Biao Long

Accurate prediction of lithium-ion battery lifespan is vital for ensuring operational reliability and reducing maintenance costs in applications like electric vehicles and smart grids. This study introduces a novel hybrid learning framework that addresses the limitations of existing methods by integrating multisource data fusion with a stacked ensemble (SE) modeling approach, specifically designed to handle the heterogeneity and variability in battery data from diverse sources. By leveraging heterogeneous data sets from the National Aeronautics and Space Administration (NASA), Center for Advanced Life Cycle Engineering (CALCE), MIT-Stanford-Toyota Research Institute (TRC), and nickel cobalt aluminum (NCA) chemistries, a variance-aware weighting mechanism mitigates variability across data sets. The SE model combines ridge regression, long short-term memory (LSTM) networks, and eXtreme Gradient Boosting (XGBoost), effectively capturing temporal dependencies and nonlinear degradation patterns. The proposed framework not only achieves superior predictive performance, with a mean absolute error (MAE) of 0.0058, root-mean-square error (RMSE) of 0.0092, and coefficient of determination (R²) of 0.9839, but also demonstrates a significant improvement over baseline models, including a 46.2% increase in coefficient of determination and an 83.2% reduction in root-mean-square error. Shapley additive explanations (SHAP) analysis identifies differential discharge capacity (Qdlin) and temperature of measurement (Temp_m) as critical aging indicators. This scalable, interpretable framework enhances battery health management, supporting optimized maintenance and safety across diverse energy storage systems.

Spatially synchronized extreme precipitation events are intensifying under anthropogenic warming. Accurate simulation of such compound extremes by global climate models underpins reliable climate projections for spatially compound risk assessment. Using complex network analysis combined with event synchronization, here we evaluate the performance of 11 Coupled Model Intercomparison Project Phase 6 (CMIP6) models in representing global synchronized structures of extreme precipitation during 1981–2014. Compared to state‐of‐the‐art reanalysis data, CMIP6 models effectively capture the scale‐dependent behavior of synchronized event pairs, particularly for teleconnections beyond 2,500 km. While the CMIP6 multi‐model ensemble mean overestimates short‐range (300–2,500 km) synchronization frequency by 5.7%, it well reproduces the overall network topology of global extreme precipitation. However, significant regional biases emerge in monsoon regions, where models systematically underestimate node connectivity by more than 20% during boreal summer, highlighting key areas for model improvement in simulating long‐distance synchronized precipitation events.

This study investigates anomalous precipitation patterns in the Taihu Basin, located in the Yangtze River Delta of eastern China, using high-resolution daily data from 1960 to 2019. Leveraging a deep learning autoencoder and self-organizing map, three spatially distinct types are identified—north type (72%), south type (19.7%), and center type (8.3%). The north type exhibits a pronounced upward trend (+0.11 days/year, p < 0.05), indicating intensifying extreme rainfall under climate warming, while the south type displays a bimodal temporal structure, peaking in early summer and autumn. Composite analyses reveal that these patterns are closely associated with the westward extension of the Western North Pacific Subtropical High (WNPSH), meridional shifts of the East Asian Westerly Jet (EAJ), low-level moisture convergence, and SST–OLR anomalies. For instance, north-type events often coincide with strong anticyclonic anomalies and enhanced moisture transport from the Northwest Pacific and South China Sea, forming favorable convergence zones over the basin. For flood management in the Taihu Basin, the identified spatial patterns, particularly the bimodal south type, have clear implications. Their strong link to specific circulation features enables certain flood-prone scenarios to be anticipated 1–2 seasons in advance, supporting proactive measures such as reservoir scheduling. Overall, this classification framework deepens the understanding of atmospheric patterns associated with flood risk and provides practical guidance for storm design and adaptive flood risk management under a changing climate.

Accurate prediction of lithium-ion battery lifespan is vital for ensuring operational reliability and reducing maintenance costs in applications like electric vehicles and smart grids. This study presents a hybrid learning framework for precise battery lifespan prediction, integrating dynamic multi-source data fusion with a stacked ensemble (SE) modeling approach. By leveraging heterogeneous datasets from the National Aeronautics and Space Administration (NASA), Center for Advanced Life Cycle Engineering (CALCE), MIT-Stanford-Toyota Research Institute (TRC), and nickel cobalt aluminum (NCA) chemistries, an entropy-based dynamic weighting mechanism mitigates variability across heterogeneous datasets. The SE model combines Ridge regression, long short-term memory (LSTM) networks, and eXtreme Gradient Boosting (XGBoost), effectively capturing temporal dependencies and nonlinear degradation patterns. It achieves a mean absolute error (MAE) of 0.0058, root mean square error (RMSE) of 0.0092, and coefficient of determination (R2) of 0.9839, outperforming established baseline models with a 46.2% improvement in R2 and an 83.2% reduction in RMSE. Shapley additive explanations (SHAP) analysis identifies differential discharge capacity (Qdlin) and temperature of measurement (Temp_m) as critical aging indicators. This scalable, interpretable framework enhances battery health management, supporting optimized maintenance and safety across diverse energy storage systems, thereby contributing to improved battery health management in energy storage systems.

