This paper experimented with a methodology of machine learning modelling using virtual samples generated by fast CFD (Computational Fluid Dynamics) simulations in order to predict the greenhouse natural ventilation. However, the output natural ventilation rates using fast two-dimensional (2D) CFD models are not always consistent with the three-dimensional (3D) one for all the scenarios. The first contribution of this paper is a proposed comparative modelling methodology between two-dimensional and three-dimensional CFD studies, regarding its validity, especially when buildings are in rows. The results show that the error of the ventilation rate prediction could exceed 50%, if 2D models are not properly used. Subsequently, in those scenarios where the 2D and the 3D models had equal accuracy, nearly one thousand samples were generated using fast 2D CFD simulations to train a natural ventilation rate regression tree model. This model is efficient to deal with the combined effect of wind pressure and thermal gradients under various vent configurations, with only four necessary inputs. In addition, by analyzing the wind speed distribution contour of the outdoor wind field around the greenhouse rows, the optimal wind speed-measuring locations were determined to eliminate interference for predicting the natural ventilation rate.

This paper experimented with a methodology of machine learning modelling using virtual samples generated by a fast CFD (Computational Fluid Duy) simulations, in order to predict the greenhouse natural ventilation. However, the output natural ventilation rates using fast two-dimensional (2D) CFD models are not always consistent with the three-dimensional (3D) one for all the scenarios. The first contribution of this paper is a proposed comparative modelling methodology between two-dimensional and three-dimensional CFD (Computational Fluid Dynamics) studies, regarding its validity, especially when buildings are in rows. The results show that the error on the ventilation rate prediction could exceed 50%, if 2D models are not properly used. Subsequently, in those scenarios where the 2D and the 3D model had equal accuracy, nearly one thousand samples were generated using fast 2D CFD simulations to train natural ventilation rate regression tree-model. This model is efficient to deal with the combined effect of wind pressure and thermal gradients under various vent configurations, with only four necessary inputs. In addition, by analyzing the wind speed distribution contour of the outdoor wind field around the greenhouse rows, the optimal wind speed measuring locations were determined to eliminate interference for predicting the natural ventilation rate.

Greenhouse microclimates exhibit substantial spatial heterogeneity. Conventional evaluations based on average measurements cannot adequately describe the environmental conditions experienced by crops. This study developed a spatial frequency analysis method to quantify long-term temperature and relative humidity heterogeneity. This method was coupled with cucumber yield and downy mildew models to evaluate the biological consequences of environmental variability. Environmental conditions in four representative greenhouse structures were monitored using distributed sensor networks under different weather conditions. The proposed method successfully identified persistent environmental hotspots. The spatial distribution of environmental heterogeneity differed among greenhouse structures, whereas solar radiation primarily controlled its intensity. Long-season monitoring data indicate that the maximum spatial variations in temperature and relative humidity within the brick-wall solar greenhouse, the assembled greenhouse, the glass greenhouse, and the plastic multi-span greenhouse reach up to 10 °C and 46%, 8.5 °C and 46%, 12 °C and 50%, and 7 °C and 32%, respectively. The assembled solar greenhouse and plastic greenhouse exhibited higher environmental uniformity, while the brick-wall solar greenhouse and glass multi-span greenhouse showed pronounced spatial gradients. Environmental heterogeneity resulted in significant within-greenhouse differences in simulated cucumber yield, with the smallest variation occurring in the assembled solar greenhouse (5.3%) and the largest in the glass multi-span greenhouse (39.2%). Disease simulations further revealed clear spatial aggregation of cucumber downy mildew risk, with the southern region of the solar greenhouse exhibiting 17–67% higher cumulative infection risk than other positions. This study establishes an integrated framework linking greenhouse environmental heterogeneity with crop productivity and disease risk, providing a practical basis for spatially differentiated precision greenhouse management and greenhouse structural optimization.

Greenhouse microclimates exhibit substantial spatial heterogeneity, whereas conventional evaluations based on average measurements cannot adequately describe the environmental conditions experienced by crops. This study developed a spatial frequency analysis method to quantify long-term temperature and relative humidity heterogeneity and coupled it with cucumber yield and downy mildew models to evaluate the biological consequences of environmental variability. Four representative greenhouse types were monitored using distributed sensor networks under different weather conditions. The proposed method successfully identified persistent environmental hotspots. Greenhouse structure determined the spatial distribution of environmental heterogeneity, whereas solar radiation primarily controlled its intensity. Long-season monitoring data indicate that the maximum spatial variations in temperature and humidity within the brick-wall solar greenhouse, the assembled greenhouse, the glass greenhouse, and the plastic multi-span greenhouse reach up to 10℃ and 46%, 8.5℃ and 46%, 12℃ and 50%, and 7℃ and 32%, respectively. The assembled solar greenhouse and plastic greenhouse exhibited higher environmental uniformity, while the brick-wall solar greenhouse and glass multi-span greenhouse showed pronounced spatial gradients. Environmental heterogeneity resulted in significant within-greenhouse differences in simulated cucumber yield, with the smallest variation occurring in the assembled solar greenhouse (5.3%) and the largest in the glass multi-span greenhouse (39.2%). Disease simulations further revealed clear spatial aggregation of cucumber downy mildew risk, with the southern region of the solar greenhouse exhibiting 17-67% higher cumulative infection risk than other positions. This study establishes an integrated framework linking greenhouse environmental heterogeneity with crop productivity and disease risk, providing a practical basis for spatially differentiated precision greenhouse management and greenhouse structural optimization.

Grape white rot is an important fungal disease that affects grapes worldwide. Recently, Coniella vitis has replaced Coniella diplodiella as the dominant pathogen causing grape white rot in China. However, its key epidemiological parameters remain undefined, hindering the development of specific disease management strategies. Thus, the effects of temperature and nutrient conditions (culture media) on the mycelial growth of C. vitis, and the combined effects of temperature and wetness duration on conidial germination were examined in this study. The results showed that the mycelial growth of C. vitis was optimal at 20°C–30°C under organic-rich nutrient conditions. On water agar, conidia germinated across a temperature range of 10°C–35°C (optimum 20°C–30°C) and required wet periods longer than 2 h. Nonlinear models described these processes with estimated optimal temperatures of 27.3°C for mycelial growth and 23.9°C for conidial germination (R² > 0.89). This study revealed fundamental differences in the responses of C. vitis to temperature and wetness duration compared with C. diplodiella, providing key information for refining existing white rot prediction models and supporting the transition from calendar-based to risk-based precision fungicide applications.