ANALYSIS OF REMOTELY SENSED DATA FOR WASTELAND MAPPING

Abstract

In India various types of soil. and land degradation problems have rendered about 175 million or 53 percent of its total geographical_ area either completely or partially unprodu-ctive, which is also ]mown as the wasteland.' Wasteland may be. defined as the land which is presently lying un--ut.ilized or not being used to its optimum potential due to various const-raints. It includes mainly ravine land, undulating upland, water logged land, salt of Eected land, shifting cu,i tivation land, degraded forest land, degraded grazing land, strip land, sands degraded plantation land, industrial wasteland, rock,' stoney wasteland, steep sloping land and sno• cov6red or glacial land. Growing population have necessiated for the optimum utilization of st,elinds. Wastelands, affect ,the economy of country by contributing to flood and premature siltation of reservoir sr In this dissertation, remote sensing techniques have been used to delineate and map the wastelands boundaries in a part of Aligarh and Etah districts, the aluvial - plains of river• Ganga. FCC of Landsat Thematic Mapper data of path row no 145 - 041, dated larch 9,1986, and. 146 - 041 dated -Nov. 1, 1936 at scale 1:50,000 as well as, KiS band 5 and band 7 imagery and FCC (NS) of path row no 156 - 041 dated Nov. 13, 1902 at scale 1:250,000 were, analysed for mapping by visual interpretation technique. Topographical map sheet nos. 54:3/13, 541/1, 541/2, 54I/6 and 541/7 at 1:50,000 scale and 54E and 541 at 1:250,000 were also used to compile base maps. Field visit for typical sites in the area was also carried out to verify the accuracy of interpreted details and post field modifications. The areal extent of each wasteland class has been compute: using dot grid shcct. The inter-pretability of various data products has. also been determined by comparing the same waste-land class, delineated from various remotely sensed data of same season and year. Change between various hypes of waste-lands e.g. salt affected land and wa. er logged land determined, and has been found increasing by about 1 S and 0.5 respecti-vely during a span of 4 years (1982 - 1986) . The accuracy of interpretation was also estimated which came out to h,-, :bout y 1 h,e f the study nicl _ ~ measures ina11 at ~ end o_ th.. ~ udy s~-.:~c ,roc._.:~rn~ ~~..m m~~:•,. in the area and an idoal~ monitoring system for such typo of studies have also boon suggested.