Accurate and timely crop type mapping is essential for smart irrigation district development, providing crucial data for crop growth monitoring, yield estimation, and agricultural ecological assessments. Deep learning models have demonstrated superior performance in crop classification research, offering great potential for automating and enhancing classification accuracy. However, deep learning models still face limitations in handling time-series remote sensing data, requiring further refinement. This study focuses on crop classification in the Hetao Irrigation District using remote sensing data from Sentinel-2 satellites. To address time-series data challenges, linear interpolation was used to synthesize data from May to October 2020. The Enhanced Vegetation Index (EVI) was derived from the imagery to create a time-series dataset tracking vegetation changes throughout the growing season. Various deep learning models were used to analyze crop classification within the Hetao Irrigation District, comparing the performance of different approaches. Among the tested methods, the Temporal Feature-based Segmentation (TFBS) model achieved the highest classification accuracy, with mean Intersection over Union (mIoU), mean precision, and mean recall values of 0.872 2, 0.924 7, and 0.9260, respectively. These metrics indicate that the TFBS model outperformed other models. The TFBS model proved robust in handling both raw remote sensing imagery and EVI data, effectively extracting temporal features for improved classification accuracy. In contrast, the Unet model failed to converge with EVI data, struggling to capture time-series characteristics. This suggests that while Unet can be effective for certain types of image classification tasks, it struggles with the complexity of temporal data derived from vegetation indices, making it less suitable for time-series remote sensing applications in crop classification. The success of the TFBS model highlights the importance of selecting models designed for time-series data when working with dynamic factors like crop growth patterns. TFBS model demonstrated not only high classification accuracy but also strong robustness in processing the varying data inputs from remote sensing and vegetation indices. This provides valuable insights for future research and practical applications, guiding the selection of appropriate models for crop classification tasks in different agricultural contexts. The crop classification results provide important technical support for agricultural management in the Hetao Irrigation District.

To investigate the effect of risk irrigation based on public weather forecasts on rice growth and the corresponding water-saving potential, and the implementation method of rice risk irrigation knowledge graph to serve intelligent irrigation decisions at the same time. Rice experiments based on risk irrigation decisions were conducted to compare the effects of conventional irrigation and two risk irrigation strategies (low and high risk) on water consumption, yield, and water productivity of rice. A knowledge graph for irrigation decisions was constructed using the Neo4j graph database. The results indicate that, compared to conventional irrigation, low-risk and high-risk irrigation decisions saved 50.40 mm and 84.35 mm of irrigation water, respectively, with field water-saving rates of 16.05% and 26.86%, and yield reductions of 0.88% and 1.19%, respectively. The irrigation decision knowledge graph can assist managers in effective query retrieval and irrigation decision-making, laying a foundation for intelligent field moisture management. The study provides new ideas and feasible technologies for the development of water-saving irrigation and intelligent irrigation systems, which is significant for water resources conservation and sustainable development.

To enhance the efficiency of water resources utilization in irrigation districts, this study proposed a theoretical framework driven by TIGGE numerical weather forecasts and public weather predictions. The Hargreaves-Samani equation was adopted combined with genetic algorithm to rate the regional parameters. Then conducting reference crop evapotranspiration predicting and real-time irrigation forecasting with forecast periods of 3 days, 7 days, and 15 days for short-term, medium-term, and long-term periods, respectively. Using the Qumutang Irrigation District in the “Hengyang-Shaoyang” Dry Corridor as a case study, the results show that: ① The accuracy of numerical weather forecasts effectively extends the irrigation forecast period. ② By optimizing the key parameters in the HS equation, the accuracy of ET ₀ can be improved with the correlation coefficient reached by 0.9. ③ Analysis of water-saving effects on paddy fields and dry crops under varying forecast periods reveals that long-term forecasts, compared to short- and medium-term forecasts, significantly reduce water wastage. Additionally, long-term forecasts mitigate risks of post-irrigation precipitation through small-volume, high-frequency irrigation events. ④ A daily rolling “Short (1~3 days) – Medium (4~7 days) – Long (8~15 days)” integrated water-saving irrigation strategy has been formulated, providing decision-making support for water resource regulation, planning and management in irrigation districts.

