This paper describes the process of formulating product standards for box-type ultrasonic open-channel flowmeters, including the establishment of the multi-channel flow measurement in the measuring tank and the determination and validation of its performance indicators. It emphasizes that the performance indicators specified in the standards must be quantifiable, measurable, and verifiable. The paper also proposes that industrially produced measuring tanks should undergo verification, engineered flow measurement devices should be calibrated, and test flumes with adjustable widths should be designed and constructed for testing and verification purposes.

In response to the problems of low efficiency, uneven mixing, and high cost of traditional mechanical fertilizer mixing devices, this study designed a pneumatic water fertilizer mixing device aimed at optimizing fertilizer mixing performance, reducing costs, and verifying its superiority. A test system was constructed to analyze the influence of key parameters—such as the aeration hole offset angle, aeration tube layout, hole aperture and number, and tank volume—on fertilizer dissolution and mixing times. The device′s performance, cost, and energy consumption were then compared with those of conventional mechanical mixing devices. The results showed that optimal performance was achieved with an aeration hole offset angle of -45 °, a central aeration tube layout, 68 holes with a 4 mm aperture, and a 300 L tank, reducing the dissolution time and mixing time to 62 s and 78 s, respectively. Compared with traditional mechanical fertilizer mixing devices, pneumatic devices significantly improve dissolution time (shortened by 15.4%~25.9%) and mixing time (shortened by 15.4%~25.9%). Furthermore, its cost and energy consumption were lower in multi-unit configurations. Pneumatic mixing technology can effectively address the issues of inefficiency and high operational costs associated with traditional fertilizer mixing devices, providing an efficient and economical new solution for fertigation systems and demonstrating significant potential for practical application.

To elucidate the evolution mechanism of the non-uniform temperature field in a novel U-shaped stacked-box aqueduct and clarify the internal forces and stress characteristics of the continuous rigid frame system under temperature gradients, this study addresses the lack of explicit guidelines in current hydraulic engineering codes regarding temperature gradient patterns for aqueduct structures. First, domestic and international calculation methods for temperature gradient patterns and their effects on bridge structures were reviewed; second, taking U-shaped continuous rigid frame aqueduct as a case study, fluid-structure coupling transient thermal analysis was conducted to obtain the temperature distribution patterns across the aqueduct section, and the transverse and vertical temperature gradient patterns were derived through curve fitting and applied to the finite element model (FEM) of the continuous rigid frame aqueduct via nodal temperature loading. Finally, the results were co Mpared with temperature effects predicted by domestic and international bridge design codes. The findings reveal that during operation, temperature differences between the top and bottom slabs and web plates exceed 10 °C, with maximum positive and negative temperature differences of 24 °C and -10 °C, respectively. The derived temperature gradient patterns reflect the nonlinear temperature gradients induced by thermal differences. Under positive temperature gradients, the box structure exhibits tensile stresses internally and compressive stresses externally, with a maximum principal tensile stress of 2.59 MPa on the inner surface of the mid-span bottom slab. Conversely, under negative temperature gradients, the stress distribution reverses, with a peak principal tensile stress of 2.89 MPa on the outer surface of the mid-span top slab. Among evaluated codes, the New Zealand code yields the highest principal tensile (6.1 MPa) and compressive (12.0 MPa) stresses. The principal tensile stress from the aqueduct’s derived temperature pattern was the second highest, peaking at 4.3 MPa on the inner surface of the mid-span top slab. Under negative gradients, the maximum principal tensile stress (3.7 MPa) in the temperature-derived pattern occurs on the mid-span bottom slab outer surface, smaller than Chinese Highway Code and New Zealand code predictions. Principal compressive stresses from the temperature-derived pattern reach 3.0 MPa at the mid-span and near edge supports, closely matching the Chinese Highway Code elsewhere.

