After a recent study by Seagrass Watch, the vital global coastal ecosystems of seagrass, known for producing oxygen and absorbing carbon, are hurtling towards destruction. The monitoring of seagrass has always been a challenge due to the murky waters and changing tides. This wonderful tool with satellites and AI lets us observe and protect the underwater meadows much more easily.
Seagrass Mapping Reimagined: Unveiling AMOSS.
The researchers from Xiamen and Tulane University have come up with an algorithm called AMOSS: Automatic Mapping through integrating Optical and SAR images for intertidal seagrass meadows. This new technology was published in the Journal of Remote Sensing. The AMOSS algorithm directly uses the info from Sentinel-2/1 SAR and optical images to differentiate seagrass from other vegetation and overcome the tide-related changes. A Multi-Sensor Method for Decoding the Underwater Landscape
The AMOSS algorithm applies the tidal, biophysical, and spectral features of seagrass with great attention, distinguishing the land cover from intertidal mangroves, salt marshes, tidal flats, and even seawater. Important to note is the difference in scale between seagrass and other vegetation. While mangroves and salt marshes have upright stems or trunks, seagrasses completely lack such vertical structures. This is what helped the researchers create a model that detects seagrass by a backscatter coefficient decrease at VH polarization, meaning it stops short of its coastal neighbors.
Completing The Pinpointing The Meadows: Accuracy and Automation
For intertidal seagrass meadows, the AMOSS system applies the Otsu algorithm, which finds low water levels best suited for locating these meadows and automates the extraction process. Also, multi-binary classification approaches are employed to further enhance the automated mapping in diverse and intricate coastal areas.
Tracking Ecosystem Change: A Time-Lapse View of Resilience
The remarkable feature of AMOSS is its capacity to detect changes over time within seagrass ecosystems. Scientific application of seagrass loss and recovery patterns monitored through the Spectral Angle Mapper (SAM) method reveals profound understanding regarding the resilience of these critical habitats under stress.
Global Accuracy, Local Impact: A Scalable Solution
AMOSS has achieved remarkable accuracy with a global of 84% absolute accuracy verified in 15 diverse global study sites, including tropical and subpolar regions. This level of reliability and accuracy strengthens the system’s potential for widespread spread application.
Efficiency Through Automation: A Scalable Future
AMOSS does not use traditional supervised classification methods where sample selection (often subjective) is done manually. AMOSS’s automated approach serves as a cost-cutting boost and enables monitoring of extensive coastal regions achieved effortlessly.
AMOSS and the Struggle Against Climate Change’s Impact on Coastal Biodiversity
The research team notes that “the AMOSS algorithm marks an advancement in seagrass monitoring. By automating the mapping process, we can now accurately and efficiently track seagrass changes on a global scale. This is critically important for conserving coastal ecosystems, which are vulnerable to climate change.”
The above quote speaks to the work done for this research project; the excerpt is quite revealing as to the scope of AMOSS. Concerned parties are now able to instantaneously have access to appropriate data, brought forth courtesy of AMOSS. Automation is proving its utility on a large scale, not for mere company profit, but for conservation and the well-being of biodiversity. As AMOSS continues evolving, it will no longer be limited solely domestically or regionally; internationally, it will soon be able to monitor and retrieve data of seagrass globally, turning AMOSS into a universal system for seagrass monitoring, capable of integrating into worldwide biodiversity programs and strategies formulated to tackle climate change.
