Soft-sensing for leaf water potential based on micro-environment factors of plant

Leaf water potential is the best parameter of estimating plant water status, evaluation from Penman-Monteith transpiration formula or retrieval from remote sensing data has complex calculations, too many parameters, poor transplantations and high costs. This paper selects accessible micro-environment factors of plant as auxiliary variables, and establishes a leaf water potential soft-sensing model with RBF neural network. Simulation result shows that this model is simple and practical, and has higher accuracy. It is one of effective methods estimating plant water status on line.