@ARTICLE{Khastar-Broujeni, author = {Taherian, Elham and khastar-Broujeni, milad and Samadi, Hossein and }, title = {Comparison of sub pixel MNDWI and AWEIshadow indices capability for shallow and narrow river extraction}, volume = {5}, number = {2}, abstract ={Preservation of large dams as one of the energy and water suppliers is of utmost importance. Sediments carried by the rivers to the reservoir can reduce the dam's useful lifetime. Construction of engineering structures in the most erodible sections can control entrance sediments flow to the dams. Locating high-risk areas for construction of engineering structures will be possible with river morphology, by means of remote sensing technology and extract river boundary using water index. Monitoring of the narrow and shallow rivers using multispectral and medium spatial resolution data as the oldest archived data is facing to the most challenges among the various water bodies. In this study sub-pixel capability of the most efficient water indices include MNDWI and AWEIshadow extracted OLI sensor Landsat was investigated by accuracy assessments statistics such as ROC curves, user accuracy, producer accuracy and commission and omission errors. The results of accuracy assessments statistics for water pure pixel showed the effectiveness for both indices, but in the case of mixed pixels, MNDWI was earned more accuracy than AWEIshadow. A large portion of commission error occurred for MNDWI was related to Rocky protrusions and for AWEIshadow to topographic shadows. Finally results shown that enhancement methods as suplementary method for commision error correction can be useful to river boundry extraction }, URL = {http://jgit.kntu.ac.ir/article-1-471-en.html}, eprint = {http://jgit.kntu.ac.ir/article-1-471-en.pdf}, journal = {Journal of Geospatial Information Technology}, doi = {10.29252/jgit.5.2.99}, year = {2017} }