Remote Sensing & GIS for Land Cover/Land Use Change Detection and Analysis in the Semi-Natural Ecosystems and Agriculture Landscapes of the Central Ethiopian Rift Valley

Band 7 der Reihe „Fernerkundung und angewandte Geoinformatik“

Bedru Sherefa Muzein


Band 7 der Reihe „Fernerkundung und angewandte Geoinformatik“
ISBN: 978-3-941216-23-5
Veröffentlicht: September 2010, 1.. Auflage, Einband: Broschur, Seiten 190, Format DIN B5, Gewicht 0.3 kg
Lieferzeit: 2 - 6 Werktage
Verfügbarkeit: Auf Lager
36,50 €

Remote Sensing & GIS for Land Cover/Land Use Change Detection and Analysis in the Semi-Natural Ecosystems and Agriculture Landscapes of the Central Ethiopian Rift Valley

Mehr Ansichten


Bedru Sherefa Muzein

Remote Sensing & GIS for Land Cover/Land Use Change Detection and Analysis in the Semi-Natural Ecosystems and Agriculture Landscapes of the Central Ethiopian Rift Valley

Band 7 der Reihe „Fernerkundung und angewandte Geoinformatik“. 190 Seiten. Format B5. Broschur. Zahlreiche Abbildungen, davon 29 farbig. Preis: 36,50 Euro. ISBN 978-3-941216-23-5. Rhombos-Verlag, Berlin 2010.

Nowadays earth observation satellites continuously acquire huge amounts of data with the effect that the capacity of current technologies to extract and analyse spatial data
can hardly cope up properly. This abundance of spatial data is especially in developing countries not adequately used because the setup of spatial data infrastructure is at its

This research work represents a valuable step towards the establishment of the relationship of spectral signatures of MODIS satellite imagery and biophysical characteristics of agricultural and semi-natural ecosystems along parts of the Ethiopian Rift Valley. Important issues regarding the assessment of multi-temporal dynamics of the encroachment of protected areas by livestock are systematically investigated by analyzing the monthly animal feed shortage or surplus as detected by MODIS imagery. Easy-to-handle empirical formulae and extended models of respective calculation are provided. The trend of degradation of ecologically important areas is accurately mapped and the annual net loss is calculated.   
Experts and stakeholders in natural resources management, in remote sensing and GIS with special regard to assessment of land degradation and sustainable land management as well as  policy makers dealing with environment governance and students aspiring to become experts in natural resources management and remote sensing and/or GIS for ecological regional development may find this book useful for enlarging their methodological as well as practical level of knowledge.

The author


Bedru Sherefa Muzein (born in 1971) studied Forestry at the Alemaya University (Ethiopia) and worked in various natural resources management related projects. In 1999 he joined the Technology University of Dresden where he completed his MSc (2001) and PhD (2007) in the use of Remote Sensing and GIS technologies for monitoring land degradation and ecosystems dynamics.
Bedru S. Muzein is an expert in the use of contemporary geospatial technologies for natural resources and environmental monitoring. He has been working in the areas of natural resources, biodiversity, water management and related fields in international and national organisations in Europe and Africa. He is a recognised practitioner and trainer of GIS and Remote Sensing methods in these fields.
This thesis was written in partial fulfilment of his Dr. rer. nat. degree received from the Institute of Photogrammetry and Remote Sensing at the Technology University of Dresden in 2007.

The editor/Der Herausgeber


Prof. Dr. techn. habil. Elmar Csaplovics leitet den Lehrstuhl Geofernerkundung am Institut für Photogrammetrie und Fernerkundung der Technischen Universiaet Dresden

TU Dresden, Institut für Photogrammetrie und Fernerkundung, Helmholtzstraße 10, 01062 Dresden

Prof. Dr. Elmar Csaplovics: "In loser Folge sollen in dieser Schriftenreihe wissenschaftliche Arbeiten, die am Lehrstuhl für Fernerkundung bearbeitet und betreut werden, in ansprechender Form veröffentlicht werden. Wir glauben, dass dadurch das Spektrum von Literatur zum Themenkreis projektorientierter angewandter fernerkundlicher Forschung nachhaltig bereichert werden kann und wünschen uns demgemäß eine kritikfreudige Schar von Leserinnen und Lesern." Professur Geofernerkundung



