•  
  •  
 

Abstract

Though the view angle range of SPOT HRV is smaller as compared to those of NOAA AVHRR, EOS MODIS/MISR and ADEOS POLDER, numerous researchers had showed significant bi-directional effect of ground reflectance for decades. Currently, there's no existing normalization method of bidirectional reflectances for multitemporal SPOT images. In this study, a normalization scheme for multi-angular SPOT data is proposed. This scheme will correct both atmospheric and bidirectional effect for short-term multi-temporal SPOT data. Five SPOT images scanned within twelve days are used. The results show that the variation of trees reflectance to viewing zenith angle is very significant. The proposed normalization scheme can satisfactorily correct viewing zenith angle effect of bidirectional reflectances of trees. Normalized vegetation indices are also developed and successfully shown to reduce the viewing angle effect of Normalized Difference Vegetation Index (NDVI). Soil Adjusted Vegetation Index (SAVI) is found to be less dependent on viewing zenith angle than NDVI and normalized VIs in the azimuthal plane away from the principal plane. This may be due to the effectiveness of SAVI in the reduction of viewing zenith angle effect induced by the soil and shadow and the error of the atmospheric correction model.

Download
COinS