Multi-Temporal Mapping of Seagrass Distribution by using Integrated Remote Sensing Data in Kung Kraben Bay (KKB), Chanthaburi Province, Thailand

Abstract

The seagrass beds are a unique marine productive ecosystem that provides a shelter, a food source for the marine community of animals and act as a biofilter in marine environments. The seagrass situation shows the current number of seagrass beds have been continuously decreasing in Thailand. The study presented the comparison of a high-resolution satellite imagery and aerial photograph by Unmanned Aerial Vehicle (UAV) to change detection in Kung Kraben Bay between 2011 and 2017. Study area was a 5.59 km², shallow (depth 2.5 m) and clear water in the Tha Mai district, Chanthaburi province, Thailand. The WorldView-2, GeoEye-1 and aerial photograph by UAV were composited to the Normalized Difference Vegetation Index (NDVI) image and classified to 3 classes such as a long seagrass leaves type (Enhalus acoroides), short seagrass leaves type (Halodule pinifolia and Halodule uninervis), and another object. The visual interpretation with in situ data and supervised classification technique assisted to seagrass detection in a very high-resolution image. The classification results showed that visual interpretation with in situ data that the overall accuracies and Kappa coefficients were higher than supervised classification with maximum likelihood such as 74.42% and 0.568, respectively. From 2011 to 2017, the total area of seagrass distribution had not changed, but the seagrass density had changed in some areas. The resultant maps provided a changing of landscape-scale seagrass dynamic data and the advantage of an aerial photograph by UAV for seagrass detection in a shallow water environment.