Characterization of brown planthopper damage on rice crops through hyperspectral remote sensing under field conditions

Field experiments were conducted to characterize the brown planthopper (BPH) (Nilaparvata lugens (Stål.) damage stress on rice crops through hyperspectral remote sensing. The BPH-damaged rice crop had higher reflectance in visible (VIS) and lower reflectance in near-infrared regions (NIR) of the electromagnetic spectrum compared with uninfested plants. Mean reflectance of the rice crop varied among different BPH damage levels in various wavebands, with the greatest variation in NIR (740–925 nm). Correlations between plant reflectance and BPH damage depicted four sensitive wavelengths, at 764, 961, 1201 and 1664 nm in relation to BPH stress on the rice crop. Three new brown planthopper spectral indices (BPHI) were formulated by combining two or more of these sensitive wavelengths. Some of the hyperspectral indices reported in the literature were also tested for their suitability to detect BPH stress on rice crops. Based on crop reflectance corresponding to the sensitive wavelengths, a multiple-linear regression model was developed (R²=0.71, RMSE=1.74, P<0.0001) and validated (R²=0.73, RMSE?=?0.71, P<0.0001) that would help to monitor BPH stress on a rice crop and to issue forewarnings to growers.