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

Accueil du site → Doctorat → Allemagne → 2015 → Interlinkages of Land Degradation, Marginality and Land Use Cover Change in Kenya - Development of an interdisciplinary framework using remote sensing and GIS

Rheinischen Friedrich-Wilhelms-Universität Bonn (2015)

Interlinkages of Land Degradation, Marginality and Land Use Cover Change in Kenya - Development of an interdisciplinary framework using remote sensing and GIS

GRAW VALERIE ANNEMARIE MARTINE

Titre : Interlinkages of Land Degradation, Marginality and Land Use Cover Change in Kenya - Development of an interdisciplinary framework using remote sensing and GIS

Auteur : GRAW VALERIE ANNEMARIE MARTINE

Université de soutenance : Rheinischen Friedrich-Wilhelms-Universität Bonn

Grade : Dr. rer. nat. 2015

Résumé partiel
Land degradation (LD) is a global problem affecting and being affected by socio-ecological systems. This thesis analyses the interlinkages of LD, marginality and land use cover change (LUCC) in Kenya based on remote sensing and geographic information systems (GIS). An interdisciplinary framework is developed using two different scales – a national scale looking at the country of Kenya and a local scale analyzing a specific area in western Kenya. LD stands for the decrease of soil fertility and, hence, land productivity. By combining biophysical and socioeconomic data we obtain a deeper understanding of internal dynamics and their relationship to processes of decreasing productivity within a coupled Human-Environment System (HES). In addition q-squared methods are used which describe the simultaneous use of quantitative and qualitative methods and thereby support insights in different disciplines. Marginality is defined as the root cause of poverty but goes beyond the solely economic perspective of poverty measurement. LUCC, based on the Global Land Programme (GLP) initiative started in the 1990s, represents another interdisciplinary concept. On the one hand land cover (LC) refers to the land surface and its biophysical determinants which can be detected with remote sensing. On the other hand land use (LU) includes an active component referring to activities on land by human impact. The question how land is e.g. used by agricultural production can be approached by gaining insight in socio-economic structures, especially via information on agricultural activities. The national study on Kenya focuses on all 47 counties of the county. Insight in the socioeconomic perspective was given with census data and household survey information while biophysical assessment on LD and LUCC was conducted via remote sensing imagery. Time series analysis of vegetation, using MODIS Normalized Difference Vegetation Index (NDVI) Terra (MOD13A1) with 500m resolution was included to analyze trends of land productivity from 2001 to 2011. Analyzing trends of LD and poverty in Kenya showed no significant relationship between both processes. While a simultaneous increase of poverty and decrease of productivity was observed in western Kenya, an exact reverse interplay was identified in northwestern and southern Kenya. Based on five indicator groups different dimensions of marginality such as health, education, access to infrastructure and information but also economy could be analyzed. Indicator groups that represent accessibility to infrastructure or information showed significant higher correlation with poverty than any other indicator groups. Finally a set of eight indicators could be detected that explains decreasing productivity trends with the use of exploratory regression and ordinary least square regression (OLS). This includes : poverty rates, population density, percent of population with basic literacy, percent of the population attending higher education, local authority transfer funds (LATF), households with access to a landline, and rates of any fertilizer use. The analysis included data from all 47 counties of Kenya. Analysis of LUCC was also based on remote sensing using MODIS Land Cover Product (MCD12Q1) also with a spatial resolution of 500m. With this dataset croplands could be detected that were affected by LD. Based on these seven counties in western Kenya were identified also with regard to food security aspects : Trans Nzoia, Bungoma, Uasin Gishu, Kakamega, Siaya, Vihiga and Kisumu.

Présentation

Version intégrale (8,4 Mb)

Page publiée le 8 janvier 2019, mise à jour le 10 novembre 2021