Landcover Change Patterns in the Volta Gorge Area, Ghana: Interpretations from Satellite Imagery

Abstract

In much of the developing world, shifting cultivation is the predominant agricultural system usually practised by farmers, which is associated with increasing the unit of land under cultivation in order to increase the output. This small-scale shifting agricultural practices based primarily on burning has been the factor responsible for the conversion of forest fringe vegetation into cropland and pasture and it initiates a continuum of changes in the land cover and consequently fragmentation of the landscape. Since the construction of the Akosombo Dam in 1965, the potential for farming and fishing has increased tremendously within the Volta gorge area and this has resulted in the influx of many settler farmers and fisher folk into the area. This brought so much pressure on the natural resource base of the area and has transformed the landscape accordingly. Within this paper we aim at the estimation of the trend in changes of the landscape using multi-temporal satellite image processing and spatial change analysis. We also seek to determine the land use / cover change in the Volta gorge area of the Volta basin of Ghana for the period 1975 to 2007. Land cover change and agricultural expansion was observed through digital processing and classification based on five multi-temporal medium resolution satellite imagery (Landsat: 1975, 1990, 2000, 2003, 2007) into five classes. From this, accurately classified pixel information was used to determine each landcover class size and the number of changed pixels into other classes through change detection. The study shows that significant changes in the landscape involved a general pattern of conversion of both Closed forest and Open forest and woodland into cropland, fallow lands, pasture, and Bare areas consisting of settlements, roads and exposed soil surfaces. For the period under consideration 1975-2007, the Agricultural class increased from 52,605 hectares in 1975 to become the land cover type with the highest proportion of cover at 221,567 hectares in 2007. The study concludes among others that the expansion of agriculture has increased in all directions with the amount of land devoted to cropland and pasture increasing for all slope categories but more especially for less steeper slopes.

Keywords: Landuse/landcover, Multi-temporal, Digital processing, Change detection