Crop growth models for greenhouse climate control.

In the last 20 years various different models have been developed to describe and explain crop growth under various conditions. These explanatory models have mainly been used for research to elucidate the often quite complex relations between crop environment and yield. Their practical significance has been mainly through this improved understanding, rather than the direct use of their output. Models, however, basically do have great potential for practical use in agriculture in general (Penning de Vries, 1983) and in horticulture in particular (Challa, 1985; 1988). In general, their use (which is still very limited) is in the field of decision-making at the three levels of farmer's involvement that are usually distinguished, depending on the decision horizon (Table 8). Spedding and van Keulen give examples of the use of models for strategic decisions (Chapters 13 and 15, respectively). Penning de Vries (1983) mentions models being used for decision support at the operational level in the case of pest management and irrigation. Decisions about process control, an important item in greenhouse culture, are usually considered within the framework of the operational decisions. In my opinion, however, process control should be considered as a special category (Table 8). The principle difference between operational decisions and process control is that the latter lacks human interference. The operator checks the process from time to time and may adjust the control procedure, but the actual control is delegated to the control system. The main reason we need to use models to control biological systems is because of the difficulty of measuring the relevant processes directly and the inherent need to interpret on-line measurements in the terms desired. In the second place it is often quite difficult to predict a required action in order to obtain a desired reaction. Process control, as described here, is characteristic for protected cultiva-