Karst peak-cluster depression is a typical karst geomorphology type area in southwest China, there are serious problems of rocky desertification, soil erosion and ecological degradation in karst area. Soil moisture is one of the key factors limiting the ecological restoration and ecosystem stability in karst peak-cluster depression. The study of soil moisture is of great significance to maintain the ecological environment and human production and life. However, the current research lacks of continuous monitoring of soil moisture under different land use modes in karst peak cluster depression area, and does not discuss the influence of rainfall infiltration on soil moisture content change based on real-time monitoring data. Therefore, this study was carried out in the typical peak cluster depression area, and monitored the soil moisture under five different land use modes of peak cluster slope. By analyzing the seasonal variation of soil moisture and the response characteristics of soil moisture under different rainfall intensities, the soil moisture change and rainfall infiltration mechanism under different land use modes in the karst peak cluster slope depression area were explored. The results showed as follows: ① The average soil moisture content in maize was the lowest, and the average soil moisture content in Erythropalum scandens was the highest. The response of soil moisture content in maize + Euonymus fortunei and Erythropalum scandens to rainfall in rainy season was not significant; ② the soil moisture content in the traditional maize planting scandens, Pitaya scandens, and the maize + Ficus pumila scandens was more stable in the dry season, while the soil moisture content in the maize + Euonymus fortunei scandens and Erythropalum scandens was more stable in the rainy season. In the rainy season, the response of soil moisture to rainfall was more sensitive in each experimental plot; ③ under heavy and torrential rainfall conditions, the soil moisture content of the five land-use modes responded significantly to rainfall. The greater the intensity of rainfall, the shorter the lag period for soil moisture to be recharged. Under heavy rainfall conditions, soil moisture was effectively recharged in all experimental plots, while under torrential rainfall conditions the soil moisture content of all experimental plots was reduced instead. ④ Soil moisture in the Erythropalum scandens was minimally affected by rainfall, in which the response of the surface soil to rainfall was weaker than that of the deep soil. In contrast, soil moisture in the traditional maize planting area was most affected by rainfall, and the deep soil was most significantly affected by rainfall.

To explore the impact of industrial solid waste fly ash application on soil fertility, this study conducted a Meta-analysis of relevant literature up to 2024, and systematically analyzed the effects of factors such as fly ash application rate, application method, and auxiliary materials on soil physicochemical properties. The results showed that: ① The application of fly ash can significantly improve the field water holding capacity, available nutrients and porosity of the soil, and reduce the bulk density of the applied soil. The application rate of fly ash with a mass fraction greater than 10% is more effective in improving of soil field water holding capacity, available phosphorus, porosity and reducing bulk density, and the field water holding capacity, available phosphorus and porosity are increased by 20.9%, 144.1% and 18.1%, respectively, and the bulk density is reduced by 19.8%, but it is not conducive to the improvement of soil available nitrogen content. ② The application of fly ash alone is more conducive to the improvement of soil porosity and the reduction of bulk density. Specifically, the porosity increased by 12.8% and the soil bulk density decreased by 7.9%. Combining fly ash with organic materials has more advantages in improving soil fertility. ③ The effectiveness of fly ash application is influenced by soil pH and texture. The application of fly ash in acidic soil is more conducive to the increase of available nutrients, whereas in neutral soil, it is more effective in increasing of porosity and the decrease of bulk density; The application of fly ash in sandy soil is more conducive to the improvement of available nutrients, while in clay, it is more conducive to the increase of porosity and the decrease of bulk density. In summary, solid waste fly ash can be used as an improver for the fertility of low yield soil, comprehensively enhancing soil fertility. When combined with organic materials, it has a significant effect on improving the available nutrient content of acidic sandy soil; Applying more than 10% fly ash alone to neutral clay can enhance pore structure.

