-

In this study, PC technique was used to reduce the number of spectral bands or spatial variables in a data set by finding their linear combinations. To apply the PCA for spectral/spatial data set, Landsat TM data recorded from 5 different areas in Central Iran, and 18 soil varaibales were used. The result of PCA transformation for TM bands revealed the importance of PC 1
in soil studies, PC2 and or PC3 for vegetation investigations. The results of
PCA for Landsat TM and soil data showed that the TM data of 7 bands and
18 soil variables were mainly compressed to just three PCs that describe more than 90% and 550/0 of spectral and spatial information respectively. Based on the obtained results we may conclude that PCA can be applied to
different sources of spectral/spatial data, for a better establishment of sampling plan and save in money and time.