectively saves computational time and costs, while the successive approximation strategy addresses the complexity of solving problems with multiple decision variables. Additionally, the experimental factors consider variations in cross-section types and hydraulic conditions along the river by setting roughness coefficients in segments. The orthogonal experimental design ensures the relevance, simultaneity, and systematic nature of parameter adjustments across all river reaches, significantly enhancing the rationality and reliability of the model parameter calibration.

Seasonal groundwater (GW) pumping and climatic variability significantly impact the dynamics of greenhouse-dominated agricultural systems, yet quantitative evaluations at the local scale remain limited. This study explores non-parametric statistical and deep learning (DL) models for analyzing seasonal GW trends and predicting GW levels near greenhouse agriculture systems in Gyeongsangnam-do, South Korea. The modified Mann–Kendall (MK) test and Sen’s slope estimator were used to estimate long-term seasonal trends for the summer (wet season) and winter (dry season), based on monthly GW-level time series from six monitoring wells. Findings indicate that seasonal asymmetry is strong (winter trends have greater magnitudes and greater variability than summer trends), and that winter trends are negative (ranging from −0.45 to +1.70 m year⁻¹) and summer trends are positive (ranging from −0.02 to +0.31 m year⁻¹). At Jinju1 and Jinju4, statistically significant increasing trends were observed in both seasons (p < 0.05), but at other stations, weak or non-significant trends were observed due to short records or high variance. Long short-term memory (LSTM) and spatio-temporal graph neural network (STGNN) models were deployed and compared to predict at the GW level. The STGNN was found to be superior to LSTM in terms of R² (0.799–0.994) and reduced RMSE of up to 64.6, especially in winter, when spatially synchronized pumping is dominant in GW behavior. Despite advanced modeling, there is a serious concern about data limitations. Findings show that combining seasonal trend analysis with spatiotemporal modeling of DLs can significantly enhance knowledge and forecasting of GW dynamics in intensive greenhouse farming.

The Mahai Basin (MHB), situated in the northern Qaidam Basin on the Qinghai–Tibetan Plateau, hosts significant Quaternary potash resources. Nevertheless, the sources and evolutionary pathways of potash-forming fluids remain controversial. In this study, a comprehensive multi-isotope dataset and online-first publications spanning the period from 2015 to 2025 were compiled to constrain the end-member characteristics and evolution of brines in the MHB. δD-δ¹⁸O indicates that the initial fluids were derived mainly from Qilian Mountains precipitation and snowmelt, delivered via surface runoff and concentrated through prolonged evaporation under arid, semi-closed conditions, forming a river-lake-brine evolution sequence. δ⁷Li (+7‰ to +40‰) systematically increases with salinity and K content, reflecting long-term low-temperature water–rock interactions and selective ⁶Li adsorption by secondary clays, while deep Ca-Cl brines represent highly evolved endmembers. Elevated ⁸⁷Sr/⁸⁶Sr ratios (0.7113–0.7122) confirm silicate weathering contributions, with intercrystalline brines acting as key intermediate end members. B, S, and Cl isotopes further highlight deep fluid ascent along faults and anticlines, driving K co-enrichment, while sandy–gravel brines inherit highly evolved paleo-lake signatures. These multi-isotope constraints define an integrated evolutionary model involving surface runoff recharge, evaporation-driven concentration with water–rock interaction, deep fluid mixing, lateral migration, and final potash precipitation.