Retrieval of leaf water content for maize seedlings in visible near infrared and thermal infrared spectra

It is very important to quantify water content of leaf for monitoring vegetation using hyperspectral remote sensing. Taking maize seedlings as example, spectra in visible near infrared region (VNIR, 0.4-2.5μm) and thermal infrared region (TIR, 8-14μm) were measured simultaneously. Narrow-band simple ratio water index (SRWI), simple difference water index (DWI) and normalized difference indices (NDWI) were calculated respectively for both VNIR and TIR spectra. According to correlation analysis and model validation, it is found that (1) the VNIR proved highly sensitive spectral region and hold promise for the estimation of leaf water content, and (2) SRWIVNIR is the most accurate index for the retrieval of leaf water content of maize seedlings (R
²
= 0.78, RMSE = 2.52%).