Crop evapotranspiration estimation is a key parameter for achieving functional irrigation systems. However, ET is difficult to directly measure, so an ideal solution was to develop a simulation model to obtain ET. There are many ways to calculate ET, most of which use models based on the Penman–Monteith equation, but they are often inaccurate when applied to greenhouse crop evapotranspiration. The use of machine learning models to predict ET has gradually increased, but research into their application for greenhouse crops is relatively rare. We used experimental data for three years (2019–2021) to model the effects on ET of eight meteorological factors (net solar radiation (R_n), mean temperature (T_a), minimum temperature (T_amin), maximum temperature (T_amax), relative humidity (RH), minimum relative humidity (RH_min), maximum relative humidity (RH_max), and wind speed (V)) using a greenhouse drip irrigated tomato crop ET prediction model (XGBR-ET) that was based on XGBoost regression (XGBR). The model was compared with seven other common regression models (linear regression (LR), support vector regression (SVR), K neighbors regression (KNR), random forest regression (RFR), AdaBoost regression (ABR), bagging regression (BR), and gradient boosting regression (GBR)). The results showed that R_n, T_a, and T_amax were positively correlated with ET, and that T_amin, RH, RH_min, RH_max, and V were negatively correlated with ET. R_n had the greatest correlation with ET (r = 0.89), and V had the least correlation with ET (r = 0.43). The eight models were ordered, in terms of prediction accuracy, XGBR-ET > GBR-ET > SVR-ET > ABR-ET > BR-ET > LR-ET > KNR-ET > RFR-ET. The statistical indicators mean square error (0.032), root mean square error (0.163), mean absolute error (0.132), mean absolute percentage error (4.47%), and coefficient of determination (0.981) of XGBR-ET showed that XGBR-ET modeled daily ET for greenhouse tomatoes well. The parameters of the XGBR-ET model were ablated to show that the order of importance of meteorological factors on XGBR-ET was R_n > RH > RH_min> T_amax> RH_max> T_amin> T_a> V. Selecting R_n, RH, RH_min, T_amax, and T_amin as model input variables using XGBR ensured the prediction accuracy of the model (mean square error 0.047). This study has value as a reference for the simplification of the calculation of evapotranspiration for drip irrigated greenhouse tomato crops using a novel application of machine learning as a basis for an effective irrigation program.

This study aimed to systematically evaluate the effects of different nitrogen application rates and irrigation practices on water-saving and yield enhancement in winter wheat production in the North China Plain (NCP) using a meta-analysis. By quantifying the impacts on crop yield, nitrogen use efficiency (NUE), and water use efficiency (WUE), the research provides a scientific basis for optimizing management practices in winter wheat production in this region. A comprehensive literature search was conducted across multiple databases, resulting in the inclusion of 94 eligible studies from 2018 to 2023. A random-effects model was employed to calculate the combined effect sizes, followed by subgroup and sensitivity analyses to further investigate the influence of nitrogen application rates, irrigation methods, and study regions on winter wheat production efficiency. The findings reveal that increasing nitrogen application rates and adopting deficit irrigation practices significantly improved winter wheat yield (combined effect size: 4.53 t·ha⁻¹), NUE (43.29%), and WUE (0.013 t·ha⁻¹·mm⁻¹). The subgroup analysis further elucidated the critical roles of nitrogen application ratios, irrigation methods, and study regions in determining winter wheat production efficiency, while the sensitivity analysis confirmed the robustness of these findings, as the pooled effect sizes decreased by merely 0.69% and increased by 0.61% after excluding small-sample or highly biased studies, respectively. The above meta-analysis did not incorporate long-term field trials; hence, two-year field experiments with designed irrigation and organic–inorganic fertilizer treatments were conducted, which provided further validation for the meta-analysis. Under short-term conditions (excluding CO₂ effects), we observed that chemical fertilizer exhibited a measurable inhibitory effect on crop water uptake and optimal water–fertilizer management was achieved with a 7:3 inorganic–organic fertilizer ratio combined with 450 m³·ha⁻¹ irrigation. This study demonstrates the effectiveness of optimizing nitrogen fertilization and irrigation management in enhancing winter wheat yield and resource utilization efficiency. The findings offer actionable insights for sustainable agricultural practices in the NCP and similar regions, contributing to improved crop productivity and resource conservation.

The improvement of the simulation accuracy of crop models in different greenhouse environments would be better applied to the automation management of greenhouse cultivation. Tomatoes under drip irrigation in a greenhouse were taken as the research object, and the cumulative evaporation capacity (Ep) of the 20 cm standard evaporation dish was taken as the basis for irrigation. Three treatments were set up in the experiment: high water treatment without mulch (NM-0.9 Ep), high water treatment with mulch (M-0.9 Ep), and low water treatment with mulch (M-0.5 Ep). AquaCrop and DSSAT models were used to simulate the canopy coverage, soil water content, biomass, and yield of the tomatoes. Data from 2020 were used to correct the model, and simulation results from 2021 were analyzed in this paper. The results showed that: (1) Of the two crop models, the simulation accuracy of the greenhouse tomato canopy coverage k_CC was higher, and the root mean square errors were less than 6.8% (AquaCrop model) and 8.5% (DSSAT model); (2) The AquaCrop model could accurately simulate soil water change under high water treatments, while the DSSAT model was more suitable for the conditions without mulch; (3) The relative error RE of simulated and observed values for biomass B, yield Y, and water use efficiency WUE in the AquaCrop model were less than 2.0%, 2.3%, and 9.0%, respectively, while those of the DSSAT model were less than 4.7%, 7.6%, and 10.4%, respectively; (4) Considering the simulation results of each index comprehensively, the AquaCrop model was superior to the DSSAT model; subsequently, the former was used to predict 16 different water and film coating treatments (S1–S16). It was found that the greenhouse tomato yield and WUE were the highest under S7 (0.8 Ep), at 8.201 t/ha and 2.79 kg/m³, respectively.

