By using the AVIRIS NG Hyperspectral data, analyzed the relationship between the
soil spectral reflectance and various degree of salt affectedness and waterlogged soils from
the soil sampling sites from Nagarjuna Sagar Left Bank Command area, Nalgonda district,
Telangana, India. The individual AVIRIS NG Hyperspectral image scenes were processed
with ENVI 5.3 software geocoded, and mosaicked, and the data is subjected to MNF,
performed PPI, n-D visualizer, and classification (mapping) for salt-affected and waterlogged
soils was attempted using the Spectral Angle Mapper algorithm. Out of the total classified
area normal, slightly saline-sodic, moderately saline-sodic, severely saline-sodic, and
waterlogged soils occupy 67.9, 17.0, 3.42, 7.27 and 0.04%, respectively. The Pearson
correlation studies showed that 1830, 1850, 1930, 1935, and 1940 nm wavelengths
significantly showed a negative correlation with EC, ESP, and CEC. The PCR model showed
the possibility of retrieval of EC, ESP, and CEC more accurately. The SAM classification for
AVIRIS NG showed a producer accuracy percentage of 76.4- 88.4 and a user accuracy
percentage of 77.4-87.9.