Fluorescence, PRI and canopy temperature for water stress detection in cereal crops

Narrow-band multispectral remote sensing techniques and thermal imagery were investigated for water stress detection in cereal crops. Visible and near infrared AISA Eagle (Specim, Finland) and thermal AHS-160 (Sensytech Inc., USA) imageries were acquired with an airborne survey on a farm-level experimental site where maize (Zea mays L.) and sorghum (Sorghum bicolor L.) were grown with three different irrigation treatments. Vegetation biophysical and eco-physiological measurements were collected concurrently with the airborne campaign. Leaf fluorescence yield (ΔF/Fₘ′) resulted to be a good indirect measure of water stress. Therefore, ΔF/Fₘ′ measurements were compared against remotely sensed indicators: (i) the Photochemical Reflectance Index (PRI), (ii) the sun-induced chlorophyll fluorescence at 760nm (F₇₆₀), retrieved by the Fraunhofer line depth method and (iii) the canopy temperature (TC) calculated decoupling soil and vegetation contributions. TC was related to ΔF/Fₘ′ with the highest determination coefficient (R²=0.65), followed by PRI₅₈₆ (reference band at 586nm) (R²=0.51). The relationship with F₇₆₀ was significant but weaker (R²=0.36). The coefficient of determination increased up to 0.54 when pigment concentration was considered by multiplying ΔF/Fₘ′ and chlorophyll content, confirming the close relationship between passive fluorescence signal, pigment content and light photosystem efficiency. PRI₅₈₆, F₇₆₀ and TC maps were produced in maize and sorghum plots. The differences in the average values of PRI₅₈₆, F₇₆₀ and TC extracted from the plots with different water treatments showed that water treatments were well discriminated in maize plots by the three remotely sensed indicators. This was confirmed by the visual observation of the PRI₅₈₆, F₇₆₀ and TC maps, while in sorghum plots, F₇₆₀ and TC appeared more sensitive to water stress compared to PRI₅₈₆.