Spatio-temporal information and analysis of land use/land cover changes in the Muni-Pomadze wetland

Abstract

Land use/land cover dynamics in wetland catchments is poorly understood; even though it is an important indicator of wetland ecological health. The study was aimed at detecting and quantifying Land Use/Land Cover (LULC) changes at the Muni-Pomadze Ramsar site in Ghana. The objective was to produce LULC maps for the area, assess the LULC dynamics during a ten year period, and establish statistical relationship between the major LULC changes and biophysical factors that influence human decisions. Using image processing techniques, an aerial photograph of January 2005, Landsat Thematic Mapper (TM) and Ehanced Thematic Mapper (ETM) data of December 1990 and February 2000 respectively were analysed to produce land use/land cover maps for the Muni-Pomadze wetland catchments. The image analysis produced an overall accuracy of 85% and a kappa coefficient of 81%. The post classification technique involving overlay operation was applied to produce land use/land cover change map and matrix which aided the analysis of LULC dynamics in the area. During the 10 year period, various land use/land cover units such as farm fields, rangeland, built-up, and barren-land areas increased by 20%, 20.6%, 40.5% and 13% respectively. However, closed forest, open forest, and water reduced by 23%, 8.9%, and 5.9% respectively. The logistic regression in SYSTAT established that altitude, distance to roads, rivers, and settlement were significant in explaining deforestation and afforestation.