MONITORING OF SALT-AFFECTED SOILS OF THE INDO-GANGETIC ALLUVIAL PLAINS USING PRINCIPAL COMPONENT ANALYSIS

The principal component analysis (PCA) of spaceborne multi-spectral da ta enables data compression and helps to delineate certain terrain fea tures of interest otherwise indiscernible. The PCA has also been found to be a powerful tool for change detection using temporal spaceborne multi-spectral data. Principal component analysis of Landsat MSS data acquired during February 1975 and March 1992 over part of the Indo-Gan getic alluvial plain covering parts of Etah, Mainpuri, Aligarh, and th e Agra districts of Uttar Pradesh, was performed after digitally co-re gistering and merging them on a MicroVAX-based ARIES-III DIPIX system. An overall significant difference in the brightness and greenness was observed during this 17-year period. However, no precise clue regardi ng temporal variation in the salt-affected soils could be observed in PC images. This may be attributed to the very concept of PCA, which us es the spectral response pattern of the entire scene contrary to the K auth and Thomas transform wherein soil and vegetation reflectances are used for generating brightness and greenness images.