A strategy for greenhouse climate control, Part II: Model validation

The present project tested a model predicting the dynamic ambient greenhous e air conditions maximizing a tomato crop yield value less the energy cost. For simplification, this yield value less energy cost is referred to as ne t profits. Net profits were equated to crop yield value, computed from the dynamic greenhouse conditions (temperature, incident radiation, CO2 level a nd relative humidity), less the costs of heating, dehumidification and CO2 injection. The physical parameters describing a Venlo-type glass greenhouse located in Quebec City, Canada, were measured to describe its heat and mas s (CO2 and water vapour) transfers and test the model. The model was used t o predict net profits for 2 months of tomato production. The measured value s were compared to that calculated by the sub-models (transpiration rate an d tomato yield) and the model itself. The sub-models and model proved to be accurate within a 3% error when used to predict crop yield and net profits for periods of 1 week or longer. The model was found to be especially sens itive to exterior temperature, affecting heating costs but not yield, then incident radiation reducing heating costs and increases yield through trans piration, and finally, relative humidity affecting crop yield and dehumidif ication costs. (C) 2001 Silsoe Research Institute.