Upstream inland rivers provide most of water resources for midstream and downstream in arid and semiarid regions. From the perspective of ecological rehabilitation, it is necessary to simultaneously ensure upstream streamflow while still maintaining economy‐ecosystem balance in upstream. However, vegetation dynamics effects caused by ecological rehabilitation cannot be well evaluated in previous studies due to the sparseness of observations at regional scales. With increased availability of vegetation and hydrological data, it is now possible to explore the relationship between water yield and ecological balance affected by vegetation dynamics. This study takes Upper Hei River Basin, part of the Asia's water tower—the Qinghai‐Tibet Plateau, as an example and uses the Eagleson's ecohydrological model to explore the effects of vegetation dynamics on streamflow and ecology in growing seasons during 1992–2015. We compare scenarios with and without vegetation changes and find that there is limited evidence that ecological rehabilitation could enhance water yield. Approximately 51.31% of Upper Hei River Basin vegetation cover (M) tends to deviate from ecological optimization state compared to equilibrium vegetation cover (M_eq). In order to explore ways to ensure streamflow and maintain ecology, different vegetation scenarios are simulated. Results show that on average, grazing in area where M is larger than M_eq could increase streamflow by approximately 7.3%/year and planting trees in other area could increase streamflow by approximately 2.9%/year. Therefore, effective control on vegetation dynamics can simultaneously ensure streamflow and maintain ecological balance, which can also provide guidance and theoretical support for upstream inland river planning.

The Inland River Area of China (IRAC), being the most water‐stressed region in the world, faced with increasingly severe ecological degradation risk and water shortage issue under global climate warming. Previous studies indicated that the climate of IRAC have changed from warm‐dry to warm‐wet, because both precipitation (P) and air temperature have increased. However, the analysis in this study, considering the effects of climate change on hydrologic cycle rather than simply on climate conditions, indicates that there has been no substantial change from a warm‐dry to a warm‐wet climate in the IRAC during 1948–2010. The Standardized Moisture Anomaly Index (SZI), which is a comprehensive multiscalar drought indicator, was used to identify the patterns of dry and wet weather. The Global Land Data Assimilation System 2 (GLDAS2) Noah model, based on advanced land surface modeling and data assimilation techniques, were used to provide input surface water‐energy variables for calculating SZI. The results show that both the forcing data and the output of the GLDAS2 Noah model are in good agreement with different types of observations. The variabilities in SZI are more consistent with evidence of multiple droughts (including residual water/energy ratio, evapotranspiration deficit, and historical drought records) than the Standardized Precipitation Index (SPI) and the Standardized Precipitation Evapotranspiration Index (SPEI). Based on SZI, it is known that increases in the water supply (P) are balanced by the increased water demand caused by significant climate warming; that is, terrestrial water storage has generally remained balanced over the past four decades.

The lack of reliable estimation of water demand in drought study has been an important obstacle in efforts for characterizing the variability of water consumption and its effects on drought monitoring and prediction, particularly over water‐stressed regions. This study evaluated the performance and the contribution of three modified water demand estimates in drought modeling, including Penman–Monteith (PM) method, dual‐source potential evapotranspiration model (2S PET model), and climatically appropriate for existing conditions precipitation ( ). The results show that Standardized Moisture Anomaly Index based on shows the highest correlations with different types of drought evidence, demonstrating that the as a water demand metric performs the best in drought modeling over water‐stressed regions. Standardized Precipitation Evapotranspiration Index based on 2S PET model performs better than that based on PM equation. Although the 2S PET model is physically superior to PM equation over agricultural or water‐stress regions where the plant canopy is not “closed,” the limitations of using PET as water demand estimates in drought study still exist. Theoretically speaking, the actual water demand in non‐humid regions should be highly associated with the climatically averaged water supply capability (precipitation, P), and the contributions of water demand and supply to drought indices are generally balanced in any region, which is realistically reflected by the relationships between P and . On the contrary, the magnitude and contribution of PET_2S and PET_PM were always higher than P over water‐stressed regions, which could result in a water imbalance and generate more systemic errors in drought identification. Copyright © 2016 John Wiley & Sons, Ltd.