Automatic detection of shoreline change: case of North Sinai coast, Egypt

摘要

This paper introduces an appropriate visualization of how to effectively digitize, quantify and predict shoreline kinematics changes. The method relies mainly on the coupling of Geographic Information System (GIS) with Digital Shoreline Analysis System (DSAS). The North Sinai coast in Egypt is selected as a case study. The proposed technique is applied over a quarter-centennial period of 27years (1989-2016). However, the years 2025, 2035, and 2050 are used for prediction purpose. Histogram threshold of band 5, Histogram threshold of band ratio, and Tasselled Cap Transformation (TCT) are initially tempted as semi-automatic shoreline extraction techniques for Landsat ETM 2010 imagery. Among of them, the TCT is found superior as a digitizing technique that attains the least normalized root mean square errors with the corresponding field data in 2010. Meanwhile, the shoreline change rates in the form of erosion/accretion patterns are automatically quantified by four statistical parameters functioned in DSAS coding. Those, namely end point rate (EPR), net shoreline movement (NSM), linear regression rate (LRR), and least median of squares (LMS). On the basis of the LRR and EPR results, this study offers to the coastal managers a highly reliable decision-support-algorism that can dynamically assist in elaborating strategies to curtail the non-affirmative consequences due to the erosion/accretion of the shoreline.