Efficient water-saving irrigation requires the use of a large number of water pump units, and their safe and stable operation plays an important role in ensuring the efficiency of the irrigation system. The safe operation and maintenance of water pump units mainly include regular maintenance, corrective maintenance, and predictive maintenance. Predictive maintenance not only requires real-time monitoring of key data, but also appropriate mathematical models, and is significant for improving the operational efficiency of pump stations and reducing potential operational risks. Based on the high correlation between bearing status and measurable physical signals, this paper analyzes the current state of bearing status monitoring in water pump units is analyzed, and an implementation scheme for intelligent bearing status monitoring based on a multi-scale time-frequency analysis method. By monitoring of bearing vibration, structural noise, and bearing temperature in real time, a mathematical model of bearing′s status is constructed to comprehensively analyze the real-time operating status and historical monitoring data of the water pump unit. This model enables bearing status fault warnings and lifetime assessment, provides a basis for water pump unit fault diagnosis and pump station energy consumption optimization, and effectively improves the operational safety and reliability of the pump units.

In the current trend of integrating hydrological monitoring and automation technologies, while traditional methods such as array-based radar flow meters have been widely used, their limitations can no longer meet the growing demand for accuracy and intelligence. To overcome the shortcomings of these traditional techniques, this paper presents an innovative dual-track automatic flow measuring device that integrates the Internet of Things (IoT) with automation technology. The device employs a unique dual-track operating system to perform high-precision flow measurement tasks above open channels while comprehensively monitoring key parameters like velocity and discharge. This dual-track measurement device not only significantly enhances accuracy and efficiency in flow measurement but also offers advanced functionalities including autonomous navigation, intelligent obstacle avoidance, and remote data transmission. This paper delves into the system’s design principles, key technological breakthroughs, and its practical value in watershed monitoring, aiming to provide a novel and efficient technical solution for water resources management and to further promote technological progress and development within this field.

To accurately simulate the growth status of rice in the Sanjiang Plain, this study used the random forest algorithm, data from 344 experimental fields in Shuangyashan and Jiamusi cities of Heilongjiang Province from 2021 and 2022, and eight vegetation indices to simulate key growth variables of rice, such as the jointing stage, flowering stage, number of tillering stems, leaf age, and yield. Additionally, the feature importance of the modeling data was analyzed using the feature attribution method. The results showed the following: ① The model achieved high simulation accuracy for the jointing stage, flowering stage, number of tillering stems, and leaf age, with the coefficient of determination (R2) ranging from 0.85 to 0.91, the Nash-Sutcliffe efficiency coefficient (NSE) from 0.65 to 0.91, and the relative root mean square error (RRMSE) from 7.22% to 9.89%. Although the yield simulation accuracy was slightly lower (R2=0.49, NSE=0.45, RRMSE=3.74%), it was still within an acceptable range. ② The results of the feature importance analysis indicated significant differences in the key influencing factors among different growth variables. The flowering stage was most affected by transplanting time. The Soil-Adjusted Vegetation Index (SAVI), Keetch-Byram Drought Index (KBDI), and Normalized Difference Vegetation Index (NDVI) were the three key vegetation indices, with importance values of 0.42, 0.37, and 0.32, respectively. The dominant factors for the jointing stage were the Ratio Vegetation Index (RVI), SAVI, and transplanting time (all with importance values exceeding 0.9), while NDVI, Normalized Difference Red-Blue Vegetation Index (NDRBI), and KBDI also showed high importance (0.72, 0.63, and 0.52, respectively). The number of tillering stems and leaf age shared the same key factors: the Difference Vegetation Index (DVI), Transformed Vegetation Index (TVI), and SAVI, all with importance values exceeding 0.65. For yield prediction, the main influencing factors were NDRBI (1.0), transplanting time (0.83), TVI (0.65), and NDVI (0.63). This study confirms that the random forest model can serve as an effective tool for simulating rice growth and analyzing data feature importance in the Sanjiang Plain, providing a scientific basis for precise water and fertilizer management in paddy fields and crop yield improvement.

