Development of a CFD crop submodel for simulating microclimate and transpiration of ornamental plants grown in a greenhouse under water restriction

Predictive models of soil-plant-atmosphere water transfers may be helpful to better manage water inputs to plants in greenhouses. In particular, Computational Fluid Dynamics appears to be a powerful tool to describe the greenhouse microclimate and plant behavior. Up until now, most models for potted plants grown in greenhouses were established for well-watered conditions. In this context, the aim of this work is to develop a specific submodel to simulate the distributed transpiration and microclimate during plants grown in pots inside greenhouses under water restriction conditions. A 2D transient CFD (Computational Fluid Dynamics) model was implemented and user-defined functions were adapted to take account of the crop interactions with the climate inside the greenhouse. The crop was considered as a porous medium and specific source terms for transpiration and sensible heat transfers were added. A specific submodel was also implemented to calculate the substrate water content based on the water balance between irrigation and transpiration. Particular care was paid to the modeling of stomatal resistance. In order to obtain the input data and to validate the CFD simulations, an experiment was conducted over 16 weeks inside a greenhouse equipped with New Guinea impatiens ornamental plants grown in containers on shelves. Both well-watered and restriction conditions were analyzed. The results of the CFD simulations showed the ability of the model to correctly predict transpiration, air and leaf temperatures as well as air humidity inside the greenhouse for both water regimes. Different irrigation scenarios were then tested, progressively reducing the water supply by providing a lesser amount of water than the growing media water capacity. The simulations made it possible to assess the model response to different irrigation regimes on plant transpiration, usual growing media water potential and climate distribution inside the greenhouse. The tests also showed that the water supply could be reduced by 20% without significantly impacting the transpiration rate and, therefore, potential plant growth. The CFD model could thus be useful to test different irrigation scenarios and better manage water inputs.