IVED 
  
  
Reusing, Matthias 
CHANGE DETECTION OF NATURAL HIGH FORESTS IN ETHIOPIA USING REMOTE 
SENSING AND GIS TECHNIQUES 
Matthias REUSING 
Land Use Planner 
German Technical Cooperation (GTZ), Ethiopia 
German Development Service (DED), Ethiopia 
Land Use Planning and Resource Management Project in Oromia Region (LUPO) 
atomatthias@hotmail.com 
KEY WORDS: Deforestation, Change Detection, Remote Sensing, GIS, Ethiopia. 
ABSTRACT 
Historical sources indicate that an equivalent of 35 % of Ethiopia's land area had once been covered by natural high 
forests. At present, no reliable information on the extend and the location of the past and/or actual forest cover was 
existing for the country. Therefore, a forest monitoring was performed based on airborne and satellite remote sensing 
data to assess the depletion of Ethiopia's natural high forests within the last 25 years. 
The change detection analysis, based on satellite images of 1973 to 1976, indicate that in the seventies, natural high 
forests covered around 4.75 % of the country. Around 10 to 15 years later, only around 0.20 % of the country was still 
covered by undisturbed natural forests. The annual deforestation rate was calculated at 163,600 ha. Today, remarkable 
forest stands can only be found in remote and/or inaccessible southern and southwestern parts of the country. A detailed 
change detection analysis in this region was therefore conducted with airphotos from 1996/97. The results give clear 
evidence that most intact high forests concentrate within the boundaries of ‘National Forest Priority Areas’. 
The ongoing deforestation is a result of the very high human pressure on the natural resources. In this context, it is 
being proposed to design a national Forest Information System (FIS) for selected pilot areas, in order to analyze the 
processes that contribute to deforestation. In conclusion, the FIS will provide veritable information for improved 
planning and decision-making in forest management. 
1 INTRODUCTION 
The natural high forests of Ethiopia have been degraded by human impact since centuries. Historical sources indicate 
that 35 %, respectively 42 Mio ha of the country’s area was once covered by natural high forest (EFAP, 1994). Dry 
coniferous montane forests naturally occur in the northern and central parts of the Ethiopian Highlands, where a semi- 
arid to sub-humid rainfall regime is prevailing. In the semi-humid to humid southern parts of the highlands, the climatic 
climax vegetation are mountain rainforests. The natural upper forest limit generally lies between 3200 and 3500 m.a.s.l.. 
In the western lowlands, along the border to Sudan, one could once find large areas of lowland rainforest. The fauna and 
the flora of Ethiopia’s forest formations are unique in the world due to a large number of endemic plant species. 
Up to now, no reliable data on the qualitative aspects, the quantitative extend and the dynamics of the forest degradation 
was available in Ethiopia. This deficit lead to the motivation to install a forest information system (FIS) at the Natural 
Resources Management and Regulatory Department of the Ministry of Agriculture. Based on multi-temporal remote 
sensing data, the FIS then served for a change detection analysis of Ethiopia’s natural forest resources. 
The results of the analysis are alarming and prove that within the last 25 years vast areas, which were then stocked by 
natural high forests, have been degraded or completely deforested. 
2 METHODOLOGY 
The countrywide change detection analysis of Ethiopia’s natural high forests was realised with multi-temporal 
LANDSAT-TM satellite images of the seventieth and the eightieth. Apart from this, a detailed forest monitoring was 
performed in the once densely forested SW of Ethiopia. In this case, multi-temporal black & white aerial photographs 
served as input. ; 
All remote sensing data that was used had been classified in scope of other projects. It was therefore necessary to 
integrate and condense the data within the FIS in order to perform the further analysis steps. The given preconditions 
restricted the forest monitoring to the following classes: 
  
International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B7. Amsterdam 2000. 1253