PARAMETER ESTIMATION AND SELECTION OF A GREENHOUSE NATURAL VENTILATION MODEL AND ITS USE ON AN ENERGY BALANCE MODEL TO ESTIMATE THE GREENHOUSE AIR TEMPERATURE

Predicting the ventilation rates of a greenhouse is very important, since natural ventilation is the main climate control method on the majority of the Portuguese horticultural greenhouses. Ventilation by stack effect or due to wind is usually achieved by opening the greenhouse windows. Quantifying the ventilation fluxes is a complex task, subject to great data variability. In this work we present a simplified model to obtain the natural ventilation fluxes in an unheated horticultural greenhouse with a PE film cover located in Lisbon. A model was selected from a set of several models previously obtained in the south of France with comparable greenhouses. The coefficients of these models were estimated based on 59 measured ventilation rate sets of data, obtained with a tracer gas technique using nitrous oxide, on the final phase of a protected tomato crop. Afterwards we proceeded with the selection of the most parsimonious model, by comparing the estimated and observed ventilation rates, using the Akaike information criteria. This ventilation model was later used on a simplified energy balance model to estimate the greenhouse air temperature, based on the external climate conditions and with an hourly time step. The energy balance coefficients were obtained using data from one experiment that took place on the same site, with a spring tomato crop, from March until July. The energy balance model was tested with data from a similar experience that took place two years earlier on the same greenhouse.