Three irrigation treatments were set up in northeast China to investigate soil water movement and root water uptake of greenhouse tomatoes, and the collected experimental data were simulated by HYDRUS-2D. The computation and partitioning of evapotranspiration data into soil evaporation and crop transpiration was carried out with the double-crop coefficient method. The HYDRUS-2D model successfully simulated the soil water movement, producing RMSE ranging from 0.014 to 0.027, an MRE ranging from 0.062 to 0.126, and R² ranging from 79% to 92%, when comparing model simulations with two-year field measurements. Under different water treatments, 83–90% of the total root quantity was concentrated in 0–20 cm soil layer, and the more the water deficit, the more water the deeper roots will absorb to compensate for the lack of water at the surface. The average area of soil water shortage in W₁ was 2.08 times that in W₂. W₃ treatment hardly suffered from water stress. In the model, parameter n had the highest sensitivity compared with parameters α and K_s, and sensitivity ranking was n > K_s > α. This research revealed the relationships between soil, crop and water under drip irrigation of greenhouse tomatoes, and parameter sensitivity analysis could guide the key parameter adjustment and improve the simulation efficiency of the model.

Irrigation and nitrogen application in an irrational manner can reduce tomato yield and resource use efficiency. To investigate the effects of different water and nitrogen combinations on yield, dry matter, ET, water use efficiency, and partial factor productivity of nitrogen, we conducted a tomato plot experiment from August to December in 2020 and 2021.Three irrigation levels (65%~75% θf(W₁), 75%~85% θf(W₂), 85%~95% θf(W₃) and four nitrogen application rates 0 (N₀), 120 kg·hm^-2 (N₁), 240 kg·hm^-2 (N₂), 360 kg·hm^-2 (N₃)) were set in this experiment. We conducted ANOVA and Pearson’s correlation analysis for a number of indicators measured in the experiment, and also utilized structural equation modeling and Entropy Weighted TOPSIS(EWTOPSIS) modeling for comprehensive analysis and evaluation. The results showed that irrigation and nitrogen application had significant effects(P<0.05) on tomato fruit yield, total dry matter content, ET and water and nitrogen utilization efficiency. Increasing irrigation and nitrogen application would achieve higher yield and total dry matter content, but was not conducive to water conservation and efficient use of water and fertilizer. Structural equation modeling indicated that ET, WUE, and dry matter all had a positive effect on yield, with WUE being the main controlling factor for yield and having the greatest influence, with a path coefficient of even 0.81. The results of the EWTOPSIS model for the two-year experiment indicators showed N₁W₃ as the treatment with the best overall benefits. Ultimately, this study found that greenhouse tomatoes could be optimized economically in Northeast China by controlling soil water content to 85% to 95% of field capacity and applying 120kg hm^-2 of N.

Irrigation and nitrogen application rates have significant effects on greenhouse tomato yields, as well as water and nitrogen use efficiencies, but little is known regarding how these rates affect plant–microbiome interactions and how the associated changes might impact tomato yields. In this greenhouse study conducted over two years, the effects of three irrigation levels (moderate deficit with 65–75% water holding capacity threshold, slight deficit with 75–85%, and sufficient irrigation with 85–95%) and four nitrogen application levels (60, 120, 240, and 360 kg ha⁻¹) on tomato growth, yield, water and nitrogen productivities, and rhizosphere microbial diversities and functions were investigated. The results demonstrated that the highest tomato leaf area, dry biomass, yield, and water and nitrogen productivities were obtained under the treatment with sufficient irrigation. With increasing nitrogen application, the tomato leaf area, dry biomass, yield, and water and nitrogen productivities showed a trend of first increasing and then decreasing. Overall, the treatment (N2W3) with sufficient irrigation and 240 kg ha⁻¹ N was associated with the highest tomato growth, yield, and water and nitrogen productivities. Moreover, optimal irrigation and nitrogen application obviously altered the structures of rhizosphere bacterial and fungal communities, particularly recruiting microbiota conferring benefits to tomato growth and nitrogen fixation—namely, Lysobacter and Bradyrhizobium. Ultimately, optimal irrigation and nitrogen application significantly increased the relative abundances of functions related to carbon, sulfur, and nitrogen metabolism, especially nitrogen fixation. In summary, optimal irrigation and fertilization enhanced tomato yield, as well as water and nitrogen productivities by increasing the nitrogen fixation functions of the rhizosphere microbiome. Our results provide significant implications for tomato cultivation in greenhouses, in terms of optimized irrigation and fertilization.

