Progress, Challenges and Implications of Land Use/Land Cover Detection Methods in Ethiopia: A Review

Abstract

Remote sensed imageries are rich in geospatial data’s pertinent for natural resources conservation. However, extracting accurate and reliable information was remained critical. This review is aimed to compile the progress, challenges and examine implications of land use/land cover detection methods executed in Ethiopia. In the diverse landscape of Ethiopia, existing satellite image classification techniques were operated. Most of the studies agreed that automatic techniques are crucial in detecting the spectral responses of features. However, it was limited under heterogeneous landscape. To retain its digital recognition, a successive hybrid of the automatic techniques were also executed then again doubted by its limitation on areas which have similar reflectance for different land covers. Other studies also applied on-screen-digitizing, which bears better accuracy but criticized as it was exhaustive, costly, and expert dependent on fine-resolution data. In order to reduce the limitations and incorporate their advantages, a hybrid of automatic and on-screen-digitizing has been effected. Even though this technique refined spectral confusion sourced from automatic operation with better accuracy, the drawback of on-screen-digitizing was followed. Indeed, if an error happen in data processing, it is obvious that decisions and modeling outputs could be doubtful. In so far, those studies showed virtuous progress in adopting the tools to Ethiopia. Nevertheless, the absence of well-organized, accessible, and up-to-date information catalog, the country was investing (in/directly) for those fragmented studies. Devising site-specific methodologies, providing accurate inputs for modeling and decision makers, organizing fragmented studies, and establishment of an accessible resource assessment database are recommended.