Optimizing irrigation management is crucial for promoting water-efficient agriculture and ensuring efficient and high-quality production of facility crops. This study aims to evaluate the applicability of the WOFOST model for simulating greenhouse tomato growth under different irrigation regimes, providing technical support for precise irrigation management and growth simulation of greenhouse tomatoes. Three irrigation treatments were designed: full irrigation (W1, with a lower irrigation limit set at 70% of the field capacity, Fc), mild deficit irrigation (W2, with a lower limit at 65% of Fc), and severe deficit irrigation (W3, with a lower limit at 55% of Fc). Using measured data, model parameter optimization was conducted under different irrigation modes by combining the NLOPT optimization algorithm with the Logistic function, and sensitivity analysis of the optimized parameters was performed using the EFAST method. The results indicated that the localized parameter sets for W1 and W2 showed superior fitting performance for the above-ground dry matter accumulation (TAGP) under full and mild deficit irrigation, with R2 values ranging from 0.90 to 0.91 and NRMSE values between 0.11 and 0.12. However, the simulation of TAGP under severe deficit irrigation was less satisfactory (R2 < 0.8, NRMSE > 0.2). For the simulation of fruit dry matter weight (TWSO), the localized parameter sets for W1 and W2 demonstrated high accuracy across all three water conditions (R2 > 0.8, NRMSE < 0.2), while the parameter set for W3 performed well only for TWSO under severe deficit irrigation (R2 = 0.95, NRMSE = 0.08). In the simulation of leaf area index (LAI), the localized parameter sets for W1 and W2 were suitable for their respective scenarios (0.83 < R2 < 0.93; 0.09 < NRMSE < 0.16), whereas the parameter set for W3 was only effective for its own LAI simulation (R2 = 0.85, NRMSE = 0.15). Overall, the localized parameter sets under different water treatments can simulate the growth process of greenhouse tomatoes according to varying irrigation needs, providing technical support for the development of smart agriculture.

Reference crop evapotranspiration (ET ₀) is an important parameter for agricultural meteorological drought diagnosis and irrigation forecasting. This study, based on observational data from three meteorological stations in southern China, forecasted ET ₀ over a 1~10 day period using meteorological factors from the GraphCast artificial intelligence weather model. The initial forecasts were then corrected using the XGBoost algorithm. The importance of the meteorological factors forecasted by the GraphCast model was also evaluated. The results indicate that the GraphCast model has high forecasting accuracy for maximum and minimum temperatures, with its performance for maximum temperature forecasts being superior to that for minimum temperature. The model's ability to forecast relative humidity is acceptable, while its forecasting capability for wind speed is relatively poor. The mean absolute error (MAE) of ET ₀ forecasted by the GraphCast model at the three stations ranged from 0.53 to 1.02 mm/d. After correction using the XGBoost algorithm, the MAE decreased to 0.44 to 0.75 mm/d, with a significant improvement in bias as well. The lack of solar radiation forecasting capability is the primary factor limiting the further improvement of the model's accuracy. The secondary factors are relative humidity and maximum temperature. Although the forecasting ability for wind speed is poor, its impact on ET ₀ forecasting is minimal.

To explore the differences in agricultural drought characteristics across different soil layers and the impact of various factors on drought in the Huang-Huai-Hai Plain, this study used the Agricultural Drought Severity Index (DSI) to analyze the spatiotemporal variation of drought in shallow (0~28 cm), middle (28~100 cm), and deep (100~289 cm) soil layers in the Huang-Huai-Hai Plain from 1959 to 2023. A random forest model was applied to assess the importance of meteorological factors (precipitation, temperature, saturated vapor pressure deficit), potential evapotranspiration, and groundwater on agricultural drought. The results showed that: ①The interannual and seasonal variation trends of DSI in different soil layers differed significantly. The DSI of shallow soil fluctuated the most, with a significant decline from 1985 to 1999. The DSI of middle soil changed relatively steadily, with a significant decline in drought se during periods such as 1960 and 1962-1966. The DSI of deep soil shows a significant upward trend, with 2004 being a turning point. Seasonal fluctuations are most pronounced in shallow soil, followed by middle soil, and least in deep soil. ② The long-term average spatial distribution of drought showed a “heavier in the north and lighter in the south, decreasing with depth” pattern. In shallow soil, the area of moderate drought accounted for 33.2%, and severe and extreme droughts accounted for 2.61% and 1.7%, respectively; in middle soil, moderate drought accounted for 13.59%; in deep soil, moderate drought accounted for 7.58%. The frequency of drought decreased significantly with soil depth. Shallow soil experienced severe drought for an average of 1.16 months per year, while deep soil experienced it for only 0.37 months. ③ Temperature was the dominant factor for all layers. The importance of current-month temperature was 18.91%, 51.91%, and 52.01% for shallow, middle, and deep soils, respectively. The importance of the previous month’s temperature was also significant for middle and deep soils, reaching 40.59% and 38.38% respectively, indicating a thermal lag effect on drought. The importance of current-month precipitation and potential evapotranspiration in shallow soil was 15.05% and 11.92%, respectively, decreasing to less than 10% with increasing depth. The relative impact of groundwater increased in the deep soil, reaching 3.46%.

