Evapotranspiration (ET) represents a significant component of the global water flux cycle, yet nocturnal evapotranspiration (ETn) is often neglected, leading to underestimation of global evapotranspiration. As for cropland, accurate modeling of ETn is essential for rational water management and is important for sustainable agriculture development. We used random forest (RF) to simulate ETn at 16 globally distributed cropland eddy covariance flux sites along with remote sensing and meteorological factors. The recursive feature elimination method was used to remove unimportant variables. We also simulated the ETn of C₃ and C₄ crops separately. The trained RF resulted in a determination coefficient (R²) (root mean square error (RMSE)) of 0.82 (7.30 W m⁻²) on the testing dataset. C₃ and C₄ crops on the testing dataset resulted in an R² (RMSE) of 0.86 (5.59 W m⁻²) and 0.55 (4.86 W m⁻²) for the two types of crops. We also showed that net radiation is the dominant factor in regulating ETn, followed by 2 m horizontal wind speed and vapor pressure deficit (VPD), and these three meteorological factors showed a significant positive correlation with ETn. This research demonstrates that RF can simulate ETn from crops economically and accurately, providing a methodological basis for improving global ETn simulations.

Accurately predicting actual evapotranspiration (${\mathrm{ET}}_{\mathrm{a}}$) at the regional scale is crucial for efficient water resource allocation and management. While previous studies mainly focused on predicting site-scale ${\mathrm{ET}}_{\mathrm{a}}$, in-depth studies on regional-scale ${\mathrm{ET}}_{\mathrm{a}}$ are relatively scarce. This study aims to address this issue by proposing a MulSA-ConvLSTM model, which combines the multi-headed self-attention module with the Pyramidally Attended Feature Extraction (PAFE) method. By extracting feature information and spatial dependencies in various dimensions and scales, the model utilizes remote sensing data from ERA5-Land and TerraClimate to attain regional-scale ${\mathrm{ET}}_{\mathrm{a}}$ prediction in Shandong, China. The MulSA-ConvLSTM model enhances the efficiency of capturing the trend of ${\mathrm{ET}}_{\mathrm{a}}$ successfully, and the prediction results are more accurate than those of the other contrast models. The Pearson’s correlation coefficient between observed and predicted values reaches 0.908. The study has demonstrated that MulSA-ConvLSTM yields superior performance in forecasting various ${\mathrm{ET}}_{\mathrm{a}}$ scenarios and is more responsive to climatic changes than other contrast models. By using a convolutional network feature extraction method, the PAFE method extracts global features via various convolutional kernels. The customized MulSAM module allows the model to concentrate on data from distinct subspaces, focusing on feature changes in multiple directions. The block-based training method is employed for the large-scale regional ${\mathrm{ET}}_{\mathrm{a}}$ prediction, proving to be effective in mitigating the constraints posed by limited hardware resources. This research provides a novel and effective method for accurately predicting regional-scale ${\mathrm{ET}}_{\mathrm{a}}$.

Timely and accurate prediction of crop traits is critical for precision breeding and regional agricultural production. Previous studies have primarily focused on single crop yield traits, neglecting other crop traits and variety-specific analyses. To address this issue, we employed a Meta-Hybrid Regression Ensemble (MHRE) approach by using multiple machine learning (ML) approaches as base learners, integrating regional multi-year, multi-variety crop field trials with satellite remote sensing indices, meteorological and phenological data to predict major crop traits. Results demonstrated MHRE’s optimal performance for rice and cotton, significantly outperforming individual models (RF, XGBoost, CatBoost, and LightGBM). Specifically, for rice crop, MHRE achieved highest accuracy for yield trait (R² = 0.78, RMSE = 0.59 t ha⁻¹) compared to the best individual model (XGBoost: R² = 0.76, RMSE = 0.61 t ha⁻¹); traits like effective spike also showed strong predictability (R² = 0.64, RMSE = 27.81 10,000·spike ha⁻¹). Similarly, for cotton, MHRE substantially improved yield trait prediction (R² = 0.82, RMSE = 0.33 t ha⁻¹) compared to the best individual model (RF: R² = 0.77, RMSE = 0.36 t ha⁻¹); bolls per plant accuracy was highest (R² = 0.93, RMSE = 2.27 bolls plant⁻¹). Moreover, rigorous validation confirmed that crop-specific MHRE models are robust across five rice and three cotton varietal groups and are applicable across six distinct regions in China. Furthermore, we applied the SHAP (SHapley Additive exPlanations) method to analyze the growth stages and key environmental factors affecting major traits. Our study illustrates a practical framework for regional-scale crop traits prediction by fusing multi-source data and ensemble machine learning, offering new insights for precision agriculture and crop management.

Ecosystem transpiration (T) is the core process in terrestrial water and carbon cycles. Accurately estimating T is critical to improving evapotranspiration (ET) models and understanding global ecosystem responses to climate change. In this study, we evaluated four ET partitioning methods (TEA, Z16, L19, and Y21) using 368 global eddy covariance (EC) sites and 15 sap flow sites. Intercomparison results showed that TEA, Z16, and Y21 maintained good consistency, whereas L19 exhibited lower agreement, primarily due to its high sensitivity to energy closure errors and poor non-linear fitting accuracy under extreme conditions. Validation against sap flow data indicated that Z16 performed best (R² = 0.45, KGE = 0.52), followed by Y21, while TEA had the lowest accuracy due to systematic overestimation driven by unremoved persistent background soil evaporation in its training dataset. Global analysis revealed that mean annual T ranged from 213 mm yr⁻¹ (Z16) to 294 mm yr⁻¹ (TEA), with annual T/ET varying between 0.45 (Z16) and 0.63 (TEA). Trend analysis further showed consistent increasing trends across all four methods for both annual T (0.33–0.83 mm·yr⁻²) and annual T/ET (0.0015–0.0019 yr⁻¹). Additionally, a notably stronger relationship was found between gross primary productivity (GPP) and T than between GPP and ET. Despite substantial differences in model structures, these methods effectively capture the temporal dynamics of T and the coupled relationships between ecosystem carbon and water fluxes. Our findings provide critical benchmarks for terrestrial water cycle modeling and sustainable water resource management under a changing climate.