“Situated in the Great Rift Valley, only 200 kilometres (124 miles) south of Addis Ababa, and in the Lake Langano recreational areas, the Abijatta Shalla Lakes National Park (ASLNP) attracts numerous visitors. Using Lake Langano as your base, it is an easy trip to visit the national park, which is 887 square kilometres in size, 482 of these being water. The altitude of the park ranges from 1540 to 2075 metres, the highest peak being Mount Fike, situated between the two lakes.
The network of tracks in this park is always developing. At present you can enter at four different points, three of which are interconnected. Approaching from Addis you first reach the Horakello entrance, where the small Horakello stream flows between lakes Langano and Abijatta.
There are two different lakes in the park. They are both terminal lakes and their beaches are unstable and saline, but they are very different in character. Abijatta is shallow at about 14 metres with a mysterious fluctuating water level. Fresh water flows into it through the small Horakello stream. The stream mouth is a source of relatively fresh water, much frequented by water birds for drinking and bathing. The lake is surrounded by gentle, grass covered slopes and acacia woodlands.
Lake Shalla  by contrast, surrounded as it is by steep, black cliffs and peaks which reflect in its waters, is the deepest lake of the Rift Valley (260 metres). It is exceptionally beautiful, with shores which give a scent of mystery with their hot sulphurous springs which bubble up and flow into the lake. (, visited 8 January 2007)”
So far so good what concerns the point of view of travel agencies. The Ethiopian Rift Valley is both part of one of the most famous geological structures worldwide, representing abundant floristic and faunistic treasures, as well as focus of competing impacts driven by migration of people along the shores of the lakes, overuse of the land and resulting conflicts with issues of nature protection versus tourism. The region was formerly divided into a patchwork of small kingdoms with specific traditions and customs, oftentimes at war among one another and successively incorporated into the emerging Ethiopian empire during the 19th century (Huntingford 1955). Political and climatic constraints of the last 30 years more and or more destabilised the regions and induced the growth of local to regional patterns of agglomerations of settlements of migrating people, overexploitation of sensible soils, destruction of vegetation cover and accelerating dynamics of degradation. On the other hand, governmental planning ignored these impacts by supporting the establishment of national parks not primarily for allowing for protection of nature and wildlife, but for creating new potentials for implementing tourism. Thus there is a two-fold problem arising and more and more disturbing the balance of ecosystems, that is taking care of nature conservation as well as of providing people with sustainable resources.
Lack of spatial information on the historical status and the status quo of land cover and land use, on the patterns of change during the harsh decennia since 1970 and on the perspectives of sound regional planning for establishing stabilised ecosystems preserving floristic and faunistic values as well as allowing the people to sustainable use the land for their livelihood is evident.
The research of Bedru Sherefa Muzein approaches these deficits in knowledge on spatio-temporal dynamics of land use and land cover in the Ethiopian Rift Valley and tries to give initial recommendations on how to use remote sensing and GIS for monitoring and assessing as well as for analysing the trends in land degradation, the losses in habitats for bird life in particular and wild life in general and the local specifications on patterns of change. Appropriate use of high-end technologies is a challenge which has to be coped urgently. Data are captured and are available, methods are developed and in action, but still these data and technologies are black boxes handled by a selected minority of adepts, who primarily do not share their knowledge deliberately with others. Bringing down spatial data acquisition and analysis to the decentralised local and regional topographical settings, thus to the regional actors, is one of the main aims and results of this research.
