Soil forms one of the vital natural resources in that it is useful for agricultural and engineering purposes as the Chief plant supporting material for producing food and fibre and as construction material for raising buildings respectively. Soil erosion caused by the action of wind, water, ice and gravity over a terrain requires, therefore, prompt detection and implementation of proper preventive/remedial measures to conserve the soil resource. Soil inventory to assess its quantity and quality by the conventional field survey methods consumes considerable manpower with time and money and the remote sensing technology offers in this context, a powerful tool to achieve the objectives of producing meaningful regional soilscape maps quickly with less amount of survey inputs.

The Landsat frame under study was of February 27, 1973 and the digital analysis was done at LARS using LARSYS software package in March - April, 1977, followed by post field check in September, 1977. While carrying out the original and limited objective of doing a quantitative exercise for assessment of Bhavani dam waterspread area and its irrigated lands by digital analysis of the Landsat scanner data, it was incidentally found out that a long stretch of land lying for length of about 70 miles and for a width of about 2 to 5 miles was revealing very light tone on the imagery. The prefield presumption for the occurrence of possible soil erosion was confirmed by post field observations. The predominant pegmatite rock associated with country rocks in the area was found undergoing disintegration due to weathering phenomenon and the quartz fragments and the mica flakes in the washed debris with high albedo revealing light tone on the imagery.

The paper discusses, among other things, the methodology adopted in the digital analysis related to soil erosion studies using the LARSYS unsupervised clustering algorithms.

Date of this Version