2014 Volume 34 Issue 1 Pages 22-32
The consistency of spectral data is essential in constellation-use of various hig-resolution satellite sensors for timely monitoring of crops and farmlands. The objective of this study was to evaluate the consistency of spectral measurements for rice paddy taken by multiple hig-resolution optical satellite sensors. We compared the reflectance spectra and four vegetation indices (NDVI, GRVI, EVI and SAVI) derived from satellite images with ground-based spectral data. The effects of atmospheric correction methods (6S, QUAC and FLAASH) were also examined. The Spectral Angle Mapper (SAM) values, as an index of the similarity, were 0.110, 0.111 and 0.312 in order of the QUAC, 6S and FLAASH, respectively. The reflectance spectra estimated by QUAC and 6S had a high degree of similarity with the ground-based spectral data. The Mean Relative Error (MRE) of NDVI, EVI and SAVI derived from satellite reflectance data against the ground-based indices was 20-40 %, whereas it was 80-90 % for the VIs derived from DN and radiance data. It therefore appears that hig-resolution optical satellite images from multiple sensors can be used for the consistent monitoring of crops and farmlands, as long as appropriate atmospheric-correction methods such as QUAC and 6S are applied.
