The Estimation and Evaluation of Shoreline Locations, Shoreline-Change Rates, and Coastal Volume Changes Derived from Landsat Images

Do, A.T.K.; de Vries, S., and Stive, M.J.F., 2019. The estimation and evaluation of shoreline locations, shoreline-change rates, and coastal volume changes derived from Landsat images. Journal of Coastal Research, 35(1), 56–71. Coconut Creek (Florida), ISSN 0749-0208.

Shoreline-change data are of primary importance for understanding coastal erosion and deposition as well as for studying coastal morphodynamics. Shoreline extraction from satellite images has been used as a low-cost alternative and as an addition to traditional methods. In this work, satellite-derived shorelines and corresponding shoreline-change rates and changes in volumes of coastal sediments have been estimated and evaluated for the case of the data-rich North-Holland coast. This coast is globally unique for its long in situ monitoring record and provides a perfect case to evaluate the potential of shoreline mapping techniques. A total of 13 Landsat images and 233 observed cross-shore profiles (from the JAaRlijkse KUStmeting [JARKUS] database) between 1985 and 2010 have been used in this study. Satellite-derived shorelines are found to be biased in seaward direction relative to the JARKUS-derived shorelines, with an average ranging 8 m to 9 m over 25 years. Shoreline-change rates have been estimated using time series of satellite-derived shorelines and applying linear regression. The satellite-derived shoreline-change rates show a high correlation coefficient (R² > 0.78) when compared with the JARKUS-derived shoreline-change rates over a period of 20 and 25 years. Volume changes were calculated from the satellite-derived shoreline-change rates using assumptions defining a closure depth. Satellite-derived volume changes also show a good agreement with JARKUS-based values. Satellite-derived shorelines compare better with in situ data on beaches that have intertidal zone widths ranging from one- to two-pixel sizes (30 m–60 m). The results show that the use of Landsat images for deriving shorelines, shoreline-change rates, and volume changes have accuracies comparable to observed JARKUS-based values when considering decadal scales of measurements. This shows the potential of applying Landsat images to monitor shoreline change and coastal volume change over decades.