The Seasonal Change Detection of Seagrass Extent using Google Earth Engine

Abstract

Seagrass ecosystems in eastern Indonesia play a crucial role in blue carbon storage, slowing climate change by absorbing and storing carbon in their sediments for decades. Nevertheless, seagrass meadows are under high pressure due to human activities. Since the calculation of emission factors requires data on the area of seagrass change, spatial time series data are necessary. One of the consistent time series of spatial data is the Sentinel-2 satellite imagery available on the Earth Engine data catalogue. This study aims to map seasonal changes of the seagrass bed at Makassar using Google Earth Engine. The image classification results from 2016 to 2024 data using Random Forest show that the area of seagrass beds in the small islands of Makassar tends to increase by about 3% during the dry season. Kodingarenglompo Island has the largest seagrass bed, with a variation of approximately 68 hectares in the dry season to 57 hectares in the rainy season. The average seagrass area on this island has a downward trend over the past eight years, as does Barranglompo Island. In contrast, Barrangcaddi Island, Langkai and Bonetambung show an increasing trend in seagrass in 2024 compared to 2016.