Reliability of greenhouse climate dynamic models by uncertainty and sensitivity analyses, calibration and evaluation

The improvement of controlled environment horticulture systems such as greenhouses hinges on reliable dynamic mathematical models. To make models trustworthy not only the postulation of their structure is necessary but also performing uncertainty and sensitivity analyses, parameter estimation and model evaluation. In this work the benefits of such overall systems modeling procedure are demonstrated by not only developing a new two-state greenhouse climate model including air temperature and humidity; but also, by carrying out a Monte Carlo uncertainty analysis. Furthermore, a global sensitivity analysis (SA) named density-based PAWN method was applied to choose the most influential model parameters on the state-variables. Next, the model was calibrated by the classical nonlinear least squares procedure which is a local search method. Finally, the model was evaluated by using an independent data set. The statistical measures MSE, RMSE, MAE and model efficiency (EF) were used during model calibration and evaluation. According to uncertainty analysis both state variables were estimated accurately, distributions turned out highly skewed with a large kurtosis value, namely, quite different from a normal distribution. Because of this a variance-based global sensitivity analysis was not possible. From a density-based SA, the most influential model parameters were the infiltration coefficient, the heat transfer coefficient of the soil, the leaf boundary layer resistance, the constant soil density, and physical properties of the greenhouse (cover and ground area). Only the three most influential model parameters were estimated. For temperature RMSE values on calibration and evaluation were 0.97 and 1.26, respectively. Whereas for humidity ratio RMSE values were 1.73 and 1.76, respectively; and for relative humidity those values were 12.3 and 15.6, respectively.