Is the Penman–Monteith model adapted to predict crop transpiration under greenhouse conditions? Application to a New Guinea Impatiens crop

Abstract

Reducing water consumption in greenhouses is one of the main concerns for growers. This implies better knowledge of the transpiration process in order to adapt water inputs to plant needs. Several simple models have been developed to estimate transpiration. The Penman–Monteith (PM) model and the Direct Method (DM) model are two of the most widespread transpiration models. Yet the PM model was primarily developed for crops grown in open field conditions; its validity in the case of sheltered crops is only seldom questioned. Moreover, the PM model is rarely compared to the DM model, although this could be a simple way to quantify its performance.The aim of the study is thus, first, to analyse the ability of the classical PM model to assess the transpiration rate inside a greenhouse with a New Guinea Impatiens crop in order to identify which modifications should be introduced to improve its efficiency. Comparison of the PM model with measurements reveals severe discrepancies. A detailed assessment of the assumptions underlying the PM model revealed that the location of the climatic parameters used to implement the model is a key factor that affects the results. The present study demonstrates that the temperature and humidity considered in the PM model should be taken inside the crop and not above the crop, leading to a so-called PM-Like (PML) model.Results show that the PM-Like and the DM models, in a one-layer configuration, give similar results and are in agreement with actual measurements. The capacity of the PM-Like and the DM models to investigate the vertical distribution of the transpiration inside the crop is also discussed. After dividing the canopy into two layers, the PML and DM models were applied and validated against experimental data. Even if such a procedure does not really improve the quality of the results compared to the one-layer models, it provides innovative information on the vertical distribution of the transpiration, which could have applications in CFD modelling.