Using a transect-focused approach to interpret satellite images and analyze shoreline evolution in Haiyang Beach, China

Abstract

This paper devises a set of simple and practical method for interpreting hundreds of free medium-resolution satellite images to study beach evolution. The method focuses on beach transects, integrates the computer-aided and subpixel recognition technologies, and derives multiple shoreline indicators and historical beach profile shape. This paper also introduces how to use the method with a case study. The case study uses 880 Landsat images from 1984 to 2020 to examine the historical evolution of the complex Haiyang Beach in China as well as its responses to onshore and offshore engineering projects.The results not only show the detailed temporal and spatial changes of the upper beach based on dry/wet and high-water lines, but also the changes of the middle and lower beach based on mean-water, low-water, and low boundary lines, as well as the changes of beach profile shape, beach slope, and intertidal width. The random errors of the dry/wet, high-water, mean-water, and low-water lines are 7.8 m, 10.5 m, 15.9 m, and 19.2 m, respectively. The random errors not only result from the geo-reference process of satellite images and the recognition process of shoreline positions, but are also affected by random water level fluctuations and irregular undulations of the beach profile. This paper also discusses the similarities, differences, and applicability of these shoreline indicators. It is found that the evolution of dry/wet and high-water lines are similar and mainly affected by onshore engineering projects while the evolution of mean-water, low-water, and low boundary lines are similar and mainly affected by offshore engineering projects.The method provided in this paper can comprehensively evaluate beach evolution based on free satellite images, and would be helpful for understanding and protection of the beaches affected by intensified or intensifying engineering projects and human activities.