Methodological improvements when using thermal imagery in viticulture

Infrared thermography (IRT) is a promising technology in viticulture using absolute or relative techniques. Methodological inconsistencies exist in the literature. Standardisation of measurements must be provided to growers before accurate deficit irrigation scheduling is implemented. Four field experiments are reported. Ambient temperature and relative humidity were similar when measured within a vineyard block and at the end of a row, but not wind speed. Wet and dry reference leaves were thermally stable 30 s after wetting and for 60 (exposed) to 90 s (shaded). The computed thermal time constant were influenced by wind puffs and ranged from 4 s to 14 s. 93% of triplicate thermographs of identical vine faces captured 2 min apart displayed a significant difference in mean temperature in all possible pairs. The confidence interval was however lower than the thermal camera accuracy error. Thermographs of heterogeneous canopies were segmented using digital photographs and the temperature range of vertical reference leaves. Most thermographs had over 40% of misclassified pixels. Finally, measurement errors of the infrared thermometer were shown to induce large error when using leaf temperature to evaluate stomatal resistance. In Conclusion, absolute canopy temperature varies over short periods of time, particularly when wind speed increases. This and the accuracy error of thermographers induce errors when computing stomatal resistance. Wind speed measurements should be taken close to the target canopy. Thermographs of reference leaves should be taken 30s after wetting. Heterogeneous canopies cannot yet benefit from thermal imagery for irrigation scheduling.