Abstract:

Coastline dynamic change is a common reflection of global change and human activities. Based on seven Landsat remote sensing images and Google Earth high-resolution images of the Guangdong-Hong Kong-Macao Greater Bay Area, Tokyo Bay and San Francisco Bay from 1980 to 2020, this paper used threshold segmentation combined with water index method and Sobel operator method to extract the coastline of each analysis period. In addition, this study also analyzed the changes of coastline in terms of length, spatial morphology, structure and utilization degree, and analyzed qualitatively and quantitatively the influencing factors with the help of geographic detectors. The results show that: (1) from 1980 to 2020, the intensity of annual average length change in the Tokyo Bay is the largest, which is 0.37%, and the coastline tends to be straight; (2) in the past 40 years, the depth of coastlines in all analyzed bay areas has remained stable, and the structures tend to be complex and the morphology tends to be dispersed. Among them, the difference of fractal dimension of coastlines in the Guangdong-Hong Kong-Macao Greater Bay Area is the smallest and the morphology is the most dispersed; (3) during the past 40 years, the natural coastline of each bay area decreased, while the length of port wharf coastline and other artificial coastline increased significantly, and the biological coastline of the Guangdong-Hong Kong-Macao Greater Bay Area fluctuated, and the coastline utilization index increased the most. The results show that temperature, wave height, tide, area and port throughput are the main factors that affect the coastline change, the interaction of any two factors is greater than the single factor.

Key words: bay area, coastline, coastline type, index of coastline length change, utilization index