COLOR ADDITIVE VIEWING TECHNIQUES FOR SMALL-SCALE SOIL MAPPING IN AN AREA OF KARIMNAGAR DISTRICT, ANDHRA-PRADESH, INDIA

Multi-spectral imagery from Landsat relating to the Karimnagar distric t of Andhra Pradesh, in South India was analysed using colour additive viewing techniques. The results were correlated with an existing soil association map prepared from aerial photo-interpretation with adequa te ground truth. The four landscapes and their subdivisions that were identified on the soil map also could be identified on the Landsat map . An overall classification performance of 83.9, 82.4, 80.9 and 95.7 p er cent respectively has been achieved for rugged (R). undulating humm ocky (U), very gently sloping (V) and river alluvium (A) landscapes. T he best class performance was in the order of 95.7, 97.6. 81.8 and 95. 7 per cent and that of the lowest class performance of 50.5, 57.3, 70. 1 and 95.7 per cent for R, U, V and A landscapes, respectively. The ov erall combined performance of 85.7 per cent has been achieved with res pect to soilscape boundary delineations for all the four landscapes. T he Student's t-test of significance revealed that for units R4, U4, an d V2 the differences were significant and the rest were found to be no nsignificant. This study has demonstrated the usefulness of colour add itive viewing techniques in the analysis of Landsat-MSS data for small -scale soil mapping and the same could be used for the preparation of small-scale soil maps of the States and the country.