Agricultural drought, as a key factor directly affecting crop growth, relies on accurate soil moisture data for effective monitoring and assessment. Aiming to address the limitations of existing soil moisture products in terms of spatial resolution and estimation accuracy, this study selected the Huaibei Plain as the study area. Based on daily observations from 49 in-situ stations during 2015-2019, we systematically evaluated four mainstream reanalysis products—ERA5-Land, GLDAS_NOAH, MERRA-2, and CLDAS—using in-situ validation and the triple collocation analysis (TCA) method. The results showed that among the four products, CLDAS exhibited the highest consistency with in-situ measurements, with a correlation coefficient (R) of 0.75 and the lowest unbiased root mean square error ( u b R M S E) of 0.032. Building upon this, we developed a high-accuracy soil moisture dataset (MSF-SSM) by integrating reanalysis products with in-situ data using a random forest and convolutional neural network fusion model (CNN-RF). The MSF-SSM demonstrated high consistency (R>0.7) and low Bias (±0.02) at most stations, significantly outperforming existing soil moisture products. Based on MSF-SSM, a standardized soil moisture index (SSMI_1) was then constructed to reveal the spatiotemporal evolution characteristics and frequency distribution of agricultural drought severity in the Huaibei Plain during 2015-2019. The study found that most areas exhibited a declining trend in SSMI_1, indicating reduced soil moisture and increased drought risk. Mild droughts occurred frequently across the region, while severe droughts were concentrated in the central part. The northeastern region of Huaibei Plain showed a high cumulative frequency of drought, indicating high susceptibility. The findings provide critical data support and technical reference for agricultural drought monitoring, meteorological services, and disaster risk management, and offer a scientific basis for optimizing water resource management measures such as water-saving irrigation.

In the context of global warming, for the Ili River Valley, known as a unique "wet island" within the arid regions of Central Asia, studying the mechanisms behind its shifting wet-dry patterns is of great scientific significance. This understanding is crucial for regional ecological security and water resource management. Based on meteorological observation data from 1961 to 2023, this study employs methods such as the Mann-Kendall abrupt change test, Morlet wavelet analysis, and the Hurst exponent to systematically reveal the spatiotemporal evolution and future trends of climatic factors in this region. The results indicate that: ① The regional climate exhibits a significant warming and wetting trend, with annual mean temperature and precipitation increasing at rates of 0.42 ℃/10 a (p<0.01) and 11.45 mm/10 a (p<0.01) respectively. Abrupt climate shifts were identified in 2000 for temperature and 1984 for precipitation. ② The most pronounced increases in both temperature and moisture occur in winter, with warming rates (0.53 ℃/10 a) and rates of moisture increase (3.78 mm/10 a) surpassing those of other seasons. ③ The Standardized Precipitation Evapotranspiration Index (SPEI) reveals a slight trend toward aridification (-0.09/10 a), with seasonal divergence showing winter wetting (+0.19/10 a) and summer aridification (-0.16/10 a). ④ The Hurst exponent predicts the continuation of the current warming and wetting trend, yet the asymmetry in the water-heat balance may exacerbate latent drought risks. The paradox of "apparent warming and wetting versus potential aridification" highlighted in this study provides crucial scientific evidence for resilient water resource management in the region.