To explore the temporal dynamics of soil moisture over time in typical plantation woodlands in Xining City, three types of woodland (Pinus tabuliformis × Picea crassifolia forest, Pinus tabuliformis forest, and Picea crassifolia forest) in Nanshan Mountain of Xining City were selected as study sites. A soil moisture monitoring site was established in each sample plot, and a tube-based soil moisture monitoring system was employed to measure the soil volumetric moisture content from May to October 2023. The measurements were taken at a depth of 100 cm, with every 10 cm representing one measurement layer. Soil moisture content was measured at the beginning, middle, and end of each month to analyze the time dynamic change pattern of soil moisture in different forest types. The results showed that: ① At the daily scale, soil moisture content in the surface layer exhibited a 'V'-shaped pattern during the daytime on sunny and cloudy days. Daily fluctuations in soil moisture content were negatively correlated with forest temperature and soil temperature. ② On a seasonal scale, soil moisture changes were divided into three periods: soil moisture stabilization, soil moisture depletion, and soil moisture gain. ③ The influence of meteorological factors on soil moisture varied with precipitation conditions. Under precipitation, the ranking of influencing factors was: atmospheric humidity > soil temperature > in-forest temperature > wind speed > solar radiation; under no precipitation weather, atmospheric humidity > wind speed > in-forest temperature > soil temperature > solar radiation.

Accurate calculation of seepage loss in earthen canals is important for estimating water conveyance efficiency and irrigation management in irrigation districts. In order to propose an accurate and efficient method for assessing canal seepage, a numerical model for two-dimensional infiltration was established with the Yichang Irrigation Area of the Hetao Irrigation District as the study area. The model was validated using ponding test data. This study further investigates the effects of canal water depth (h), bottom width (b), slope coefficient (m), and saturated hydraulic conductivity of the layered soils (k_{s 1}, k_{s 2}, and k_{s 3}) on the soil water infiltration process in the canals. A multivariate linear regression analysis was employed to create a multifactor model based on infiltration parameters, using the Kostiakov-Lewis model as the foundational model. The results indicate that the R2 values of the numerical model exceed 0.928, RMAE and RMSE values are both below 14.57% and 0.048 cm³/(cm·s), respectively. This suggests that the model is capable of accurately simulating the infiltration process of canal seepage. The factors influencing the stable seepage rate of the canal was ranked in the following order: k_{s 1}> h > k_{s 2} > k_{s 3}> b (where m was not significant). The established two-dimensional multi-factor mathematical model for earthen canal infiltration achieved an R2 value greater than 0.87, with a mean value of 0.96. The mean values for RMSE value, RMAE value, and ME value are 0.05 cm³/(cm·s), 14.13%, and 0.027 cm³/(cm·s), respectively. The evaluation indices indicate that the results are favorable, with simulated and predicted values showing excellent consistency, confirming the model's feasibility. This study quantified the effects of various influencing factors on the stable seepage rate of the canal. It was found that the soil layer adjacent to the canal bed and the water depth significantly affected canal seepage. Additionally, the established practical formula for seepage in earthen canals can accurately and efficiently calculate the seepage of canals at all levels under layered soil conditions.

This study explores the influence of changes in the structural parameters of guide vane cyclones on their internal hydraulic characteristics. Numerical simulation and physical experiments are used to analyze the hydraulic performance of cyclones with different guide vane heights. The results show that the axial velocity is lower near the surface of the guide vane in the straight section and the less twisted section of the guide vane, but higher in the center of the pipe and far away from the guide vane. For the twisting section with a large degree of twisting, the axial velocity gradually increases from the frontal surface to the backwater surface, and increases with the increase of the height of the guide vane. The radial velocity of the straight section of the guide vane is low, while the radial velocity of the twisted section increases with the increase of the degree of twisting. Near the frontal surface, the radial velocity is towards the pipe axis, while near the backwater is towards the pipe wall, and the two velocity modes alternate; The increase of the degree of distortion of the guide vane leads to the increase of the circumferential velocity of the cross-section. The circumferential velocity near the backwater and headwater is higher than in the area away from the guide vanes, and the velocity on the backwater is higher than the headwater. A negative circumferential velocity region exists at the inner edge of the guide vane. The absolute value of its minimum decreases with increasing guide vane height. Local mechanical energy loss at the leading and trailing edges of the guide vanes increases with increasing guide vane height. The mechanical energy loss along the guide vane section remains basically unchanged, but the higher the height of the guide vane, the faster the loss rate.The mechanical efficiency of the cyclone decreases with the increase of the guide vane height and increases with the increase of the Reynolds number. The conclusions of this paper provide a theoretical basis for further optimization of the structural parameters of the cyclone.

