Abstract
The paper is aimed at the accuracy assessment of coastline extraction from optical imagery in studying the short period dynamics and development trends of the coastal zone of the Sambia Penin-sula using high- and ultra-high resolution satellite imagery. The paper presents the methodological issues relating to the studies of the Sambia Peninsula coastal dynamics using field data and optical images. It gives the error estimates for satellite remote sensing data based on comparison of those data with the data of field measurements of the shoreline coordinates. It has been established that the accuracy of shoreline extraction from optical image could be that of only several metres. It is shown how the Change Detection algorithm can be used to detect spatio-temporal coastline changes on multi-temporal images. The analysis of the quantitative assessment of changes in the area of erosion and accretion on the Sambia Peninsula coast during the research period of 2010-2017 showed the prevalence of abrasive processes over accumulative ones, which has also been confirmed by field observations during coastal zone surveys conducted after storms in 2012-2017. The prevailing trend on the greater part of the peninsula over the last five years has been established as coastal retreat of 6-36 m on the average.