Abstract:
This study examines the use of Geo-spatial techniques to model and predict urban growth of
Hossana town between 2003 and 2018 so as to detect and analyze the change that has taken in
the town between these periods. In order to achieve these the remotely sensed Google Earth
Image of 2003, Ortho-photo of 2013 and 2018 have been obtained and pre-processed using
EARDAS IMAGINE. Supervised image classification has been used to generate land use/land
cover maps. Land use/land cover classification, accuracy assessment, and change and prediction
map in ERDAS IMAGINE and IDRSI software. For the accuracy of the classified LULCC maps
the confusion matrix was used to generate it in standard form. The overall accuracy and kappa
coefficient results were 89.50% for 2003 and 92.46% for 2018 which had been above the
minimum and acceptable threshold level. The result revealed that annual aggregate rate of
changes of urban growth of Hosanna town has been occurred within 15 years from 2003-2013
and 2013-2018 were 135.35% and 121.36% respectively. Though the period of 2003 from 2018
there were dramatic change in built-up areas, whereas agricultural, open, and forest land has
decreased in area. Accordingly, more land was occupied under built up. The results also shows
that increase in built-up coverage of the town was due to population pressure, road
infrastructure network development, and other physical and proximity factors on land. So the
rate of change has been dynamic and it needs also infrastructure and other facilities
development in the town. Generally, spatial analysis and modelling enables for sustainable
managements of urban growth through proper planning, wise decision making, monitoring of
urban expansion and development