This study investigates the effect of the inverted cone bottom angle of the dropper on local head loss in drip irrigation pipes. CFD numerical analysis was used to simulate drip irrigation pipes with 8 different drip head inclination angles. The influence of the inverted cone bottom angle of the drip head on the flow field and local head loss of the drip irrigation pipe within the range of Reynolds numbers from 2 700 to 83 000 was analyzed. The results show that as the inverted cone bottom angle increases, both the local head loss and the increase in head loss due to higher inlet flow velocity first decrease, then increase, and decrease again. At an inlet flow velocity of 0.2 m/s, the optimal drip head inverted cone bottom angle for minimizing local head loss is 78°. When the inlet flow velocity exceeds 0.2 m/s, the optimal angle increases to 81°. Changes in the drip head inverted cone bottom angle affect the flow velocity components in the Y-axis and Z-axis directions. The turbulent kinetic energy losses in both directions jointly influence the magnitude of the local head loss in the drip irrigation pipe. When designing drip heads, that the inverted cone bottom angle be around 81°.

Water and soil resources are key elements in maintaining the stability and sustainable development of oases in inland arid zones. The establishment of a harmonious, stable, and sustainable oasis requires further research on the appropriate scale of the oasis, based on water resources carrying capacity (WRCC) and its alignment with specific economic and technical conditions. This paper examines the current status, carrying capacity, and stability of oasis development in Xinjiang watersheds using the spatial Norenz curve, the maximum capacity model, and the oasis water-heat balance equation. The results show that: the overall Gini coefficient of the study area is high, and the match between oasis area ratio and non-agricultural water use is poor. 80% of the watersheds are under light and medium development, and are dominated by agricultural watersheds. 90% of the watersheds have non-agricultural water use of less than 40%, and the urbanization rate is low, and the development is relatively slow. The overloading of oasis areas in the study region is severe, and the stability of the northern oasis is destabilizing. The entire region has no potential for further development and depends on inter-basin water transfer to maintain the current status. The findings of the study serve as a foundation for decision makers in formulating rational oasis development plans.

In order to study the current situation and future development trend of water resources in Tianjin, the water resources system of Tianjin is divided into four subsystems, namely, population, economy, environment and resources, based on the system dynamics theory, and a system dynamics model is constructed to simulate the water resources supply-demand system of Tianjin according to the causal relationship between the variables. The model is tuned and validated against the model using the data of the 2012-2019 and 2020-2023 periods, respectively. The absolute values of the relative errors of each test variable in the validation period are less than 10%, which indicates that the model can accurately simulate the relationship between water resources supply and demand in Tianjin. On this basis, five development scenarios, namely the continuation of the current situation (S₁), economic priority (S₂), environmental priority (S₃), increasing supply and reducing demand (S₄) and comprehensive development (S₅), are set to predict the future water resources situation of Tianjin. The ecological footprint method is used to quantitatively evaluate the current situation of water resources in Tianjin (from 2010 to 2023) and the future situation under different scenarios (2024-2035) from the perspectives of water resources pressure, water resources utilization efficiency and development coordination. The results show that the water resources situation in Tianjin has been in a dangerous state for a long time from 2010 to 2023, the ecological surplus/deficit of per capita water resources has been in a deficit state for many years (with a multi-year average of -0.199 hm²), and the average value of the water resources ecological pressure index is 3.37. In future development, the water resources ecological pressure index under different scenarios will decrease. Increasing the amount of water transferred from outside will significant help alleviate the water resources predicament in Tianjin. However, unreasonable planning schemes (such as only focusing on the economy or blindly increasing the ecological water consumption) will lead to the imbalance of water resources allocation and the reduction of the degree of water resources security. By 2035, the water resources ecological pressure index of the current trend scenario S₁ and the comprehensive development scenario S₅ will be 2.84 and 2.77 respectively, and the intensity of water resources ecological footprint will be 247.8 hm²/ (100 million yuan) and 239.3 hm²/ (100 million yuan) respectively. This indicates that the comprehensive development scenario (S₅) takes into account both water resources security and economic development, ensuring the sustainability of water resources in Tianjin.

