Valley of Southwest Montana, USA. The technique was very useful
in detecting a variety of wetlands and riparian zones present in
the landscape.used high resolution SPOT-5 and Landsat Thematic
Mapper (TM) images to detect and classify wetlands employing
a decision tree technique, their findings revealed that combination
of different knowledge and techniques is an effective and
promising method of identification and classification of wetlands.Swaziland and Ethiopia,
respectively. The wetlands covered in these studies were, however,
larger and techniques used also varied from the current study.
Although there is increasing importance of monitoring wetland
and their dynamics based on optical and radar satellite datasets,
only some few studies have investigated the potential of very high
to high resolution optical datasets for the detection of small wetlands
in the dry savanna areas of East Africa (Haack, 1996; Kashaigili
et al., 2006; Owor et al., 2007). Therefore this study is needed tocontribute in improving our understanding on (1) delineation of
small wetlands using high to very high resolution optical datasets
and on (2) mapping of wetland’s natural vegetation and agricultural
use patterns by applying adapted classification methods.