With the constant advancement of irrigation technology and the continuous expansion of irrigation areas, non-point source pollution (NPS) caused by agricultural activities has posed a persistent threat to ecosystems and biological safety. Against this backdrop, it is imperative to lay scientific foundations for green, sustainable, and high-quality agricultural development through a thorough review of the relevant research progress. In this study, bibliometric methods are adopted to comprehensively analyze and visualize the current state and key literature on agricultural irrigation and NPS pollution from 2010 to July 2024. The focus of this study is specifically on summarizing the research hotspots and development trends of different irrigation methods and the mechanisms behind their impacts on NPS pollution. The results indicate that publications from the United States and China account for 63.8% of the total, but the fragmentation of research efforts remains, suggesting a necessity to strengthen international and regional collaboration. There are three institutions with the highest publication output, namely Northwest A&F University, Hohai University, and the Chinese Academy of Sciences. The subjects identified as the key areas of research on irrigation-related NPS pollution (IRR-NPS) include precision irrigation, rapid water pollution response, spatiotemporal management, interdisciplinary integration, wastewater treatment, and crop models. Regarding future research, it is necessary to focus attention on real-time precision irrigation, standardized crop models, data accuracy, spatiotemporal pollution coordination, pollution purification technology development, interdisciplinary integrated governance, and the innovative applications of soil improvement technologies. In addition to offering theoretical support and practical guidance for the management of agricultural NPS pollution, this study also provides management and technical support for policymakers, which is beneficial for advancing agricultural irrigation technology and environmental preservation.

To investigate the mechanism and influence of applying gamma-polyglutamic acid (γ-PGA) on soil water infiltration, laboratory experiments and numerical simulations were conducted using Hydrus-1D. These studies assessed the impact of various application rates of γ-PGA on soil water characteristic parameters. Orthogonal simulation experiments on soil bulk density, γ-PGA application rates, and burial depths were performed utilizing predefined soil water characteristic values (twelve groups: nine groups of numerical simulation experiments and three groups of laboratory verification tests), and the soil infiltration characteristics were analyzed. Concurrently, an empirical model was developed to elucidate the relationships between the empirical model parameters and influencing factors, as well as to examine the sensitivity of these factors to changes in soil infiltration rate. The relationship between cumulative infiltration and the distance of wetting front movement, based on the water balance equation, was refined. The results indicated that γ-PGA significantly affected soil water characteristic parameters, where the saturated water content and the reciprocal of soil intake suction increased with rising γ-PGA applications (p < 0.01), while the saturated hydraulic conductivity and the parameter n decreased (p < 0.01), with no notable changes in the retained water content (p > 0.05). The trend in cumulative infiltration influenced by various factors could be modeled by a capacitive charging model function, which yielded a superior fit. A negative correlation existed between the sensitivity index and all the influencing factors (p < 0.05), with the order of influence being soil bulk density, γ-PGA application rate, and γ-PGA burial depth, respectively. Utilizing the modified water balance equation, the ratio of cumulative infiltration to wetting front migration distance corresponded more closely with a power function. These findings provide a theoretical foundation for further studies on the effects of γ-PGA on crop growth characteristics in fields and the optimization of γ-PGA technical element combinations.

Accurate, high-resolution irrigation-related spatial information is paramount to diverse applications, including water resources management, food security, and agricultural planning. To address this need, our study leveraged machine learning algorithms and integrated multi-source data to extract and analyze land use types and spatiotemporal dynamics of irrigated farmland across provinces in the lower reaches of the Yellow River Basin over the 2008–2022 period. The results indicate that cultivated land remained dominant and largely stable, although localized losses occurred in peri-urban areas due to urban expansion. Construction land increased significantly, particularly in Shandong where it expanded by more than 15%, while forest and grassland areas grew under national ecological programs. The Random Forest (RF) algorithm achieved robust performance in identifying irrigated farmland, with overall accuracy exceeding 85% and regression with statistical irrigation data yielding R² values above 0.9 over the past 15 years at the city level. Spatiotemporal analysis showed strong variability in Henan, with irrigated area declining by 8–12% during drought years and recovering in wetter years, while Shandong experienced relative stability but a gradual 5% decline since 2015, driven by groundwater depletion and stricter regulation. The findings suggest irrigation expansion has reached near-saturation, given stable cultivated land and continuous improvements in water use efficiency. Future strategies should prioritize water use efficiency, water saving technologies, and equitable allocation to ensure sustainable agricultural development.

Leaf nitrogen content (LNC) is an important physiological indicator for evaluating citrus nutritional status, photosynthetic capacity, and fertilization demand. However, conventional LNC determination mainly relies on field sampling and laboratory chemical analysis, which are destructive, time-consuming, labor-intensive, and limited in spatial continuity, making them unsuitable for large-scale real-time nitrogen monitoring in complex orchard environments. To achieve rapid and non-destructive estimation of citrus LNC, this study developed a UAV multispectral inversion framework integrating object-based canopy extraction and machine learning models. Field experiments were conducted in a citrus orchard in western Hubei Province, China. Multi-temporal UAV multispectral images were collected from April to October 2025, and ground measurements of citrus LNC were collected simultaneously. First, minimum distance classification (MDC), maximum likelihood classification (MLC), and object-based image analysis (OBIA) were used for land-cover classification of citrus orchard images, and their canopy extraction performance under complex orchard backgrounds was compared. Subsequently, multiple vegetation indices were calculated from the extracted citrus canopy spectra, and sensitive spectral features were selected through correlation analysis. Finally, seven models, including simple linear regression, quadratic regression, partial least squares regression (PLS), back propagation neural network (BP), extreme learning machine (ELM), particle swarm optimization-extreme learning machine (PSO-ELM), and particle swarm optimization-back propagation neural network (PSO-BP), were constructed to systematically evaluate the inversion performance of citrus LNC across the entire growth period. The results showed that: (1) OBIA achieved higher classification accuracy and temporal stability in citrus orchard land-cover classification, with overall accuracy ranging from 68.86% to 85.65% and Kappa coefficients ranging from 0.56 to 0.72, outperforming MDC and MLC. This indicates that OBIA can effectively reduce the interference of bare soil, grass, shadows, and other non-target objects on canopy spectral extraction. (2) The correlations between vegetation indices and LNC varied markedly among different growth stages, suggesting that the spectral response of citrus LNC has strong phenological dependence and that a single vegetation index is insufficient to stably characterize LNC variation across the whole growth period. (3) At the whole-growth-period scale, multi-index fusion models outperformed single-index models, among which EVI, TVI, and MTVI showed relatively strong cross-stage sensitivity. (4) Optimized machine learning models generally outperformed traditional regression models and unoptimized machine learning models. Among them, PSO-BP achieved the best performance, with a validation R² of 0.68 and an RMSE of 1.54 g kg⁻¹, representing an increase in R² of 23.64% compared with the PLS model and 25.93% over the baseline BP model in terms of R². Overall, this study demonstrates that OBIA-based canopy spectral quality improvement combined with PSO-optimized machine learning can effectively improve the stability and reliability of UAV multispectral estimation of citrus LNC under complex orchard backgrounds. The proposed framework provides technical support for citrus nitrogen diagnosis, precision fertilization, and intelligent orchard management.