Computational water stress indices obtained from thermal image analysis of grapevine canopies

Thermal imaging of crop canopies has been proposed more than a decade ago as a sensitive methodology to determine the water status of different crops. This paper describes the development of a semi-automated and automated methodology using MATLABA (R) programming techniques to analyse the infrared thermal images taking into consideration the pitfalls pointed out previously in the literature. The proposed method was tested in an irrigation reduction and recovery trial for Chardonnay in the 2010-2011 season and in the 2009-2010 season from seven varieties in field conditions. There was a clear separation (assessed by principal component analysis) between control and recovery compared to stress treatments using leaf area index (LAI), stomatal conductance, stem water potential and indices derived from canopy temperatures measured by infrared imaging. High and significant correlations were found between canopy temperature indices and other measures of water stress obtained in the same vines that were independent of LAI. Furthermore, a fully automated analysis method has been proposed using ancillary weather information obtained from the same locations of infrared thermal images. This paper is a first step towards automation of infrared thermography acquisition and analysis in the field for grapevines and other crops.