Traditional ground irrigation and fertilization methods are commonly used in Fenhe Irrigation District, and the use of submembrane drip irrigation for corn planting is still in the early stages of development. In order to explore the effects of submembrane drip irrigation on soil water and nitrogen distribution and corn water and nitrogen utilization in this area, this study carried out a spring corn water and nitrogen coupling experiment under membrane drip irrigation in Wenshui County, Shanxi Province. In order to put forward a suitable drip irrigation and fertilization system under film which can ensure higher yield and water and nitrogen utilization efficiency of maize. Three irrigation amounts were set in the experiment: from the lower limit of irrigation (70% of seedling stage and 75% of remaining growth stage) to the irrigation amount required for 100% of field water capacity (W100), 60% of W100 irrigation amount (W60), and no irrigation water (W0); Three nitrogen application rates: 300 kg/hm² (N300 local conventional nitrogen application), 200 kg/hm² (N200), 0 kg/hm² (N0). The results show that: Soil moisture content and nitrogen reduction rate of 33.3% under W100 were more suitable for local corn growth. After corn heading, leaf area index under N200W100 treatment increased by 6.4%~25.9% compared with N0W100 and N300W100. The soil water content decreased by 1.3% to 18.3% because the larger leaf area dominated the transpiration water consumption of the plants. After harvest, the content of nitrate nitrogen and ammonium nitrogen in the 0~60 cm soil layer under N200 remained largely unchanged compared to pre-sowing levels, while N300 levels were 24.7%~35.9% and 17.2%~78.9% higher, respectively, increasing soil nitrogen accumulation in the corn root zone. Compared with the local conventional nitrogen application, N200W100 treatment achieved the highest spring maize yield and nitrogen use efficiency (14 459 kg/hm² and 36.9%, respectively) and water-use efficiency (38.9 kg/hm²/mm) under the condition of saving 33.3% of nitrogen fertilizer. In conclusion, N200W100 water and nitrogen management scheme is recommended for spring maize under drip irrigation in Fenhe irrigation District, which can not only reduce nitrogen fertilizer input but also obtain higher water-nitrogen utilization efficiency.

This study explores the appropriate irrigation system for cauliflower under underground membrane-controlled irrigation conditions. For underground membrane controlled irrigation (G), three single irrigation quotas of 190 m³/hm² (L₁), 125 m³/hm² (L₂) and 95 m³/hm² (L₃) were set, and three control film burial depths of 25 cm (S₁), 20 cm (S₂) and 30 cm (S₃) were set. The irrigation quota of surface drip irrigation (F) without regulation film (S₀) was L₁, L₂, L₃ as the control, a total of 8 treatments. The results showed that the total irrigation amount during the entire growth cycle was 15%~25% lower than that of surface drip irrigation. Additionally, subsurface membrane-controlled irrigation led to higher yield and water use efficiency. Water use efficiency was 5.36%~8.56% higher in spring and 2.13%~8.19% higher in autumn compared to surface drip irrigation. Growth indices, yield and water use efficiency of cauliflower treated with S₂L₁-G were the highest under subsurface membrane-controlled irrigation conditions. Plant height, stem diameter, leaf area, dry matter weight and biological yield of cauliflower treated with S₂L₁-G were positively correlated with single irrigation quota and negatively correlated with subsurface membrane buried depth, and there were significant differences among most treatments. The biological yield and water use efficiency of cauliflower under S₁L₁-G and S₁L₁-G treatments were the highest in both spring and autumn. The values were 26.752 t/hm2 and 24.46 kg/m³， 27.860 t/hm2 and 24.91 kg/m³ in spring, and 24.496 t/hm2 and 23.62 kg/m³, 25.267 t/hm² and 24.21 kg/m³ in autumn. Although no significant difference was observed between the two treatments, considering the impact of a 20 cm membrane burial depth on field operations, the S₁L₁-G treatment with a membrane burial depth of 25 cm is recommended. By comprehensive analysis, the lower limit of broccoli irrigation can be set as 70% field water retention, the upper limit is 100% field water retention, and the depth of control film is 25 cm.

