Gross primary productivity (GPP) quantifies the rate at which plants convert atmospheric carbon dioxide into organic matter through photosynthesis, playing a vital role in the terrestrial carbon cycle. Machine learning (ML) techniques excel in handling spatiotemporally complex data, facilitating accurate spatial-scale inversion of forest GPP by integrating limited ground flux measurements with Remote Sensing (RS) observations. Enhancing ML algorithm performance for precise GPP estimation is a key research focus. This study introduces the Random Grid Search Algorithm (RGSA) for hyperparameters tuning to improve Random Forest (RF) and eXtreme Gradient Boosting (XGB) models across four major forest regions in China. Model optimization progressed through three stages: the Unoptimized (UO) XGB model achieved $R^2$ = 0.77 and $RMSE$ = 1.42 g Cm⁻² d⁻¹; the Hyperparameter Optimized (HO) XGB model using RGSA improved performance by 5.19% in $R^2$ (0.81) and reduced $RMSE$ by 9.15% (1.29 g Cm⁻² d⁻¹); the Hyperparameter and Variable Combination Optimized (HVCO) XGB model with selected variables (LAI, Temp, NR, VPD, and NDVI) further enhanced $R^2$ to 0.83 and decreased $RMSE$ to 1.23 g Cm⁻² d⁻¹. The optimized GPP estimates exhibited high spatial consistency with existing high-quality products like GOSIF GPP, GLASS GPP, and FLUXCOM GPP, validating the model’s reliability and effectiveness. This research provides crucial insights for improving GPP estimation accuracy and optimizing ML methodologies for forest ecosystems in China.

Subsurface drip irrigation is an advanced technique that significantly enhances agricultural water efficiency and conserves irrigation resources. The Heilonggang region is highly representative of the maize–wheat rotation system in China. This region was selected for conducting operations and maintenance experiments on subsurface drip irrigation systems. The primary objective of this study was to determine the most suitable type of drip tape for application in the North China Plain and to identify specific maintenance measures necessary to ensure the long-term functionality of subsurface drip irrigation systems. The experiment was conducted in Jing County, Hengshui City. Anti-blocking drip tape (Netafim Co., Ltd. Beijing, China) with automatic functionality was evenly laid in the test area. The experimental area was divided into six rotational irrigation groups. The key parameter examined in this single-factor experiment was the drip tape wall thickness, with values of 0.2, 0.225, 0.25, 0.28, 0.31, and 0.38 mm. Drip tape treatments were randomly sampled in rotational irrigation groups, and there were three repeat plots in each treatment. Each replicate plot contained ten drip irrigation belts spaced 60 cm apart, with an interval of more than 2 m between adjacent plots. The subsurface drip irrigation system was installed in October 2023. Prior to irrigation, drip tape troubleshooting was conducted and recorded manually on 12 March 2024 (wheat regreening stage) and 29 June 2024 (maize-sowing stage). The experimental findings indicated that the primary factor influencing the stability of the irrigation system was the wall thickness of the drip tapes, while other system components operated efficiently. A significant correlation was observed between the wall thickness of the drip tape and the number of water leakage points (p < 0.05), with an absolute correlation coefficient exceeding 0.9. The number of leakage points in drip tapes with wall thicknesses of 0.2–0.28 mm (267 instances) was significantly higher than those with wall thicknesses of 0.31–0.38 mm (29 instances), primarily due to damage caused by mole crickets and wireworms. Following the injection of 40% phoxim, 2.5% lambda-cyhalothrin, and 70% imidacloprid insecticides (at a cost of 16.7 USD·ha⁻¹) into the subsurface drip irrigation system, the insect pests were nearly eradicated within one month. A cumulative cost evaluation over a 6–10-year period recommended the use of drip tapes with a wall thickness of 0.31 mm and the application of insecticides every 1–2 months to maintain optimal system performance in this region. These measures can effectively support the stable operation of this irrigation technique at a relatively low cost.

Nowadays, because of the severe contradiction between water supply and demand, a large amount of ecological water resources are occupied by other water-using sectors, resulting in the rapid degradation of the ecological environment, especially in arid and semi-arid areas of northwestern China. Therefore, in order to deal with the above problems, a multi-objective water resources optimization allocation model based on ecological water demand is established, which not only focuses on ecological water demand, but also can effectively deal with the conflict among multiple objectives. A case study was applied in Wuwei City, Gansu Province, China, which had maximum economic benefit and ecological benefit as objectives. A series of optimal water resources distribution schemes were obtained under different representative hydrological years. From the result, as representative hydrological years changed from wet to dry, economic benefit and ecological water deficit would vary from CNY [52.82, 36.32] × 10⁸ and [2.69, 5.51] × 10⁸ m³, respectively. It indicated that water resources have become one of the factors restricting the sustainable development of Wuwei City. Even when p = 25%, it still cannot meet the water demand. This indicates that Wuwei city needs to aggressively develop water-saving measures and new water resources in the future to deal with the acute water scarcity scenario. In addition, no matter what representative hydrological years are used, the results of the established multi-objective programming model are always in the middle of the results of the individual objective, indicating that the established multi-objective programming model has the advantage of dealing with water competing conflict among different objectives.