Spatial analysis by means of low-cost remote sensing and geographic information systems (GIS) proved that the establishment of parks does not imply protection of the environment and of biodiversity at all. It appears that the natural and semi-natural ecosystems of ASLNP and the Important Bird Areas (IBAs) are subjected to dynamics of land cover land use change (LULC) in the same way as the non-protected areas. Spatial changes in the regions of interest show that the parks cannot continue to protect the biological resources within their borders any more – the national park concepts are not working! On the other hand the application of advanced but easy-to-handle and low-cost methodologies of remote sensing data interpretation and analysis sets up the opportunity to support managers of parks to raise awareness of political decision makers. Moderate Resolution Imaging Spectroradiometer (MODIS) data integrated in the research is available free of cost. An extended evaluation of the level of fulfilment of objectives of the respective research highlights the reliability of the adopted approaches to land cover and land use assessment based on remotely sensed data, image analysis and spatial data analysis at low-cost and under minimisation of external inputs, thus following the theorems of appropriate technologies as set up by Kohr (1957) and Schumacher (1973). It is important to say that Bedru did not hesitate to also focus on the limitations of the approaches implemented in the research. Especially the poor data on habitat statistics like bird counts, but also on human population and lack of spatial data of ecologically important areas like IBAs or lack of inventories of ecosystems themselves form obstacles which can only be overcome by attracting attention of political decision makers and thus by raising funds for building sound strategies for protection and management of Ethiopian national parks.
Bedru does not hesitate to call a spade a spade. This is important and scientifically correct, as it is grounded on the outstanding results of in-depth research. Patterns of land degradation clearly make visible that ASLNP is paralysed by dynamics of anthropozooic pressure on the land. On the one hand IBAs should be preserved by stabilising lake levels, on the other hand developing agencies and governmental politics support people in establishing - how they call it - ecologically sound ways of using limited water resources by small-scale irrigation projects. It can be proven that these projects are ecologically sound only at first view, because they contribute to the shrinking of the lakes themselves, as highlighted by shoreline classification of respective remote sensing imagery. Institutional coordinators should carefully study these facts and reduce exaggerated implementation of small-scale irrigation schemes. This approach - if affecting water levels of the lakes - is neither in coincidence with the regional climatic situation nor with the human impact and additionally keeps local people from developing and/or adapting suitable appropriate methods of water collection and use. Bedru stresses the limited willingness and preparedness of institutional/governmental representatives in the environmental sector to integrate spatial data capture and analysis much more efficiently into national programmes of mapping and monitoring. The presented outputs satisfy the specific needs for supporting protection of ASLNP and also correspond to international standards of national park monitoring and spatial management. Networking of park authorities and local/regional institutions in general as well as full-scale biodiversity monitoring in particular are neither implemented on European nor on global levels (Csaplovics and Wagenknecht 2006). It is thus evident that increasing loss of awareness concerning the destruction and neglect of natural and cultural heritage due to uncontrolled consumption of land caused by purely economical interests is driven by a distinctive global dimension.
The research conducted by Bedru Sherefa Muzein represents a most valuable, important and innovative contribution to proving for the wide-spread reliability of remote sensing image analysis and multi-thematic spatial data analysis for monitoring, assessing and reasoning in fragile ecosystems of African national parks particularly. The results and recommendations serve as a profound basis for setting up strategies of protective management and periodical monitoring of state and dynamics of land use and land cover as well as of ecological parameters like habitat suitability on the one hand and biomass as well as productivity on the other. The study perfectly coincides with the fact, the ecosystems have to be understood, assessed and managed as holistic units both in terms of topography and land cover as well as in terms of varieties of demands driven by ecological versus economical interests. It is recommended to implement the results of research into governmental strategies of land and national park management in Ethiopia.
Elmar Csaplovics, Stefan Wagenknecht, Ecological Landscapes. GeoConnexion International Magazine 5(10): 59-61, 1996
George Wynn Brereton Huntingford, The Galla of Ethiopia - the Kingdoms of Kafa and Janjero. International African Institute, London, 1955
Leopold Kohr, The breakdown of nations. Routledge and Kegan, London, 1957
Ernst Friedrich Schumacher, Small Is beautiful - economics as if people mattered. Harper and Row, New York, 1973
Dresden, July 2010
Prof. Dr. habil. Elmar Csaplovics


