This work explores the possibilities of distinguishing saline from gypsifer
ous soils, using remote sensing data, especially the Landsat TM sensor, wit
h it's seven spectral bands which was practically useful. Using supervised
image classification, makes it feasible to differentiate gypsiferous from s
aline soils and from other soils as well. This study demonstrates that the
best way to differentiate between gypsiferous and saline soils is to use th
e thermal band of the Landsat TM sensor. It was shown by using the TM therm
al band that mapping of the soils containing gypsum can be done in a relati
vely fast and accurate way. The final output can easily be stored in a GIS
for further practical applications. (C) 1998 Elsevier Science B.V. All righ
ts reserved.