PREDICTING URBAN SPRAWL PATTERNS AROUND ELDORET TOWN USING REMOTE SENSING AND GIS TECHNIQUES

ODHIAMBO, SAMSON (2022-07)
xmlui.dri2xhtml.METS-1.0.item-type
Thesis

One of the most rapidly growing urban phenomena in the 21st Century is emergence of sprawling settlements. Such settlements provide essential services but also cause some strain on these centers. Controlling and managing the growth of such settlements is necessary in order to maximize the services they provide and mitigate against the stresses they cause. This requires an understanding of both spatial and temporal sprawling patterns and factors that control them with a view to enabling planners to predict, guide and minimize patterns of urban sprawls. Satellite data plays a vital role in studying urban growth and sprawl patterns. However, despite most sensors delivering medium and high resolution satellite imageries and development in computer technology, a great percentage uses pixel based classification techniques. These techniques overlook variations in soft classifiers, sub-pixel classifiers and spectra un-mixing. This study therefore used classifiers that combine spectral and structural characteristics, involving rule-based object classification, to determine urban sprawl patterns around Eldoret Town and isolate factors that governs or controls the patterns in order to use them in predicting possible future sprawl patterns around Eldoret town in 2029. The findings revealed that different sprawl areas have different patterns for example Kapseret, Maili nne, Jua Kali showed linear patterns while Soy, Kuinet, Kipkorgot showed leapfrog pattern and attracted by economic factors during 2000-2020. Five of the selected factors, that is, distances to roads, powerline, waterline, employment centers and population density were the most significant factors contributing by 84.12% to sprawling patterns while three of the factors, that is, distance to restricted areas with 6.02%, elevation and slope tying with 4.94% each contributing least to sprawling patterns. The CA-Markov chain and AHP models predicted that sprawl areas would take different sprawl patterns of linear and leapfrog, increasing from 138.91 km2 to 154 km2 during 2020 to 2029, respectively. The study recommends that Uasin Gishu County government should direct areas destined for development by supplying roads, energy and water and control these services in areas that are not in order to control urban sprawl.

Éditeur
University of Eldoret
Collections:

Preview

Nom:
SAMSON ODHIAMBO.pdf



Fichier(s) constituant ce document

Thumbnail
Thumbnail

Les fichiers de licence suivants sont associés à ce document :

Attribution-NonCommercial-NoDerivs 3.0 United States
Excepté là où spécifié autrement, la license de ce document est décrite en tant que Attribution-NonCommercial-NoDerivs 3.0 United States