University of Twente Student Theses


Analysis of forest degradation in Miombo Woodlands using remote sensing and GIS: Case study of Kitulangalo Forest in Morogoro, Tanzania

Karenget, Job Araba-son (2009) Analysis of forest degradation in Miombo Woodlands using remote sensing and GIS: Case study of Kitulangalo Forest in Morogoro, Tanzania.

[img] PDF
Abstract:Forest degradation refers to changes within the forest which negatively affect the structure or function of the stand or site, and thereby lower the capacity to supply goods and services The use of remotely sensed data in forest degradation becomes an essential means in determining the extent and density of forest degradation and at the same time forest management options sought Monitoring forest degradation can be achieved through repetitive measurements of biophysical attributes that characterize land cover and should be surface characteristics measured in space for example vegetation cover and biomass. This study aims at analysis of the degradation of part of Miombo woodlands in Tanzania using Remote Sensing and GIS. Forest inventory data was done in which biomass and volume were determined using allometric equations of tree diameter. Regressions were used to relate vegetation indices NDVI, MSAVI, SR and PVI with stand parameters. Kriging interpolation was done to spatially analyze using indicators of forest degradation. Maximum likelihood classification SPOT-5 image was performed. The spectral reflectance and vegetation indices from SPOT-5 were poorly correlated to stand parameters and could explain only 13% of their variance. While vegetation indices NDVI, MSAVI, SR and PVI showed a fair correlation of up to 47%. Forest degradation indicators identified in Miombo woodland were: basal area, volume, stem density and biomass Forest parameters were decreasing in the public land while increasing in the forest reserve for the period between years 2004 and 2008. Spatial distribution of forest degradation indictors indicated that lower values were concentrated within the middle part of the area. Gradual increases of medium values are realized towards the middle part from both the north and south part. The trends of forest parameters increased in the reserve as opposed to the public land. Spatial distribution of forest parameters indicated the extent of degradation coupled with classification of remotely sensed data. Although the potential ground based biomass estimation combined with remotely sensed data proved difficult, other practical works can provide adequate and reliable information regarding biomass resource particularly firewood through biomass equations. Such information will guide decision making process in order to provide sound policies for protection and conservation of natural woodlands. Understanding different rates of ecosystem change growth in biomass is important in developing accurate estimates of factors contributing to atmospheric concentration in carbon dioxide and other green house gases (GHG). Key words: degradation, kriging interpolation, forest parameters, vegetation indices, remote sensing.
Item Type:Essay (Master)
Faculty:ITC: Faculty of Geo-information Science and Earth Observation
Programme:Geoinformation Science and Earth Observation MSc (75014)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page