Accurate quantification of evapotranspiration (ET_c) and its components are critical for enhancing water use efficiency and implementing precision irrigation. A two-year experiment was conducted for greenhouse-grown tomatoes under mulched drip irrigation with three irrigation treatments during 2020–2021 in Northeast China. Three different irrigation treatments were applied by setting upper and lower soil moisture irrigation thresholds (i.e., W1, 65%θ_FC–75%θ_FC, W2, 75%θ_FC–85%θ_FC, W3, 85%θFC–95%θ_FC, respectively, where θ_FC is field capacity). In this study, a modified dual crop coefficient (K_c) model was proposed to simulate daily ET_c, plant transpiration (T_r) and soil evaporation (E_s). The simulations of the model were validated against observed data from the sap flow system combined with the soil water balance method. The controlling factors on the variations of evapotranspiration and its components were also identified by using the path analysis method. Results showed that the modified dual K_c model can accurately simulate daily ET_c, E_s, and T_r for the greenhouse tomato under different irrigation conditions, with the coefficients of determination ranging from 0.88 to 0.98 and the index of agreement higher than 0.90. The seasonal cumulative ET_c of tomato for W1–W3 were 138.5–194.4 mm, of which 9.5–15.8% was consumed by E_s. Path analysis showed that the net radiation (R_n) was the dominant factor controlling the variations of T_r and ET_c during the growing seasons. The canopy coverage degree (K_cc) was the dominant controlling factor of E_s, while the temperature (T_a) was the primary limiting factor affecting E_s. This study can provide reference information for developing proper irrigation management in a greenhouse-grown tomato in the north cold climate regions.

The growth environment of corps requires necessary improvements by Chinese solar greenhouses with Pad–Fan Cooling (PFC) systems for reducing their high temperatures in summer. Although computational fluid dynamics (CFD) could dynamically display the changes in humidity, temperature, and wind speed in solar greenhouses, its computational efficiency and accuracy are relatively low. In addition, the use of PFC systems can cool down solar greenhouses in summer, but they will also cause excessive humidity inside the greenhouses, thereby reducing the production efficiency of crops. Most existing studies only verify the effectiveness of a single machine learning (such as ARMA or ARIMA) or deep learning model (such as LSTM or TCN), lacking systematic comparison of different models. In the current study, two machine learning algorithms and three deep learning algorithms were used for their ability to predict a PFC system’s cooling effect, including on humidity, temperature, and wind speed, which were examined using Auto Regression Moving Average (ARMA), Autoregressive Integrated Moving Average (ARIMA), Long Short-Term Memory (LSTM), Time Convolutional Network (TCN), and Glavnoe Razvedivatelnoe Upravlenie (GRU), respectively. These results show that deep learning algorithms are significantly more effective than traditional machine learning algorithms in capturing the complex nonlinear relationships and spatiotemporal changes inside solar greenhouses. The LSTM model achieves R² values of 0.918 for temperature, 0.896 for humidity, and 0.849 for wind speed on the test set. TCN showed strong performance in identifying high-frequency fluctuations and extreme nonlinear features, particularly in wind speed prediction (test set R² = 0.861). However, it exhibited limitations in modeling certain temperature dynamics (e.g., T6 test set R² = 0.242) and humidity evaporation processes (e.g., T7 training set R² = −0.856). GRU delivered excellent performance, achieving a favorable balance between accuracy and efficiency. It attained the highest prediction accuracy for temperature (test set R² = 0.925) and humidity (test set R² = 0.901), and performed only slightly worse than TCN in wind speed prediction. In summary, deep learning models, particularly GRU, offer more reliable methodological support for greenhouse microclimate prediction, thereby facilitating the precise regulation of cooling systems and scientifically informed crop management.

In the design of gel phase change composite wall materials for solar greenhouses, the alteration of material composition could directly affect the thermal performance of gel phase change composite wall materials. In order to obtain better suitable gel phasechange composite wall material for solar greenhouses, Na₂HPO₄·12H₂O-based gel phasechange materials with different content of ingredient (Na₂SiO₃·9H₂O, C₃₅H₄₉O₂₉, KCl, and nano-α-Fe₂O₃) were obtained via the Taguchi method and machine learning algorithms, such as Support Vector Regression (SVR), Random Forest (RF), and Gradient Boosting Trees (GBDT). The result shows that the GBDT is more suitable for the thermal performance optimization prediction of gel phase change composite wall materials, including time cooling (TC), latent heat of phase change (ΔH_m), supercooling degree (ΔT), and phase change temperature (T_m). The determination coefficient (R²) of time cooling (TC), latent heat of phase change (ΔH_m), supercooling degree (ΔT), and phase change temperature (T_m) by GBDT are 0.9987, 0.99965, 1, and 0.9995, respectively. The mean absolute error (MAE) coefficient percentage of supercooling degree (ΔT), phase change temperature (T_m), latent heat of phase change (ΔH_m), and time of cooling (TC) by GBDT are 0.32%, 0.25%, 0.17%, and 0.26%, respectively. The root mean square error (RMSE) of supercooling degree (ΔT), phase change temperature (T_m), latent heat of phase change (ΔH_m), and time of cooling (TC) by GBDT are 0.41%, 0.32%, 0.19%, and 0.35%, respectively. The optimal result predicted by GBDT is Na₂HPO₄·12H₂O + 5% Na₂SiO₃·9H₂O + 12% KCl + 0.2% Nano-α-Fe₂O₃ + 3% C₃₅H₄₉O₂₉, which was verified by experiments.