This paper evaluates linear regression analysis, BP neural network, and a random forest prediction model for the prediction of jujube water demand. The results highlight that the R2 of the random forest is 0.941 and the residual distribution is the most stable. Hence, the random forest is more suitable for prediction, and therefore, an intelligent irrigation system is established employing random forest, where the cloud server is the upper computer and a Raspberry Pi is the lower computer, and at the same time, a PC and a mobile interface was built to present various information about the developed irrigation system.

Organic fertilizers have been widely applied in Chinese agriculture. Soil aggregate composition and pore size distribution indicate the effects of fertilizer practices on soil quality. While the effects of the long-term application of organic fertilizers on total soil pore space have been reported, few studies have examined the distribution of connected pore spaces (of critical importance for solute transport in the soil). Soil physical and chemical properties influence plant growth and soil processes. We compared the effects of organic fertilizer application (10 years, 5 years, and no application) on soil connectivity pore structures and physicochemical properties within the root zone (20–40 cm depth) around date palms through CT scanning or core samples. Specifically, when the number of years of organic fertilizer application increased, the proportion of 2–0.25 mm particle size aggregates in the soil increased, and soil connectivity pore structure became more complex and stable, with higher connectivity, pore numbers, and numbers of pore throats. The connectivity of particles of the 0–300 μm size class of pores increased and the proportion of the 500–1000 μm class of pore throats increased. Soil organic matter and enzyme activity were significantly related to soil pore structure characteristics. In conclusion, the application of organic fertilizer improved soil pore structures, and the effects became more pronounced with the increased duration of application. These results provide theoretical and practical support for the application of organic fertilizer to improve soil structures in arid areas, and the findings have significant implications for sustainable agriculture in arid and semi-arid regions.

To address pressure-drop-induced safety risks in high-drop gravity-fed irrigation pipelines, this study investigates coordinated prevention and control strategies that integrate air release and vacuum valve groups with flow-adaptive valve closure rules. A large-scale self-pressurized irrigation network (1.33 × 10⁸ m²) in Karamay, Xinjiang, China, is selected as a representative case study. Based on one-dimensional transient flow modeling, pressure drop and negative-pressure characteristics induced by inlet valve closure in the main pipeline are analyzed using wave speed theory, governing differential equations, and the finite difference method. A coordinated protection framework is proposed that explicitly links valve operating patterns with the spatial configuration of protective devices. Unlike conventional schemes that rely on empirical layouts and fixed closure rules, this study introduces a critical-flow-velocity-based valve grouping method combined with flow-dependent valve closure strategies. Simulation results demonstrate that a strategically optimized configuration of air release and vacuum valves along the main pipeline is sufficient to eliminate negative pressure under all operating conditions. For flow rates below 6 m³/s, linear valve closure ensures safe operation, whereas a two-stage closure is required for higher flow rates (6–10 m³/s). As flow increases, reducing the fast-closure ratio and extending the total closure time effectively suppress pressure-drop-dominated transient effects at vulnerable inlet sections. By effectively mitigating transient pressure surges, the proposed coordinated “valve closure-protection device” strategy improves system adaptability to flow variability and provides practical engineering guidance for the safe operation of gravity irrigation systems, particularly high-gradient self-pressurized networks.

This study investigates the impact of Aerated Subsurface Drip Irrigation (ASDI) on the growth and yield of mulched cotton, aiming to identify the optimal water-air combination pattern for ASDI in cotton cultivation. Conducted during 2021–2022, the experimental setup involved two aeration modes (aerated A₁ and unaerated A₀) and four irrigation quotas (W₁, W₂, W₃, and W₄), organized in a two-factor randomized block design resulting in eight distinct treatments. The findings revealed that ASDI significantly promoted soil moisture depletion from 0 to 40 cm during the cotton flowering and boll opening stages. Specifically, aerated A₁ reduced soil water content by 5.84% to 7.83% during the flowering stage and 7.45% to 13.39% during the boll opening stage compared to unaerated A₀. Additionally, both aerating and increasing irrigation quotas not only enhanced the cotton leaf area index (LAI) but also delayed leaf area decay, contributing to prolonged photosynthetic activity. Aerating also favorably influenced the distribution of above-ground biomass in cotton towards budding and boll stages, with the biomass share of buddings, flowers, and bolls averaging 62.98% under aerated conditions versus 62.27% under non-aerated conditions during the boll opening stage. Furthermore, aerating combined with increased irrigation quotas resulted in higher seed cotton yields, with aerated irrigation boosting yields by 1.79% in 2021 and 4.43% in 2022 compared to non-aerated irrigation. This approach also increased cotton’s water demand and average daily water consumption significantly (p < 0.01). Importantly, aerating improved IWUE, achieving 1.72 kg/m³ in 2021 and 1.62 kg/m³ in 2022 for ASDI, versus 1.69 kg/m³ and 1.57 kg/m³ for unaerated subsurface drip irrigation, respectively. In conclusion, from a water conservation and yield enhancement perspective, an irrigation quota of 337.4 mm during the reproductive stage under ASDI is recommended as an effective strategy for “one film three tubes and six rows” mulched cotton in Southern Xinjiang.

Leaf nitrogen concentration (LNC) exhibits pronounced vertical heterogeneity across canopy layers, which affects the accuracy of nitrogen diagnosis derived from UAV-based remote sensing imagery. To address the differential contributions of leaf nitrogen from distinct canopy strata and the limitations associated with single-source features, this study proposes an integrated framework that combines cumulative LNC indicators across canopy layers with multi-source feature sets (vegetation indices and texture features). Centered on three core technical innovations—(1) incorporating canopy-layer aggregation logic into LNC modeling, (2) integrating spectral and structural information through CNN-based feature fusion, and (3) combining deep feature extraction with gradient boosting regression to improve robustness under multi-stage conditions—the framework systematically evaluates three machine learning algorithms: Random Forest (RF), a Convolutional Neural Network–Extreme Gradient Boosting hybrid model (CNN_XGBoost), and K-Nearest Neighbor (KNN) for cotton LNC estimation across multiple growth stages. The results demonstrate that cumulative canopy-layer nitrogen indicators more effectively represent overall plant nitrogen status than single-layer measurements. The integration of multi-source features further enhances model performance. Under both single-variable inputs and combined VI–TF feature sets, the CNN_XGBoost model consistently outperforms the other models in calibration accuracy and stability across all growth stages. Its optimal performance occurs during the cotton flowering and boll stage, achieving a calibration R² of 0.921. Overall, the proposed framework substantially improves the estimation accuracy of cotton LNC and provides both a theoretical foundation and technical support for precision nitrogen management and sustainable agricultural development.