Technical complexities and the high cost of satellite images have hindered the adoption of remote sensing technology and tools for nature conservation works in Ethiopia as in many developing countries. The terrestrial and aquatic ecosystems in Abijjata Shala Lakes National Park (ASLNP) and the Important Bird Areas (IBAs) around the park are considered to be one of the most important home ranges for birds. However, little is known about the effect of land use/land cover (LULC) dynamics, due to lack of technical know how and logistical problems. However, it has been shown in this study that sophisticated image management works are not always relevant. Instead a simple method of utilizing the thermal band has been demonstrated. A new approach of long-term dynamics analysis method has also been suggested. A successful classification of images was achieved after such simple enhancement tests. It has been discovered that, there were more active LULC change processes in the area in the first study period (1973 to 1986) than during the second study period (1986-2000). In the first period nearly half of the landscape underwent land cover change processes with more than 26% of the entire landscape experiencing forest or land degradation. In the second period the extent of the change process was limited to only 1/3 of the total area with a smaller amount of degradation processes than before. During the entire study period, agriculture was responsible for the loss of more than 4/5 of the total terrestrial productive ecosystem. More than 37.6% of the total park area has been experiencing this loss for the past 3 decades. Only 1/5 of this area has a chance to revive, the remaining has undergone a permanent degradation. Lake Abijjata lost half of its size during the past 30 years. In the Zeway-Awassa basin 750 km², 2428km² and 3575km² of terrestrial lands and water bodies are within a distance of 10km, 20km and 30km from IBAs respectively. There are ecologically important areas where two or more IBAs overlap. In areas where more than two to five IBAs overlap, up to 85km² of areas have been recently degraded. High livestock density is one of the reasons for degradation. Using a monthly MODIS data from 2000-2005 and a series of interpolation techniques, the productivity of the area as well as the standing biomass were estimated. Moreover, a new method of spatially accurate livestock density assessment was developed in this study. Only 0.3% of the park area is found to be suitable for productive livestock development but nearly all inhabitants think the area is suitable. Feed availability in ASLNP is scarce even during rainy seasons. Especially the open woodlands are subject to overgrazing. Such shortage forces the inhabitants to cut trees for charcoal making to buy animal feed and non-food consumables. While more than 95% of the inhabitants in the park expanded their agriculture lands, only 13.3% of the farmers managed to produce cereals for market. The application of low cost remote sensing and GIS methods provided ample information that enables to conclude that low productivity and household food insecurity are the main driving forces behind land cover changes that are negatively affecting the natural and semi-natural ecosystems in the central and southern Rift Valley of Ethiopia. The restoration of natural ecosystems or conservation of biodiversity can be achieved only if those driving forces are tackled sustainably.

