Analysis of Mucilage Levels Build Up in the Sea of Marmara Based on Unsupervised Unmixing of Worldview-3 Data

Abstract

<p><span style="color: rgb(51, 51, 51); font-family: &quot;HelveticaNeue Regular&quot;, sans-serif; font-size: 18px;">Increasing uncontrolled industrialization causing untreated wastewater and marine traffic, coupled with specific weather conditions makes the Marmara Sea face major challenges, such as the mucilage outbreak in the summer of 2021. Cleaning up the Sea of Marmara will continue to be critical to the long term survival of the fishing industry and tourism, and new methods and techniques for monitoring its behaviour will be important for benchmarking. In this specific context, this paper discusses the potential of using unmixing approaches for the analysis of multispectral images in order to understand and map the abundance of mucilage coverage, in connection with its chemical compositions related to varying levels of aggregation. A WorldView-3 multispectral image, acquired on May 25, 2021, was chosen for the study due to its high spatial resolution and VIS/NIR spectral region. Unmixing provides the spectral characteristics of endmembers (i.e., water, ship and different types mucilage), and the abundance associated with each endmember. Preliminary results were discussed and some research aspects for understanding mucilage phenomena using EO were put forward.</span><br></p>