Investigation of deforestation in East Africa on regional scales

• Main Author: Wu, Yi-Hua
• Format: Others
• Language: English
• Published: Stockholms universitet, Institutionen för naturgeografi och kvartärgeologi (INK) 2011
• Online Access: http://urn.kb.se/resolve?urn=urn:nbn:se:su:diva-63543

Tropical forests contain abundant natural resources and play an important role in the balance of the ecosystems and environment. Depletion of forests could destroy habitats of endangered plants and animals and cause biodiversity loss. Rapid deforestation is a major problem in East Africa and seriously affects desertification and climate change in East Africa. More monitoring of the deforestation in East Africa are emergent. Therefore, this study was conducted to identify and evaluate the spatial and temporal distributions and determinants of deforestation in East Africa. Two kinds of satellite image datasets, including Landsat images and GIMMS data were used to map the deforestation in Kenya, Tanzania and Uganda. Possible drivers of deforestation were analyzed, including population statistics, economic and climate data. The analysis of Landsat images was focus on the forests, including Mount Kenya, Mao forest, Aberdares forest as well as Mount Kilimanjaro in Tanzania and its surroundings. Supervised classification was carried out on the images comprising PCA component images and Tassel Cap transformed images to identify forest area and non forest area. High Kappa coefficient of the classification indicated that using the images that comprising the enhancement images transformed from original images would be a better approach to mapping forest areas. The obvious deforestation was observed in Mau forest, Mountain Kilimanjaro and Aberdares forest close to Nairobi city from 1980s to 2000s. The analysis using the GIMMS NDVI dataset did not show a significant decline of NDVI values during the study period. The results indicate that the GIMMS NDVI is not a good proxy of total forest areas because of the coarse resolution of GIMMS dataset and the characteristics of NDVI. Future studies should use higher resolution satellite images and collect enough information to monitor deforestation.