To explore the effects of different rain-storing intermittent irrigation regimes on rice yield improvement, water saving, pollution control, and lodging resistance, and to identify the optimal irrigation strategy, a lysimeter experiment was conducted based on two irrigation lower limits: slight intermittent (SA) and heavy intermittent (HA). An orthogonal design was applied across rice growth stages to establish water management schemes by superimposing different gradients of water depth as the post-rainfall upper limits. The entropy-weighted TOPSIS method was used to comprehensively evaluate 11 indicators covering yield, water saving, pollution control, and lodging resistance. The results showed that rain-storing intermittent irrigation had a significant condition-dependent effect on yield: among 36 treatments, 17 increased yield (with an average increase of 6.46%) and 19 reduced it (with an average decrease of 12.91%). Achieving synergy between yield improvement and water saving was key to enhancing irrigation water use efficiency. The treatment with the highest efficiency exceeded that of traditional flooding irrigation by 46.67%, and was 90.27% higher than the lowest-efficiency treatment. Rain-storing intermittent irrigation also helped control agricultural non-point source pollution, with 35 out of 36 treatments showing lower average total nitrogen concentrations in percolated water at 40 cm depth than the flooding control. Reasonable rain-storing intermittent irrigation improved rice lodging resistance by up to 17.3% compared with traditional flooding. Regulation of the post-rainfall upper limit during the greening stage had the most significant impact on the comprehensive score index. When irrigation water was not properly distributed across growth stages, the negative effects under heavy intermittent irrigation were more pronounced, leading to lower comprehensive scores. Under the heavy intermittent irrigation regime, controlling the post-rainfall upper limits at 50 mm, 80 mm, 100 mm, 120 mm, and 90 mm during the greening, early tillering, jointing–booting, heading–flowering, and milky stages, respectively, yielded results closest to the ideal outcome of high yield, water saving, pollution control, and lodging resistance.

This study aimed to explore the effects of different water-retaining agents combined with organic carbon fertilizer on the growth, flue-cured tobacco, and its water and fertilizer use efficiency. A field experiment was conducted with eight treatments: a control with no amendments (CK); organic carbon fertilizer alone (T1, 375 kg/hm²); commercial water-retaining agent (WRA) alone (T2, 90 kg/hm²); self-made AM/CMC WRA alone (T3, 90 kg/hm²); self-made AM/HAK WRA alone (T4, 90 kg/hm²); and their combinations with organic carbon fertilizer (T5, T6 and T7, respectively). We investigated the effects on soil nutrients, moisture content, tobacco growth, water and fertilizer use efficiencies, and tobacco leaf quality. The results showed that the combined application of AM/HAK WRA and organic carbon fertilizer (T7) yielded the best results. At 75 days, the soil moisture content in T7 increased by 18.52% and 9.74% respectively compared with the CK and T5 treatments, respectively, and the soil nutrient content was also significantly higher than that of the CK and T5 treatments. The combined application of AM/HAK WRA and organic carbon fertilizer promoted the growth and development of flue-cured tobacco, and the biomass of tobacco plants was significantly increased compared with the treatments of CK and T5. It also significantly increased the net photosynthetic rate, stomatal conductance, and intercellular carbon dioxide concentration of tobacco leaves. In the curved leaves, the total sugar, reducing sugar, and potassium contents increased significantly by 34.13%、23.45% and 32.65% respectively compared with CK, and significantly increased by 9.80%、6.25% and 16.07% respectively compared with T5 treatment. The water use efficiency was significantly increased by 44.32% and 6.09% compared with the CK and T5 treatments, respectively. The use efficiencies of N, P₂O₅ and K₂O were also significantly higher than CK and T5. In conclusion, the combined application of AM/HAK compound WRA and organic carbon fertilizer can promote the growth of flue-cured tobacco and improve its water and fertilizer use efficiencies and quality.