The unique geographical and climatic conditions of the northwestern region contribute to the frequent occurrence of droughts, making it crucial to understand the evolution characteristics of droughts for regional disaster prevention and reduction. Additionally, limited seasonal studies exist on the effects of droughts on vegetation and their periodicity. Therefore, this study adopts the Standardized Precipitation Evapotranspiration Index (SPEI) to characterize meteorological drought. Utilizing continuous wavelet transform and Rotating Empirical Orthogonal Function (REOF), this research examines the seasonal evolution characteristics of droughts in northwestern China from 1961 to 2020. Furthermore, Pearson correlation analysis, cross wavelet analysis, and wavelet coherence analysis are employed to investigate the response of the Normalized Difference Vegetation Index (NDVI) to SPEI from 1982 to 2020. The results indicate: ① REOF divides the study area into four subregions based on wet and dry conditions, with distinct spatial distribution differences. There are significant seasonal variations in the years of wet and dry transitions. ② SPEI exhibits notable periodicity at the seasonal scale, with oscillation periods concentrated between 1991 and 2001, though the lengths of these cycles vary. ③ The frequency of drought occurrence shows clear seasonality and regionality, with summer > spring > autumn. In terms of drought severity, moderate drought > mild drought > severe drought > extreme drought. ④ Seasonal drought significantly impacts vegetation growth in the northwestern region, with over 50% of the area demonstrating a positive correlation between SPEI and NDVI. The relationship between SPEI and NDVI variations, analyzed through wavelet coherence and cross wavelet methods, exhibits clear interannual periodicity.

In order to explore the influence of the nonlinearity of monthly mean potential evapotranspiration (ET_p ) estimation methods on the calculation results, a nonlinear study was conducted on seven commonly used estimation methods. Using ET_p calculated using daily meteorological data as the standard, the spatio-temporal characteristics of the deviations generated when using monthly scale data were analyzed. The results showed that the Hargreaves method produced the maximum absolute deviation ΔET_p 0.89 mm/d (relative deviation RE: 7.78%) in July, and the Blaney-Criddle method produced the minimum ΔET_p -0.51 mm/d (RE: -10.58%) in July. Penman-Monteith method, Hargreaves method and Makkink method tend to overestimate ET_p in spring and summer, but underestimate ET_p in autumn and winter. The annual ΔET_p values of the Hargreaves method and Priestley-Taylor method are larger. Among the 11 temperature belt representative stations, most of the methods have the largest ΔET_p at Dunhuang station. The main factors influencing ΔET_p of Hargreaves, Priestley-Taylor and Blaney-Criddle methods were maximum temperature, sunshine hours and average temperature, wind speed and minimum relative humidity, respectively. The study found that when calculating monthly ET_p, the direct use of monthly meteorological data will produce a significant deviation, and the influence of spatiotemporal characteristics caused by the nonlinearity of the estimation method must be considered appropriately.

To investigate the effects and differences of alternating drought and flood conditions at different growth stages on the yield and quality of winter wheat, a pot experiment was conducted using Huaimai 55 during the jointing and booting stages. Nine treatments involving varying levels and durations of drought (65% field capacity for 4, 8, and 12 days) and flood (water level 10 cm above ground for 4, 8, and 12 days) were established, along with a control group (CK). The study aimed to elucidate the impact patterns of alternating drought and flood during the jointing to booting stage on the quality characteristics and yield of winter wheat. Principal component analysis results were used in conjunction with the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) to quantitatively evaluate yield performance and overall quality. Findings indicated that a 12-day drought followed by a 12-day flood had the most significant effect on plant height, resulting in a suppression rate exceeding 40%. The most severe yield reduction was observed under treatments involving 12 days of drought followed by 4, 8, or 12 days of flooding, with a yield loss exceeding 90%. Drought and flooding intensities significantly influenced the quality of winter wheat. Notably, yield was most negatively correlated with protein content (r=-0.93) and most positively correlated with starch content (r=0.96). Under alternating drought and flood conditions, the treatment with 4 days of drought followed by 8 days of flood demonstrated the best comprehensive performance in terms of yield and quality, while the treatment with 12 days of drought followed by 12 days of flood performed the worst. These findings provide a robust scientific basis for enhancing the quality of winter wheat and optimizing strategies for drought and flood resistance.