Integration of Drone-based Imaging for Coastal Ecosystem Mapping: A Case Study of Shallow Coastal in Phang Nga Bay, Thailand

Abstract

Efficient management of marine and coastal resources, particularly in shallow areas with sensitive ecosystems such as coral reefs and seagrass meadows, requires continuous and rapid monitoring. These ecosystems, which provide critical habitats, are increasingly vulnerable to climate change, highlighting the need for non-invasive, systematic assessments to support conservation efforts. This study explores the feasibility of using multispectral imaging-equipped unmanned aerial vehicles (UAVs) as an effective alternative to traditional survey methods for evaluating marine resources. Focusing on shallow coastal ecosystems in Phang Nga Bay, Thailand, the study examines methodologies, processes, and data analysis techniques involved in using UAVs for coastal ecosystem mapping. Surveys were conducted in three distinct sub-areas of Phang Nga Bay, each characterized by unique land cover types. The first area, north of Koh Khai Yai, included six land cover types, with underwater rock and coral serving as key indicators of ecological richness. The second region beneath Koh Khai Yai featured seven classifications, depicting both natural and human-influenced elements. The third region, on Koh Khai Nui, consisted of four cover types with distinct rock formations. By comparing UAV-derived data with existing marine resource databases, the study demonstrated the ability of UAV multispectral imaging to deliver high-resolution, detailed data for resource distribution. The results confirm that integrating UAV-based imaging with the ISODATA unsupervised classification method offers a fast, up-to-date, and efficient approach to resource monitoring, reducing ecological disruption and enhancing conservation management for fragile coastal ecosystems.