STATIC REGRESSION MODELS FOR PLANNING GREENHOUSE PRODUCTION

For planning greenhouse production a regression model at high abstraction level was developed.
It is based on experiments in greenhouses with successive plantings during the season and varied set points for temperature, in other experiments additionally with different CO₂ -concentrations.
By successive harvestings growth curves were derived (lettuce transplants in fig. 1) and parameterized by the Feldmann growth function.

Using a non-linear regression function with multiplicative terms (fig. 2), response surfaces were calculated which allow to derive mean growth rates and thereby growth periods as a function of irridiance and temperature (fig. 3 for lettuce transplants with different set points for heating) and possibly CO₂-concentration, as a function of weight at transplanting, of weight at harvesting and of the season with special or long term weather conditions.

This model allows to derive:

• Dates of emergence or planting and harvesting, respectively, as a function of the season and the set points for heating (fig. 3) and possibly CO₂ -concentration, based on mean long term weather conditions or the weather of a particular year.

Examples for lettuce transplants in fig. 3: Planting date March 15., using set points of 14/10° C (day/night) emergence at beginning of February, for set points of 2/2° C at the middle of November.
Or with emergence middle of November transplant size is reached using set points of 14/10° C at beginning of February and for set points of 2/2° C at middle of March.

• The variability of the growth period as a function of the weather conditions of a particular year and the set point of heating (fig. 4) or other growth promoting procedures.
  This implies the risk of planning.

Symposium on High Technology in Protected Cultivation

H. Krug, H.-P. Liebig

https://doi.org/10.17660/ActaHortic.1988.230.56