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

Accueil du site → Doctorat → Royaume-Uni → 1996 → The application of high spatial resolution satellite data for land use and land cover mapping and inventory in north east Nigeria

Cranfield University (1996)

The application of high spatial resolution satellite data for land use and land cover mapping and inventory in north east Nigeria

Lawan, A.I

Titre : The application of high spatial resolution satellite data for land use and land cover mapping and inventory in north east Nigeria

Auteur : Lawan, A.I

Université de soutenance : Cranfield University

Grade : Doctor of Philosophy (PhD) 1996

Résumé
The acquisition of information about the land resources in semi-arid North East Nigeria is an important issue due to the need to plan for sustainable land resources utilization. This is becoming increasingly important in the face of the declining land quality in the region. SPOT multispectral and Landsat Thematic Mapper satellite images have been compared in terms of their effectiveness for land use and land cover mapping and inventory. Several image processing techniques were applied to examine the impact of such techniques on classification accuracy. Field survey based on stratified random sampling was carried out between September 1992 and August 1993 in North-East Nigeria in support of the remote sensing data. The analysis of the relative efficiency of the stratification over random sampling showed that for floodplain based land use and land cover classes, the precision of the ground survey was enhanced. Multitemporal ground data were acquired, computer processed and analysed for spatial and temporal changes. Results of the analysis showed significant inter-class changes between most of the classes.Due to lack of a suitable existing classification scheme, a new land use and land cover classification was developed. The 22-class scheme was evolved after one year of field survey.The overall classification accuracy of the initial 21 classes gave an average of 50% due to spectral overlap between classes. Subsequently the classes were combined into 6 along spectral and land cover similarity. This improved the overall classification accuracy to an average of 65%.The ground data were used in combination with the satellite data to produce area estimates of the land use and land cover classes.The present study will be useful for the development of satellite remote sensing for land resources mapping and inventory in the North East Arid Zone Development Programme (NEAZDP) in North East Nigeria.

Proquest Dissertations & Theses

Page publiée le 2 mars 2015, mise à jour le 18 octobre 2018