A new water index proposal for coastline extraction from Landsat 9 OLI images

The extraction of the coastline from remotely sensed images is fundamental for many operational fields, e.g., coastal zone management, environmental monitoring and oceanography. Among different approaches available to support the process for coastline extraction from remotely sensed images, methods based on the use of indices aimed at distinguishing water from no-water play a very important role as evidenced by the numerous applications described in the literature. This article proposes a new index to highlight the difference between the two considered classes (water and nowater) and therefore extract the coastline from Landsat 9 OLI images. The new index, called SWIR Minus Blue Index (SMBI), produces one layer which, submitted to Maximum Likelihood Classification (MLC), allows to clearly distinguish the presence of water in the analyzed scene and to extract the coastline as a relative boundary. The applications are conducted on the north-eastern coast of the Gulf of Salerno (Italy). The results demonstrate the effectiveness of the proposed method compared with the performances of other four indices present in the literature and applied in this study.