Abstract
Global climatic changes increase areas that are influenced by drought. Remote sensing based on the spectral characteristics of reflected light is widely used to detect the action of stressors (including drought) in plants. The development of methods of improving remote sensing is an important applied task for plant cultivation. Particularly, this improvement can be based on the calculation of reflectance indices and revealing the optimal spectral bandwidths for this calculation. In the current work, we analyzed the sensitivity of broadband-normalized difference reflectance indices and RGB indices to the action of soil drought on pea and wheat plants. Analysis of the heat maps of significant changes in reflectance indices showed that increasing the spectral bandwidths did not decrease this significance in some cases. Particularly, the index RI(659, 553) based on the red and green bandwidths was strongly sensitive to drought action in plants. The normalized red-green index (NRGI), which was the RGB-analog of RI(659, 553) measured by a color camera, was also sensitive to drought. RI(659, 553) and NRGI were strongly related. The results showed that broadband and RGB indices can be used to detect drought action in plants.