Use of thermal imagery to detect water stress during berry ripening in Vitis vinifera L. 'Cabernet Sauvignon'

The combination of uncertain and low water availability, the increasing cost of water for irrigators and the use of water as a lever to control yield and fruit composition has driven a need for improved irrigation scheduling in southern Australia. Remotely sensed canopy temperature assessment offers a potentially reliable and cost effective means for improving irrigation scheduling. Crop water stress index (CWSI), based on canopy temperature was assessed during berry ripening for ‘Cabernet Sauvignon’ exposed to deficit irrigation in central Victoria, Australia. CWSI was calculated from canopy temperature (T(c)), a dry reference surface (T(d)) and a wet reference surface (T(w)) using the method of Jones et al. (1999). An infrared camera was used to measure the temperature of grapevine canopies and reference surfaces. Measurements were taken of grapevines irrigated at intervals of 3-4 days (wet) or not irrigated for 18 days (dry). CWSI calculated from either northern (sunlit) or southern (shaded) facing canopies identified differences in grapevine water status on all five occasions except on a day of high VPD (5.8 kPa). On this day, and at time of measurement stem water potential of both irrigated and non-irrigated vines was quite low (<-1.5 MPa). The standard deviation of T(c) was greater for southern compared to northern facing canopies and did not vary with irrigation treatment suggesting that the use of canopy temperature of partly shaded canopies is unreliable for detecting the onset of water stress. A software program was developed and tested to identify sunlit leaves to improve repeatability and reliability of extraction of T(c) data from thermal images. The software (auto) calculated T(c) within 1°C of that of the labour intensive manual method (manual) and therefore, shows great promise as a tool for rapid image processing of thermal images.