Effect of cirrus cloud on normalized difference Vegetation Index (NDVI) and Aerosol Free Vegetation Index (AFRI): A study based on LANDSAT 8 images

The present study is an attempt to identify the influence of cirrus cloud on NDVI and AFRI values and to check the scope of replacing NDVI with AFRI in cirrus affected satellite images. Multi-spectral channels of LANDSAT-8 satellite image collected for two different dates of acquisition are used in the current study. Reflectance values of cirrus band are used for correcting red and near infra-red spectral channels based on standard cirrus correction algorithm. The NDVI values obtained after cirrus correction is found to be significantly more than that of NDVI values without cirrus correction. In the case of AFRI(2.1m), the difference in values between before and after cirrus correction is found to be considerably lower than that of NDVI variation. These outcomes show the suitability of AFRI(2.1m) over NDVI for cirrus affected satellite images. The results of analysis also reveal that NDVI difference is directly proportionate with optical thickness and ice particle size distribution of cirrus clouds.