Integrated system for detecting algal blooms in the Arabian/Persian Gulf: a multi-sensor approach using the Sentinel-2 and Sentinel-3 satellites data

Details

Supervisors

Hanert, Emmanuel

Faculty

Faculté des bioingénieurs

Degree label

Master [120] : bioingénieur en sciences et technologies de l'environnement, à finalité spécialisée

Abstract

Harmless and harmful algal blooms present significant threats to marine environments, economies, and the health of exposed populations. This is particularly critical in the Arabian/Persian Gulf, where such events are frequent. The region's scarce freshwater sources necessitate reliance on seawater desalination, making the establishment of an early warning system for algal blooms vital for the efficient management of desalination plants and other economic activities. In this master thesis, we explore the performance of a multi-sensor optical approach combining Sentinel-2 and Sentinel-3 images to detect algal blooms. For the first time, the study integrates the OLCI Terrestrial Chlorophyll Index (OTCI) to Sentinel-3 for algal bloom detection, and uses the Algae Bloom Detection Index (ABDI) with near-infrared (NIR) data from Sentinel-2. Through meticulous calibration, we simulated conditions for the year 2023, identifying risk zones across the Gulf with enhanced precision on specific dates when possible. Our findings demonstrate the efficacy of using OTCI and Sentinel-3 for daily detection of algal blooms and validate the use of the combination of ABDI and NIR for precision detection with Sentinel-2, highlighting well algal blooms convergent fronts. However, the method primarily highlights the area on the water surface, potentially underestimating the blooms' total extent. While further refinement is needed to enhance the model's spatial and temporal accuracy, this research represents a decisive first step toward developing an early detection system. Despite its limitations, the study holds promise for achieving more comprehensive and large-scale application in the future.