Remote Sensing Database Preparation from Irs Data and Lulc Change Monitoring over Indo-gangetic Basin

This paper describes the methodology developed and implemented for Remote Sensing DataBase (RSDB) preparation from multitemporal and multi-sensor data from Indian earth observation satellites. The characteristics of the RSDB prepared over parts of IndoGangetic Basin and Central India (IGB&CI) are summarised. An approach for extraction of long term (1988-89 to 2004-05) Land Use / Land Cover (LULC) changes using RSDB over selected sites is demonstrated. Design of analysis framework and selection of a suitable projection scheme with optimization of scale factors has been discussed. The major components of procedure are multi-date RS data geo-referencing, radiometric normalization, top of atmosphere (TOA) reflectance images preparation, multi-image compositing and mosaicing. An RSDB for three rabi (winter) seasons i.e. 1988-89 (226 IRS 1A LISS-II quadrants), 1993-94 (251 IRS 1B LISS-II quadrants) and 2004-05 (three bi-monthly sets, 170 IRS P6 AWiFS quadrants) has been prepared over IGB&CI. Long term LULC changes in terms of increase or decrease in area under agriculture, urban, forest and water bodies have been mapped and quantified using temporal composite images of 1988-89 and 2004-05 rabi seasons. Increase in agricultural area due to conversion of water bodies into crop area (parts of Rajasthan) of the order of 6021 ha (overall increase in crop area from 22383 ha to 43596 ha) and intensification of agriculture into river flood plains (parts of Punjab) of the order of 38.2 percent (of the study area) have been mapped. On the other hand, decrease in agriculture and other land cover area due to urban sprawl has been detected and mapped in the surrounding of Agra town in Uttar Pradesh. ∗ Corresponding author: [email protected] : [email protected]