Table of Contents


Chapter 1    15
1    Introduction    17
1.1    Remote Sensing for Ecosystem Management    17
1.2    Alteration of Ecosystems and its Consequences in the Ethiopian Rift Valley    18
1.3    Aim and scope of the study     20
1.3.1    Conceptual Framework of the Study     20
1.3.2    General Objective    20
1.3.3    Specific objectives    21
1.4    Organisation of the Dissertation    21
Chapter 2    23
2    The Evolution and Revolution of Remote Sensing
and GIS for Nature Conservation    25
2.1    Historical Development of Application of Remote Sensing and GIS in Conservation     25
2.1.1    History of Remote Sensing    25
2.1.2    History of GIS    27
2.2    Mapping for Monitoring and Conservation of Natural Ecosystems     28
2.2.1    The Emergence of Remote Sensing and GIS as a Major Tool for Ecosystems Monitoring    28
2.2.2    Mapping Habitat Ranges in GIS    32
2.2.3    Distribution of Migratory Species Rich Localities and their Conservation Status    34
2.3    Conservation Status of Migratory Species Rich Ecosystems    37
Chapter 3    39
3    Social and physical attributes of the study area    41
3.1    Location of the Study Area and Its Extent    41
3.2    Physical Attributes     42
3.2.1    Climate    42
3.2.2    Hydrology    43
3.2.3    Soils    44
3.2.4    Vegetation    44
3.3    Wildlife and Tourism    45
3.3.1    Protected Areas     45
3.3.2    Important Bird Areas (IBAs)    47
3.4    The Socio-cultural and Economic Situation of Abijjata Shala National Park and its Surroundings    47
3.4.1    Population    47
3.4.2    Farming Systems and Land Use    49
3.4.3    Relationship of the Inhabitants to the Protected Area System    49
3.4.4    Government Policies and the Natural Ecosystem    49
Chapter 4    51
4    Data Organization and Methodology    53
4.1    Materials Used    53
4.1.1    Satellite Imagery    53
4.1.2    Topographical Maps    57
4.1.3    Digital Elevation Models (DEM)    58
4.1.4    Metrological Station Data and Climate Map Formation    58
4.1.5    Field Surveying    62
4.1.6    Ecological Important Areas    65
4.1.7    Livestock Data    66
4.2    Pre-Classification of Digital Image Processing    66
4.2.1    Radiometric and Atmospheric Correction    66
4.2.2     Temporal Normalisation    70
4.2.3    Geocoding and Georeferencing    72
4.2.4     Topographic Normalisation    72
4.2.5    Tasselled Cap Transformation    75
4.2.6    Thermal Bands for Use in Land Cover Classification    75
4.3    Land Cover Classification Methods and Their Applicability to this Study     77
4.3.1    Deterministic Classification     77
4.3.2    Fuzzy/Soft Classification    81
Chapter 5    83
5    Multi-Temporal and Multi-Scale Land Cover Classification and Biophysical Information Extraction by Means of Low-Cost Remote Sensing    85
5.1    Results of Land Cover Classification     85
5.1.1    Derivation of Land Cover Classes from MODIS Datasets    85
5.1.2    Classification of Landsat Imagery    88
5.2    Accuracy Assessment    90
5.3    Procuring Biophysical Information through Low-Cost Remote Sensing    92
5.3.1    Spectral Characteristics of Plant Leaves    92
5.3.2    Estimating Leaf Area Index for Wide Area Coverage    94
5.3.3    Establishing Empirical Relation Ship Between
Remote Sensing Variables and LAI    95
5.3.4    MODIS-LAI Product    103
5.3.5    Standing Biomass and NPP Estimation through Low-Cost Remote Sensing Products    103
5.3.6    Assessing the Potential Productivity by Means of Remote Sensing Variables and Physical Attributes of the Area    108
5.4    Summary and Discussion    116
Chapter 6    119
6    Livestock Centred Land Use and
the Natural Ecosystem    121
6.1    Local and National Perspectives of Priorities of Land Use     121
6.2    The Land Cover Dynamics in and around the ASLNP    122
6.2.1    Land Cover Change Process    122
6.2.2    Major Land Use/ Land Cover Change Driving Forces     126
6.3    Identification of Critical Sites for Conservation around the
Semi-natural Ecosystems in the Zeway-Awassa basin    126
6.3.1    Geographic and Topographic Distribution of Important Changes in the Landscape    126
6.3.2    Multivariate Gradient Analysis    128
6.4    Remote Sensing of Livestock Based Farming System: The Issue of Land Suitability and Carrying Capacity of Semi-Natural Ecosystem    129
6.4.1    Distribution of Livestock in the Study Area    130
6.4.2    Daily Average Feed Requirement in the Rift Valley    134
6.4.3    Land Suitability Assessment     134
6.4.4    The State of Other Protected Area Networks    136
6.4.5    Carrying Capacity of Livestock in and around Protected Areas    137
6.4.6    Carrying Capacity of Land Cover Classes     138
6.5    Environmental Implications of Land Cover Changes
and Livestock Impact    139
6.6    Summary and Discussion    140
Chapter 7    143
7    Overall Conclusions, Discussion and Recommendations    145
7.1    Conclusions and Discussion     145
7.2    Limitation of the Study    149
7.3    Recommendations and Outlook for Future Studies    150
7.3.1    Remarks on Strategic Approaches to Save ASLNP     150
7.3.2    Suggestions to Preserve the IBAs    150
7.3.3    Institutional Coordination    151
7.3.4    Limits of Spatial Technology Adoption for Nature Conservation     151
7.4    Indications for Further Studies    152
8    References     153
9    Appendices    167

