Multitemporal Land Use/Land Cover Change Detection for the Batena Watershed, Rift Valley Lakes Basin, Ethiopia

Abstract

A majorityof therural population in Ethiopiadepends on agriculture. Land use changes during the past couple of decades are mostly linked to agricultural development attributed to factors such as population pressure and environmental changes. Mapping land use/land cover (LULC) to analyze the type, rate, and extent of changes in land use patterns has far reaching significance for policy/decision makers and resource managers to provoke the wide range of applications at regional scales for erosion, landslide, land planning, forest management, and ecosystem conservation. The focus of this chapter is to depict quick and practical approaches to generate spatially and temporally quantified information on land cover dynamics using high-resolution satellite images for the years (1973–2008) in Batena watershed and its environs in southwestern Ethiopia. To quantify the magnitude of LULC change, supervised classification technique was applied using Landsat Thematic Mapper (TM) and Enhanced Thematic Mapper Plus (ETM+) images employing Bayesian maximum likelihood classifier (MLC) with the aid of ground truth training sites. A majority/minority analysis was used for smoothing the classification results and the accuracy of image classification was carried out by means of a confusion matrix generated through geographic information system (GIS) overlay of the classified maps and the test samples. The classification accuracy was further verified by the strong kappa statistical estimate of more than 90 % as a measure of overall agreement between image and reference data. The final output of remote sensing imagery revealed five land cover classes: Grazing land, bush land, mixed forest, dominantly cultivated agricultural land, and water body. It has been discovered that, there were more active LULC change processes in the area in the first study period (1973–1984) than the second study period (1984–1995) and the third study period (1995–2003). On the other hand, areal extent of cultivated and uncultivated agricultural land has been on a steady decline from 39.7 % in 1995 to 41.4 % in 2003 and a mere 50.1 % in 2008. In the first period, nearly half of the landscape underwent land cover change with more than 17 % of the entire landscape experiencing agricultural expansion. In the second period, the extent of the changes was limited to less than 1/3 of the total area with a smaller amount of agricultural area expansion than before. Though the rate of land cover change was observed to vary across the three periods of study, a general decline of forest cover and amplified increase of agricultural lands of more than 41.7 % was found in the area.