Articles
SOME REMARKS CONCERNING THE USE OF MODELS IN GREENHOUSE CLIMATE RESEARCH
Consequently considerable effort has been put into the field of greenhouse climate research.
However, progress in this field is hampered by the difficulty of generalizing the results obtained experimentally.
This difficulty is caused by the complex relationships and interactions that exist between the factors that can be manipulated and the simultaneous reactions that are observed.
Furthermore, the situation is complicated by the great discrepancy between the response time of the observed crop performance (weeks) and the response time of the climate control required for adequate adaptation to variations in the weather conditions (hours).
In many cases manipulation of a factor gives rise to positive effects with one process and simultaneously negative effects with another process.
Hence optimization is important in many aspects of greenhouse climate research.
A major obstacle in this optimization is the lack of quantitative knowledge necessary to evaluate and predict the effect of climate on crop performance.
It is likely that in this field models can play an important role as they have done already for example in the technical sciences.
A model is usually a simplified representation of a system.
Models can be applied to the solution of a wide range of problems, but not every one.
In the case of greenhouse climate research there are problems where models can probably contribute little to their solution, for example fruit-set of tomato at low night-temperature, poor quality of flowers at high temperatures and low light intensity and tip-burn with lettuce.
These problems are examples where empirical knowledge forms the frame-work in which optimization through quantitative models can play a role.
Models used for greenhouse climate-research may be divided in two categories:
- Empirical models in which statistics are used to describe the relationship between environment, crop response and energy costs in some empirically obtained equations (black-box models).
- Mechanistic models where the system is described on the basis of physical and physiological theory (explanatory models).
The development of empirical models usually requires less time than the construction of mechanistic ones.
Furthermore, these models are simpler and the predictive value is less affected by erroneous assumptions or factors and processes that have not been taken into account.