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

Accueil du site → Doctorat → Royaume-Uni → 2008 → Assessment of Land Cover Change in North Eastern Nigeria

Cranfield University (2008)

Assessment of Land Cover Change in North Eastern Nigeria

Garba, Samuel Sule

Titre : Assessment of Land Cover Change in North Eastern Nigeria

Auteur : Garba, Samuel Sule

Université de soutenance : Cranfield University

Grade : Doctor of Philosophy (PhD) 2008

Résumé
Land cover change provides a means of understanding and managing the problems of degradation and shortage of land and water resources and the conflicts therewith in the north eastern Nigeria. This research assessed how tree, shrub grass, bare ground changed from 1986 to 2005 using the NigeriaSat-1 and Landsat images calibrated with field survey data. Thirteen subclasses of the land cover were spectrally analysed and classified severally, however uncertainties in the classification made the merger into four classes necessary. Changes were analysed according to persistence, swapping, loss and gain analysis, multi-year transition of each land cover in succession, location of intensive change, and regional change density. Uncertainties were analysed by confusion and transition error matrices. The overall accuracies of the classifications were between 60% and 75%, and the transition and change accuracies were between 45% and 60%. Approximately 60% of the area of study remained unchanged during the period. Of the remainder, approximately 11% of the area interchanged between shrub grass and bare ground. Shrub grass was found to be the most unstable category and the source of most misclassification. The loss of tree was general but more intensive in the Fadama making it the most vulnerable. How local people perceived land cover change was sought through group interview and the results concurred generally with the assessment of the changes. NigeriaSat-1 imagery was tested for its quality and whether the addition of the middle infrared wavebands improved the classification. NigeriaSat-1 failed to classify the 13 classes and the middle infrared did not improve the classification, thus comparable to Landsat data, although the test was done with dry season images and the result may likely be different for wet season imagery. The 8 km AHVRR-NDVI was found to be useful in assessing the timing of image acquisition, but the data could not provide sufficient spatial resolution to warrant its usage for local scale studies.

Présentation

Version intégrale (22,34 Mb)

Page publiée le 19 octobre 2018