Using Remote Sensing to Detect Change in the Ndop Floodplain Wetlands of Cameroon
Stephen K. Ndzeidze；R. A. Mbih；G. F. Bamboye
Change Detection ； Wetlands ； Landsat ； Geographic Information System
International Journal of Remote Sensing Applications
|Volume or Term/Year and Month of Publication||
6卷（2016 / 03 / 01）
146 - 158
Using remote sensing and GIS technology to determine change in the Ndop floodplain wetland area from 1973 to 2010, six Landsat MSS, TM and Landsat ETM+ images were classified. Twelve different observed land cover and land use patterns were classified and grouped into four major categories based on the Supervised Maximum Likelihood Classification algorithm. These included the humid floodplain wetland area, agro-pastoral zones, montane forest zone and settlements. Within the wetland area in the floodplain, the reservoir shows evidence of significant fluctuations in surface area since the construction of the Bamendjin reservoir in 1975. Within the reservoir area, acute siltation has been observed since 1988 (1.3%) and this has increased in area by 4.07% in 2002 and by 4.4% in 2010. These increases account for the observed drop in the level of water in the reservoir. A significant drop has been registered in the area occupied by permanent flooded prairies of 11.19% in 1978 and 2.01% in 2010 as well as in that of seasonally flooded prairies of 12.2% in 1978 and 5% in 2010. Areas under irrigated farmland also show decreasing trends from 1988 to 2010. The swamp forest equally exhibits significant corresponding drops in area cover, which directly correlates with the draining of the flood plain for swamp rice cultivation and irrigation. Concerning the agropastoral landscape, the upland grazing areas are generally decreasing in area, while the mixed farming area contrarily increased from 1978 to 2010. This study thus provides base data for monitoring human impacts on the Ndop floodplain wetlands in the Upper Noun drainage basin and natural habitats, especially within and around the wetland area.