Abbreviations and acronyms

APAR    Absorbed Photosynthetically Active Radiation
ARTEMIS    The Africa Real Time Environmental Monitoring Information System
ASTER    Advanced Spaceborne Thermal Emission and Reflection Radiometer
ASLNP    Abijjata Shala Lakes National Park
AVHRR    Advanced Very High Resolution Radiometer
CASA    Carnegie-Ames-Stanford Approach
CBD    Convention on Biological Diversity
CCA    Canonical Correspondence Analysis
CITES    Convention on International Trade in Endangered Species
CMS     Convention on Migratory Species
COST    Cosine of Solar Zenith Angle. Correction
CSA    Central Statistics Agency
DCA    Detrended Correspondence Analysis
DEM    Digital Elevation Model
DEPHA    Data Exchange Platform for Horn of Africa
DN    Digital Numbers
DOS    Dark Object Subtraction
DOY    Date Of Year
EMA    Ethiopian Mapping Agency
EROS    Earth Resource Observation
ESDI    Earth Science Data Interface
ESRI    Environmental Systems Research Institute
ETM+    Enhanced Thematic Mapper plus
EWCO     Ethiopian Wildlife Conservation Organisation
EWNHS     Ethiopian Wildlife and Natural History Society
FAO    Food and Agriculture Organisation
FAPAR    Fraction of Absorbed Photosynthetically Active Radiation
FDRE    Federal Democratic Republic of Ethiopia
FEWS    Famine Early Warning System
GCP    Ground Control Points
GIEWS    Global Information and Early Warning System
GIS    Geographic Information System
GLCF    Global Land Cover Facility
GPP    Gross Primary Production
GROMS    Global Register of Migratory Species
Ha    Hectare
HS    Habitat Suitability
IBA     Important Bird Areas
IDWA    Inverse Distance Weighted Average
ILRI    International Livestock Research Institute
IR    Infrared
ISODATA    Iterative Self Organizing Data Analysis Technique
IUCN    International Union for the Conservation of Nature and Natural Resources
KBC     Knowledge Base Classification
kg    Kilogram
LAI    Leaf Area Index
LSR     Least Square Regression
LUE    Light Utilisation Efficiency
LULC    Land Use Land Cover
MFF    Modified Flat Field
MIR     Mid Infrared
MJ    Mega Joule
MLC     Maximum Likelihood Classifier
MODIS    Moderate Resolution Imaging Spectroradiometer
MSS    Multispectral Scanner
NASA    National Aeronautics and Space Administration
NDVI    Naturalized Differential Vegetation Index
NDVIc     Corrected Naturalized Differential Vegetation Index
NIR    Near Infrared
NPP    Net Primary Production
PAR    Photosynthetically Active Radiation
PCA    Principal Component Analysis
RSR    Reduced simple Ratio
RS    Remote Sensing
R²    Goodness of Fit
SAR     Synthetic Aperture Radar
SAVI    Soil Adjusted Vegetation Index
SD    Standard Deviation
SPOT    Satellite Pour l’Observation de la Terre
SR    Simple Ratio
SRTM    Shuttle Radar Topography Mission
SWIR    Short Wave Infrared
TCT    Tasseled Cap Transformation
TIN    Triangular Irregular Network
TLU    Tropical Livestock Unit
TM    Thematic Mapper
TRFIC     Tropical Rainforest Information Center
UNCCD    United Nations Convention to Combat Desertification
UNEP    United Nations Environmental Programme
UNITAR     United Nations Institute for Training and Research
UNOOSA    United Nations Office for Outer Space Affairs
UNOPS     United Nations Office for Project Services
UNOSAT    United Nations Satellite (image providing service)
USGS    United Sates Geological Service
VI    Vegetation Index
WCMC    World Conservation Monitoring Center
WHC    World Heritage Convention



Gewicht 0.3000
Lieferzeit 2 - 6 Werktage