Seagrass and Seaweed Detection Approaches Using Remote Sensing and Google Earth Engine: A comparative Analysis of Different Machine Learning Techniques
Abstract
Seagrasses and seaweed habitats contribute to crucial ecological services globally, fromcapturing carbon dioxide and supporting 20% of the world’s largest fisheries to sustainingthe small, but many coastal communities [1]. Across the globe, an alarming decline intheir wild distribution has been recorded, attributed to climate change and direct pollution[2]. Current estimates of how much the loss is are uncertain and mapping and monitoringefforts are costly, data-intensive, and lack scalability. Thus, freely available data andsoftware in remote sensing, coupled with Machine Learning (ML) are deemed importantmeans to leverage existing mapping of seagrasses and seaweed spatial distribution [3, 4].This thesis explored a free and scalable workflow by comparing three different MLtechniques mainly on Overall Accuracy (OA) and Tau(e) in classifying seagrass, seaweed,and water. These are supervised, unsupervised, and semi-supervised learning (SSL) whichused data from the satellite, Sentinel-2 Level-2A, applied to a novel area of study, fromBiddeford Pool to Small Point at the Coast of Maine, United States of America. Resultsshowed that the SSL achieved the highest OA of 76% and Tau(e) = 0.72 on the hard test,in line with previous work. To the best of the author’s knowledge, this work contributesto the field of science by being the first in its field to use the geospatial analysis package’geemap’, along with the software Google Earth Engine, and SSL for classifying seagrassand seaweeds. Through demonstration, this work shows the potential of free data in remotesensing, leveraged by ML to aid community monitoring in the environmental managementof seagrass and seaweed. The results here can be considered as a starting point for furtherexploring the SSL paired with freely available data and community monitoring to lowercosts, handle data scarcity, and scale up in the field of aquatic vegetation mapping andmonitoring. Seagrasses and seaweed habitats contribute to crucial ecological services globally, fromcapturing carbon dioxide and supporting 20% of the world’s largest fisheries to sustainingthe small, but many coastal communities [1]. Across the globe, an alarming decline intheir wild distribution has been recorded, attributed to climate change and direct pollution[2]. Current estimates of how much the loss is are uncertain and mapping and monitoringefforts are costly, data-intensive, and lack scalability. Thus, freely available data andsoftware in remote sensing, coupled with Machine Learning (ML) are deemed importantmeans to leverage existing mapping of seagrasses and seaweed spatial distribution [3, 4].This thesis explored a free and scalable workflow by comparing three different MLtechniques mainly on Overall Accuracy (OA) and Tau(e) in classifying seagrass, seaweed,and water. These are supervised, unsupervised, and semi-supervised learning (SSL) whichused data from the satellite, Sentinel-2 Level-2A, applied to a novel area of study, fromBiddeford Pool to Small Point at the Coast of Maine, United States of America. Resultsshowed that the SSL achieved the highest OA of 76% and Tau(e) = 0.72 on the hard test,in line with previous work. To the best of the author’s knowledge, this work contributesto the field of science by being the first in its field to use the geospatial analysis package’geemap’, along with the software Google Earth Engine, and SSL for classifying seagrassand seaweeds. Through demonstration, this work shows the potential of free data in remotesensing, leveraged by ML to aid community monitoring in the environmental managementof seagrass and seaweed. The results here can be considered as a starting point for furtherexploring the SSL paired with freely available data and community monitoring to lowercosts, handle data scarcity, and scale up in the field of aquatic vegetation mapping andmonitoring.