Extraction of soil characteristics and monitoring of agricultural land use in the semi-arid tropics of India using remote sensing

In the semi-arid tropics of India, agricultural land use has been extended to meet the demand for food to match the increase of the population. A considerable portion of land used for agriculture in this area is based on rain fed cultivation and its extension has been influenced by the changes in physical conditions. It is important to analyze the conditions of the land used and to monitor annual changes of cropping in terms of sustainability of land use. This study aims at the development of methods to estimate the land conditions suitable for agricultural purposes and to monitor the temporal changes of land use using remote sensing techniques. The study includes the processing of remote sensing and geographic data by a computerized system and field verification by measuring the spectral reflectance of various types of surface conditions. As a result, two indices, the redness index (RI) and brightness index (BI) calculated from Indian Remote Sensing Satellite (IRS) multi-spectral data, were found to be applicable to estimate the land suitability especially on black soils (Vertisols) as well as to discriminate black soils from red soils (Alfisols). In order to discriminate the cropped area from other land uses and to monitor the annual changes of cropped area, the normalized vegetation index (NDVI) was employed as discriminator. NDVI value calculated from IRS data obtained during the year can be corrected by employing the representative profile of seasonal NDVI for each land use category. It was observed that the ratio of agricultural land use in the post-rainy (Rabi) season in the objective area had almost doubled during the period between 1989 and 1996 and showed a positive correlation with the amount of rainfall around the sowing season. It was also found that the area more suitable for agriculture showed a higher rate of increase of cropped area during this period