Student researchers in the UK have been learning how to use Copernicus Sentinel data for applied projects in Marine Science, in areas such as automated vessel traffic monitoring, detecting belts of sargassum seaweed, and tracking internal waves that protect coral.
Freely available data from the Sentinel-1 and Sentinel-2 missions of the EU’s Copernicus programme were the inspiration for second-year research projects in an introductory course to Marine Remote Sensing, at the School of Biological and Marine Sciences at the University of Plymouth, UK.
Second-year students following an introductory course in Marine Remote Sensing, at the School of Biological and Marine Sciences at the University of Plymouth, UK.
Copyright:University of Plymouth
What better way to teach the boundless possibilities of Earth observation than by using remote sensing data to do real science? The aspiring researchers were given only three weeks to complete their projects, where they learnt to use Sentinel data in an applied approach. The results were impressive – with three of the 12 projects showcased here.
In the Atlantic and Caribbean, sargassum is an invasive seaweed with widespread negative impacts. When it blooms, it can form enormous floating mats that extend over kilometres. Using free online neural network object detection models and Sentinel-2 data, a team of young researchers showed that it is possible to automate the monitoring of Sargassum invasions.
Sentinel-2 imagery were used to train neural models to detect Sargassum invasions. Sargassum mats are shown here using a False Colour filter, and these can be seen as kilometre-long red blobs floating towards Barbados.
Copyright:Copernicus/ESA
Jack Allen-Davis, Brandon Callican, Samuel Lea, Alex Mallinson and Aslak Varela used Sentinel-2 data acquired over the Caribbean Sea during the 2018 Sargassum invasion.
False colour images from the Sentinel-2 MultiSpectral Instrument were downloaded using Copernicus Browser and then uploaded to the Roboflow platform, to train three versions of YOLOv8 object detect ion and classification computer vision models.
Incremental improvements to accuracy were made by increasing the number of images for training. The best model of the three was run at a 20% confidence level and proved successful at discerning sargassum mats and windrows from cloud.
Monitoring vessel traffic in the Suez Canal
In March 2021, one of the largest container ships ever built, the Ever Given, became stuck in the Suez Canal for nearly one week. The resulting ‘traffic jam’ cost billions in trade per day, and according to Maritime Agency, Leth, by the time the Ever Given was freed on 30 March, a record 171 vessels were queued up in the Gulf of Suez in the Red Sea.
To assess the accuracy of this count, young researchers Poppy Firth, Emily Robinson and Eleanor Smith, trialled an automated object detection machine learning algorithm.
Using freely available Synthetic Aperture Radar (SAR) data acquired by the Copernicus Sentinel-1 satellite, the team demonstrated the value of the data and machine learning methods for automated vessel traffic monitoring.
Dual polarised Sentinel-1 SAR data show vessels in the Suez canal, discernible as bright spots.
Copyright:Copernicus/ESA
In Roboflow, vessels were detected and counted in 10 SAR images collected between 21 March and 08 April 2021, using the YOLOv8 convolutional neural network. Sentinel-1 SAR data in dual polarisation were used to serve increase signal strength, while also reducing speckle noise and star-shaped artefacts.
Dual polarisation also enabled the vessels to be clearly discernible from water, showing visibly more uniformity in shape, ensuring robust training imagery for the automated object classifier. The model counted up to 224 vessels queued in the gulf – 53 more than reported by Leth.
Identifying internal waves that protect coral
Internal waves are large amplitude gravity waves caused by variations in ocean water density. They propagate at the interface between a layer of warm water and a layer of cooler water, such as the pycnocline, about 40 m below the surface.
Internal waves in the Andaman Sea, west of Thailand, have been shown to reduce coral bleaching, especially during high Sea Surface Temperature (SST) events. A group of Plymouth researchers - Olivia Cox, Lucy Bryan and Anya Kirkbride - were inspired to use Sentinel-1 SAR imagery to detect the presence of internal waves in the region and investigate their protective effect on corals.
Using EO Browser to visualise 45 Sentinel-1 SAR images collected from January to December 2019, the team detected internal waves off the Similan Island chain. Optical data from Sentinel-2 were used to validate these detections.
Identification of internal waves using surface backscatter data collected by Sentinel-1 over Andaman Sea, with a focus on the Similan Island chain to the West of Phuket, Thailand.
Copyright:Copernicus/ESA
The researchers accessed the 40-year ‘Global Ocean OSTIA Sea Surface Temperature and Sea Ice Reprocessed’ timeseries (1981 to 2022), from the Copernicus Marine Service, to identify a clear warming trend in the area.
Highest temperatures generally coincided with the start of the hot wet season in April, exactly when internal waves were shown to be largely absent. The results suggest that while internal waves have been protecting coral in the Andaman Sea, these positive impacts may be disrupted over time due to rising SSTs.
Dr. Lauren Biermann, Marine Remote Sensing Lecturer at the University of Plymouth, says, “I started this role in September last year with the aim of sharing my boundless enthusiasm for Earth Observation satellite science, and teaching the skills needed to use such data through an applied approach. To this end, I tasked the second-year students in my OS209 ‘Introduction to Marine Remote Sensing’ module with research projects that they would have to start, finish, and write up in 3 short weeks. This was no small challenge, but they still managed to produce inspired work!”
About the Copernicus Sentinels
The Copernicus Sentinels are a fleet of dedicated EU-owned satellites, designed to deliver the wealth of data and imagery that are central to the European Union's Copernicus environmental programme.
The European Commission leads and coordinates this programme, to improve the management of the environment, safeguarding lives every day. ESA is in charge of the space component, responsible for developing the family of Copernicus Sentinel satellites on behalf of the European Union and ensuring the flow of data for the Copernicus services, while the operations of the Copernicus Sentinels have been entrusted to ESA and EUMETSAT.
Did you know that?
Earth observation data from the Copernicus Sentinel satellites are fed into the Copernicus Services. First launched in 2012 with the Land Monitoring and Emergency Management services, these services provide free and open support, in six different thematic areas.
The Copernicus Marine Environment Monitoring Service (CMEMS) provides regular and systematic reference information on the physical and biogeochemical state, variability and dynamics of the ocean and marine ecosystems for the global ocean and the European regional seas.