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

Accueil du site → Doctorat → Allemagne → 2018 → Forage supply of West African rangelands - Towards a better understanding of ecosystem services by application of hyperspectral remote sensing

Rheinischen Friedrich-Wilhelms-Universität Bonn (2018)

Forage supply of West African rangelands - Towards a better understanding of ecosystem services by application of hyperspectral remote sensing

Ferner, Jessica

Titre : Forage supply of West African rangelands - Towards a better understanding of ecosystem services by application of hyperspectral remote sensing

Auteur : Ferner, Jessica

Université de soutenance Rheinischen Friedrich-Wilhelms-Universität Bonn

Grade : Doktorgrades (Dr. rer. nat.) 2018

Résumé
Grazing is the predominant type of land use in savanna regions all over the world. Although large savanna areas in Africa are still grazed by wild herbivores, the West African Sudanian savanna region mainly comprises rangeland ecosystems, providing the important ecosystem service of forage supply for domestic livestock. However, these dryland rangelands are threatened by global change, including a predicted in-crease in climatic aridity and variability as well as land degradation caused by overgrazing. In this context, the international research project WASCAL (West African Science Service Centre on Climate Change and Adapted Land Use) was initiated to investigate the effects of climatic change in this region and to develop effective adaptation and mitigation measures. This cumulative dissertation aims at providing a methodology for a regular knowledge-driven monitoring of forage resources in West Africa. Due to the vast and remote nature of Sudanian savannas, remote sensing technologies are required to achieve this goal. Hence, as a first step, it was necessary to test whether hyperspectral near-surface remote sensing offers the means to model and estimate the two most important aspects of forage supply, i.e. forage quantity (green biomass) and quality (metabolisable energy) (Chapter 2.1). Evidence was provided that partial least squares regression was able to generate robust and transferable forage models. In a second step, direct and indirect drivers of forage supply on the plot and site level were identified by using path modelling within the well-defined concept of social-ecological systems (Chapter 2.2). Results indicate that the provisioning ecosystem service of forage supply is mainly driven by land use, while climatic aridity exerts foremost indirect control by determining the way people use their environment. Building on these findings, upscaling of models was tested to generate maps of forage quality and quantity from satellite images (Chapter 2.3). Here, two different available data sources, i.e. multi- and hyperspectral satellites, were compared to serve the overall objective to install a regular forage monitoring system. In conclusion, preliminary forage maps could be created from both systems. An independent validation would be a research desiderate for future studies. Moreover, both systems feature certain shortcomings that might only be overcome by future satellite missions

Mots clés  : West Africa, field spectroscopy, partial least squares regression, rangeland, metabolisable energy, green biomass, forage provision, climate change land-use pressure, social-ecological system (SES), structural equation modelling (SEM), partial least squares path modelling (PLS-PM), field spectroscopy, monitoring, Hyperion, Sentinel-2

Présentation

Version intégrale (5,2 Mb)

Page publiée le 24 novembre 2021