The UK Hydrographic Office (UKHO) is launching the world’s highest resolution, most up-to-date global view of the world’s critical mangrove forests: the ADMIRALTY Global Mangrove Data Set.
Mangrove forests are a collection of saltwater tolerant plants that grow in tropical coastal environments, providing biodiversity and carbon sequestration. They are a crucial part of coastal ecosystems and act as natural barriers protecting human communities and infrastructure against the impacts associated with rising sea levels, larger waves, and intensifying storms - among other ocean-related natural phenomena.
Acting as a natural carbon sink, mangroves are known to sequestrate 3-5 times more carbon than inland tropical forests and contain the highest carbon density of all terrestrial ecosystems, potentially making them a vital component of the fight against climate change.
The new data set, which will be updated proactively and automatically, is created by machine learning being applied to global satellite imagery. From this imagery, UKHO has been able to create a data set showing the extent of global mangroves; the first step to identifying where these vital carbon sinks could be under threat.
Mangroves are also important features for navigation, situational awareness, and coastline and marine sustainability planning. The ADMIRALTY Global Mangrove Data Set aims to provide timely and accurate information on mangrove presence to mariners and the UK and overseas governments. The data set has been launched as a beta service. This means that we will be continually testing and improving it, ensuring it fulfils user needs.
A new level of precision
A vastly improved level of precision and reliability compared to existing mangrove data sets has been achieved by establishing a cross-functional delivery team and using artificial intelligence and data science techniques, which allowed the UKHO’s Data Science, Data Engineering and Remote Sensing teams to create a method to identify mangrove from satellite imagery that ensures automated, scalable, and repeatable analysis.
The data set was created by using a neural network that learned to identify mangroves from satellite imagery. This network was initially trained using over 7000 coastal images distributed across the world, and its capability for detection will continue to improve as further images are collected and processed.
Importantly for UKHO, the new ADMIRALTY Global Mangrove Data Set is a transformational product and is a truly proactive data set. Under the ‘reactive’ model of building a mangrove dataset, changes to mangrove extent were manually digitised from satellite imagery; a process that was unscalable, unrepeatable, and time-consuming.
This model has now changed, and the ADMIRALTY Global Mangrove Data Set will be updated regularly and automatically. The new processes brought into place for this dataset will vastly improve the UKHO’s ability to identify mangrove forests regardless of geographical location.
How does it work?
Every six months, the UKHO searches for the latest available satellite images from the European Space Agency's Sentinel-2 mission. From these images, we select examples of coastal scenes that are clear and cloud-free, which will be processed to detect mangrove forests. Once images have been selected, they are analysed using the new mangrove detection model described above, which recognises the presence of mangroves in each individual pixel.
The ADMIRALTY Global Mangrove Data Set harnesses global, open data, and applies serverless applications and deep learning to create a comprehensive data set that is actively contributing to the quest for geo-generalisability in mangrove forests.
I’m interested in learning more. What should I do next?
The ADMIRALTY Global Mangrove Data Set is available in ESRI Shapefile and other industry standard formats – including raster and vector format - and is updated every 6 months. If you would like to learn more about ADMIRALTY Global Mangrove Data Set, please visit our dedicated web page: https://discover.admiralty.co.uk/access-data/mangroves
To download an extract of the UKHO Global Mangrove Data Set and try the data, please visit the ADMIRALTY Marine Data Portal: https://datahub.admiralty.co.uk/portal/apps/sites/#/marine-data-portal
Our work to detect mangroves is constantly evolving and improving as we collect new data and experiment with the latest cutting-edge technologies. User feedback is crucial for us to understand how the model is performing in different geographical regions and use this information to incorporate new training data into the next iteration of the model.
Insights and suggestions from our customers ultimately help our Data Scientists to target their research and apply updates in areas of interest to meet user requirements. We would like to hear from you. If you would like to share your valuable feedback and suggestions or if you have any questions regarding the ADMIRALTY Global Mangrove Data Set, please reach out to our Customer Services team at firstname.lastname@example.org.