After calibration, GTa_CropGrowth simulated total fresh yield well as the difference between simulated and observed total fresh yield reduced from 23% to 0.3% in 2012. It was mainly improved by changing DMC from 0.05 to 0.04. The difference in 2013 and 2014 was 1.8% and 4.4% respectively.
After calibration, GTa climate modules simulated temperature in 2012 well with a correlation coefficient of 0.98. Simulation of CO2 concentration was largely improved (correlation coefficient from 0.20 to 0.98) by changing set-point.
However, simulation for Absolute Humidity was still poor with a correlation coefficient of 0.69.
CO2 injection was largely under-estimated for at least 34%, and gas consumption was largely over-estimated for at least 4.5 times more than that in reality.
The control of screen and window opening did not match practice in Venlow greenhouse. It was therefore poorly simulated.
Climate control options should be improved to better match the use of resources. Although GTa_CropGrowth is a simple model based on light use efficiency, applicability for prediction of crop growth is promising. With respect to prediction of greenhouse climate and resource use the applicability of the model is limited and further validation is recommended.