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

Titre : Landscape change dynamics in a semi-arid part of Baringo District, Kenya, based on Landsat-TM-Data and GIS analysis

Auteur : Benjamin Mwasi

Université de soutenance : Universiteit van Amsterdam

Grade : Doctor 2004

Résumé
Thee study area, a semi-arid area in the Baringo district of Kenya, is inhabited by communitiess that mostly derive their livelihood directly from the environment, through subsistencee agriculture and pastoralism. Consequently, the landscape in this area is managedd with two major goals. The primary, and most compelling, goal being maximisationn of the production function of the landscape. The second goal, which is oftenn seen as secondary, is related to landscape’s information carrier function, where conservationn of both biotic and abiotic resources for future use is considered to be important.. Ideally, these objectives should be complementary for sustainability. However,, in practice they tend to be competitive at best, otherwise conflicting. With properr planning and management it is possible to approach this ideal situation. This study investigatedd the possibilities of reducing conflicts between these objectives by increasing thee understanding of the landscape change dynamics related to human exploitation of the naturall resource base. Thee study recognises the fact that landscape changes as a result of human utilisationn of natural resources is inevitable. However, land degradation or continuous loweringg of landscape quality, can be avoided through planning and management of landscapee changes. Effective management requires, among other things, a) knowledge of thee spatial distribution of resources, b) ability to monitor changes in quantity and quality off these resources, c) knowledge of the factors causing or controlling these changes and, d)) an understanding of how these factors interact to cause such changes. Satellite remote sensingg and image processing software are capable of providing accurate and up-to-date informationn relevant for locating and monitoring the quality of resources. By combining temporallyy distributed land resource data with similar spatially referenced socio economicc data in a geographic information system, the relationship between the two is explored.. This relationship is analysed further using multivariate statistics to identify the mostt important drives of land cover changes. Armed with this knowledge and appropriate spatiall dynamic models the process of landscape change is modelled. The resultant informationn can enable planners to predict future change patterns and also provide managerss and policy makers a means for directing or controlling land cover changes throughh deliberate introduction or elimination of specific change driver

Annonce de la thèse

Version intégrale