This study presents an innovative unmanned aerial vehicle (UAV)-based intelligent detection method utilizing an improved Faster Region-based Convolutional Neural Network (Faster R-CNN) architecture to address the inefficiency and inaccuracy inherent in manual wheat spike counting. We systematically collected a high-resolution image dataset (2000 images, 4096 × 3072 pixels) covering key growth stages (heading, grain filling, and maturity) of winter wheat (Triticum aestivum L.) during 2022–2023 using a DJI M300 RTK equipped with multispectral sensors. The dataset encompasses diverse field scenarios under five fertilization treatments (organic-only, organic–inorganic 7:3 and 3:7 ratios, inorganic-only, and no fertilizer) and two irrigation regimes (full and deficit irrigation), ensuring representativeness and generalizability. For model development, we replaced conventional VGG16 with ResNet-50 as the backbone network, incorporating residual connections and channel attention mechanisms to achieve 92.1% mean average precision (mAP) while reducing parameters from 135 M to 77 M (43% decrease). The GFLOPS of the improved model has been reduced from 1.9 to 1.7, an decrease of 10.53%, and the computational efficiency of the model has been improved. Performance tests demonstrated a 15% reduction in missed detection rate compared to YOLOv8 in dense canopies, with spike count regression analysis yielding R² = 0.88 (p < 0.05) against manual measurements and yield prediction errors below 10% for optimal treatments. To validate robustness, we established a dedicated 500-image test set (25% of total data) spanning density gradients (30–80 spikes/m²) and varying illumination conditions, maintaining >85% accuracy even under cloudy weather. Furthermore, by integrating spike recognition with agronomic parameters (e.g., grain weight), we developed a comprehensive yield estimation model achieving 93.5% accuracy under optimal water–fertilizer management (70% ETc irrigation with 3:7 organic–inorganic ratio). This work systematically addresses key technical challenges in automated spike detection through standardized data acquisition, lightweight model design, and field validation, offering significant practical value for smart agriculture development.

Accurate short-term forecasts of daily reference evapotranspiration (ET₀) are essential for real-time irrigation scheduling. Many models rely on current and historical temperature data to estimate daily ET₀. However, easily accessible temperature forecasts are relatively less reported in short-term ET₀ forecasting. Furthermore, the accuracy of ET₀ forecasting from different models varies locally and also across regions. We used five temperature-dependent models to forecast daily ET₀ for a 7-day horizon in the North China Plain (NCP): the McCloud (MC), Hargreaves-Samani (HS), Blaney-Criddle (BC), Thornthwaite (TH), and reduced-set Penman–Monteith (RPM) models. Daily meteorological data collected between 1 January 2000 and 31 December 2014 at 17 weather stations in NCP to calibrate and validate the five ET₀ models against the ASCE Penman–Monteith (ASCE-PM). Forecast temperatures for up to 7 d ahead for 1 January 2015–19 June 2021 were input to the five calibrated models to forecast ET₀. The performance of the five models improved for forecasts at all stations after calibration. The calibrated RPM is the preferred choice for forecasting ET₀ in NCP. In descending order of preference, the remaining models were ranked as HS, TH, BC, and MC. Sensitivity analysis showed that a change in maximum temperature influenced the accuracy of ET₀ forecasting by the five models, especially RPM, HS, and TH, more than other variables. Meanwhile, the calibrated RPM and HS equations were better than the other models, and thus, these two equations were recommended for short-term ET₀ forecasting in NCP.

The combined effects of tillage and organic amendments on microbial respiration and its contribution to soil hydraulic conductivity are still uncertain in the 0–40 cm layer of a loess soil. We conducted a two-year field experiment to explore the effects of organic amendments, tillage and their interaction on soil microbial respiration, aggregate stability, pore parameters, and hydraulic conductivity on the Loess Plateau. Three tillage methods (conventional tillage (CT), deep tillage (DT) and no tillage (NT)) plus five fertilizer treatments (mineral fertilizer (control) alone and along with 20 t ha⁻¹ wheat straw (MWS), wheat husk (MWH), farmyard soil (MFS) and bioorganic fertilizer (MBF)) were set up as experimental treatments. The findings demonstrated that the organic amendments significantly increased the soil microbial respiration and saturated hydraulic conductivity compared to the control in the 0–10 cm and 10–20 cm layers. Soil microbial respiration had indirect effects on hydraulic conductivity by improving the water aggregate stability and macroporosity. Additionally, the interaction effects of tillage and organic amendments on the pore and hydrological parameters were significant in the 20–40 cm layer. NT-MBF resulted in the greatest saturated hydraulic conductivity, which was directly correlated with the soil’s strong pore organization. Given the issue of subsurface soil compaction in our study area, it is recommended that local farmers adopt NT-MBF to enhance the soil’s microbial, structural and hydrological properties.