Saline-alkali stress is a major abiotic factor limiting rice growth, yield, and quality. As the world′s third-largest country in saline-alkali soil area, China faces urgent challenges in enhancing the agricultural productivity of saline-affected lands to ensure food security. To systematically review the research progress and evolutionary trends of rice under saline-alkali stress, this study performed a bibliometric analysis of literature from 2002 to 2023 using the Web of Science Core Collection and CNKI databases, aided by CiteSpace and VOSviewer tools. The results revealed a continuous growth in global research output, with Chinese institutions playing a significant role. Research trends have evolved from phenotypic assessments to in-depth studies on physiological and molecular mechanisms, focusing on osmotic regulation, antioxidant defense, and ionic homeostasis. Rice responses to saline-alkali stress are growth stage-specific: moderate tolerance at germination, high sensitivity during the seedling stage, and significant grain filling and quality reduction under stress in the reproductive phase. However, current studies face limitations in simulating combined stress conditions, the insufficient integration of molecular mechanisms, and bridging the gap between theory and application. Future research should emphasize the mechanisms of response to combined stresses, reproductive-stage regulation, and the adaptability evaluation of salt-tolerant rice varieties to promote the efficient translation from basic research to field applications, thereby achieving sustainable rice production in saline-alkali environments.

This study investigated the removal efficiency of nitrogen pollutants in tidal flow constructed wetlands using zeolite substrates with different particle sizes and gradations. Twelve vertical subsurface flow constructed wetland (VSFCW) units were established in six parallel groups. Three zeolite particle sizes (2~4, 4~8 and 8~16 mm) were used as packing material in three different configurations: homogeneous, well-graded (positive gradation), and inversely-graded (reverse gradation). The systems were operated in two phases under four different inflow modes to analyze the effects on nitrogen removal and porosity changes. For NH??-N removal, both tidal flow and continuous flow showed high and similar efficiency in the inversely-graded wetlands. The highest removal rate (98.14%) was achieved in the inversely-graded system under tidal flow with a 2∶1 influent-to-idle ratio. In contrast, for the homogeneous wetlands, continuous flow demonstrated a significant advantage. The overall NH??-N removal efficiency ranked as follows: inversely-graded ≈ well-graded > homogeneous. For NO??-N, continuous flow was generally more effective than tidal flow. The highest removal rate (71.36%) occurred in the inversely-graded system under continuous flow. No clear superiority was observed among the different gradations, but the efficiency of different inflow modes was ranked as: continuous flow > 2∶1 > 1∶2 > 1∶1 influent-to-idle ratio. For Total Nitrogen (TN), the tidal flow with a 2∶1 influent-to-idle ratio achieved better removal across all three gradation types compared to continuous flow, while other tidal flow ratios showed no significant advantage. The TN removal efficiency ranking was: inversely-graded ≈ well-graded > homogeneous. Additionally, the porosity in all VSFCW units gradually decreased over the course of the experiment.

To improve the non-destructive monitoring of soil moisture and nutrients in agricultural fields, this study explored the utility of unmanned aerial vehicle (UAV) multispectral imagery for estimating soil water and nitrogen content. The study was conducted in a field that was plowed after the spring corn harvest. A DJI M3M UAV was used to collect multispectral data, while surface soil samples were collected simultaneously to measure soil moisture content, soil nitrate nitrogen, and soil ammonium nitrogen. The multispectral images were processed through mosaicking and radiometric correction to extract spectral reflectance. Prediction models linking spectral data to soil properties were then constructed using three machine learning algorithms: support vector regression, random forest regression, and partial least squares regression. The results showed that models predicting soil moisture content achieved high accuracy. The random forest regression performed best, with an R ² of 0.776 on the training set and 0.661 on the testing set. In contrast, the models developed to predict soil nitrate nitrogen and ammonium nitrogen both yielded low accuracy. This study provides a theoretical basis for selecting UAV multispectral inversion models for soil water and nitrogen information in autumn-plowed farmland, and offers data support for the implementation of smart farming.