Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Allemagne → 2015 → Remote sensing-based analysis of land cover detrimental for ecosystem services in fragile lands of Ethiopia

University of Bayreuth (2015)

Remote sensing-based analysis of land cover detrimental for ecosystem services in fragile lands of Ethiopia

Ayanu, Yohannes

Titre : Remote sensing-based analysis of land cover detrimental for ecosystem services in fragile lands of Ethiopia

Auteur : Ayanu, Yohannes

Université de soutenance : University of Bayreuth

Grade : Doctoral thesis, 2015

Fragile lands such as mountainous regions and drylands are highly vulnerable for land degradation and subsequent decline in productivity resulting from anthropogenic and natural causes. In developing countries of the tropics like Ethiopia, the human-induced impact is aggravated by the increasing population of subsistence farmers living in these areas. Land use/land cover type mainly determines the sustainability of supplies of ecosystem services and biodiversity supported. Land use decisions made without caution often deteriorate ecosystems in fragile lands and have detrimental impacts on supplies of ecosystem services. Therefore, continuous monitoring and assessment of land use/land cover types in fragile lands is essential to ensure sustainable supplies of ecosystem services in such environments that are liable for deterioration. Remote sensing provides fast and recurrent data for assessing land cover and ecosystem services. The main objective of the dissertation is identifying the potentials and limitations of remote sensing for assessing ecosystem services and map two major landcover types detrimental for ecosystem services in fragile lands of Ethiopia. Ethiopia was chosen for the case studies due to the ongoing pressure on fragile lands of the country which is triggered by population growth, large-scale agricultural land acquisition and problems arising from invasive species. The thesis is organized in series of chapters described below.Overview of the thesis highlighting the research questions, methods and major findings is presented in Chapter 1. Following the general overview (Chapter 1), potentials andlimitations of remote sensing in quantifying and mapping ecosystem services are reviewed (Chapter 2). The review showed that there is uncertainty involved in quantifying and mapping ecosystem services with remote sensing data which calls for more research to find the link between ecosystem services and image spectra. Moreover, while selecting remote sensing data, factors such as resolution, sensor types, and financial and technical capacity of users need to be considered. In Chapter 3, the trendsin Prosopis juliflora invasion of the Awash basin of Ethiopia were mapped using Landsat ETM + and ASTER images for the years 2000, 2005, 2010 and 2013, and potential impacts on ecosystem services were assessed. Results showed that over the past decade P. juliflora spread rapidly and has had negative impacts on the supplies of ecosystem services such as provisioning and cultural services. Further research is needed to understand drivers of P. juliflora invasion, quantify its impacts on ecosystem services and identify controlling mechanisms. Chapter 4 discusses undercover cropland inside forests of the Bale Mountains of Ethiopia and its influential factors. Land use/land cover classes were derived by classifying RapidEye images using Random Forests classification approach. Undercover cropland was mapped using Boosted Regression Trees on field observed percent cover, topographic and location and parameters. The influential factors of undercover cropland are elevation, distance to settlements, slope, East aspect and distance to national park with elevation being the most important factor. Therefore, ecosystem management efforts in such mountainous areas should be based on the relative importance of these influential factors. In the last section (Chapter 5), the studies in the thesis are synthesized and presented. Besides, recommendations about monitoring of fragile lands and ecosystem services, management of invasive species, mountain regions, and future prospects of remote sensing in ecosystem services assessment are provided.

Mots clés  : Remote sensing ; Fragile lands ; Ecosystem Services ; Land cover ; Invasive species ; Croplands


Version intégrale (22 Mb)

Page publiée le 5 décembre 2015, mise à jour le 30 décembre 2018