Detection Of Fruit Tree Water Status In Orchards From Remote Sensing Thermal Imagery
1P. J. Zarco-Tejada, 1V. Gonzalez-Dugo, 2J. Girona, 1E. Fereres, 2J. Bellvert
1. Instituto de Agricultura Sostenible (IAS), Consejo Superior de Investigaciones Científicas (CSIC), Córdoba, Spain
2. Water Use Efficiency programme, Institut de Recerca i Tecnologia Agroalimentàries (IRTA), Lleida, Spain
In deciduous fruit trees there is a growing need of using water status indicators for scheduling irrigation and adopt regulated deficit irrigation (RDI) strategies taking into account spatial variability of orchards. RDI strategies have been successfully adopted for many fruit trees as a means for reducing water use and because yield and quality at harvest are not sensitive to water stress at some developmental stages. Although water status is generally monitored by measuring tree water potential, other alternative methodologies have been implemented to determine plant water status at large scale. Canopy temperature has been a useful technique to identify and assess plant water status in large areas. The study reported here was conducted to evaluate the feasibility of detecting plant water status of a peach orchard using remotely sensed canopy temperature obtained from airborne thermal images.
One approach for quantifying water stress by using canopy temperature is the `Crop water stress index´ (CWSI). In 2012, CWSI was empirically developed by installing infrared temperature sensors over well-watered peach trees transpiring at potential rate. The CWSI algorithms were also validated with results from high resolution airborne thermal imagery, and `ground truth´ with leaf water potential (Ψ
L) measurements at different phenological stages. Six flights were conducted over a 2-ha peach orchard using an aircraft carrying a thermal sensor. The thermal sensor has a resolution of 640x480 pixels, equipped with a 24.5 mm f1.0 lens, connected to a computer via USB 2.0 protocol. Flight altitude was 150 m and images obtained had 30-cm spatial resolution, enabling the capture pure vegetation pixels and excluding soil, background targets, and shadows. Relationships between remotely CWSI and measured Ψ
L had a coefficient of determination (R
2) ranging from 0.58 to 0.75. In 2013, nine flights were conducted in the same peach orchard from April to August and Ψ
L was remotely estimated by using the CWSI algorithms and relationships developed during the previous season. Therefore, Ψ
L of each individual peach tree was determined from high resolution thermal remotely sensed images. Remotely sensed Ψ
L maps were generated to identify spatial variability of water status within the peach orchard. However, as irrigation scheduling depends on the irrigation sector design, information of the averaged Ψ
L of each irrigation sector was also provided as a useful tool for irrigation purposes. Implementation of this technology for scheduling irrigation in fruit tree orchards will allow the adoption of RDI strategies and in determining how much water to apply in each irrigation sector to reach the same plant water status.
