14 Sep 2018 Story Oceans & seas

Cloud computing to speed up stocktaking of Northwest Pacific blue carbon sinks

Cloud computing technology can speed up the assessment of Northwest Pacific seagrass beds which nurture biodiversity, purify seawater and mitigate climate change but are threatened by human activities and natural disasters, says a study by the UN Environment Regional Seas Programme in the region.

Ten species of seagrass reported in the Northwest Pacific seas that border Japan, People’s Republic of China, Republic of Korea and the Russian Federation are included in the International Union for Conservation of Nature red list of threatened species, according to the Feasibility Study Towards Assessment of Seagrass Distribution in the NOWPAP Region. The study was published by the Toyama-based Special Monitoring and Coastal Environmental Assessment Regional Activity Centre, which was set up under the UN Environment Action Plan for the Protection, Management and Development of the Marine and Coastal Environment of the Northwest Pacific Region.


Seagrass habitats in the region are threatened by land reclamation, port construction, destructive digging, aquaculture, oil exploitation and typhoons, says the study.

Seagrasses absorb huge amounts of carbon and are the focus of the International Blue Carbon Initiative for climate change mitigation, the International Partnership for Blue Carbon and the UN Environment-led Blue Forests Project co-funded by the Global Environment Facility. Mapping of seagrass beds in the Northwest Pacific supports the commitment by all nations at the June 2017 United Nations Ocean Conference to compile a global blue carbon database.

Seagrass beds also have crucial ecological functions such as providing habitats for marine life, spawning, nursery and breeding grounds for aquatic biota, and purifying seawater by absorbing nutrients (nitrogen and phosphorous).

Mapping seagrass deep under water using satellite remote sensing is more difficult than satellite mapping of mangroves and tidal marshes that are easily visible on the surface. Water turbidity, sun glint and epiphytes covering the grass blades can dilute the spectral reflectance signal of seagrasses, reducing the ability of instruments to see through water.


Case studies conducted by the Special Monitoring and Coastal Environmental Assessment Regional Activity Centre in selected Northwest Pacific sea areas during 2015-2016 using satellite imagery found that it was too expensive and time-consuming to use high-resolution data from commercial satellites and conventional methods of analysis. The process involved purchasing images from private satellite operators, scientists spending several months on each case study to analyze the satellite images using field data for classifying sea floor substrates, and removing sun glint and the effect of water column from the satellite images.

“Taking into account costs and time spent in the case studies in the selected sea areas, it was considered unrealistic to apply the same method to map distribution of seagrass in the entire Northwest Pacific coastal zones,” says the study.

Instead, it recommends using cloud computing technologies to analyze freely available multispectral satellite images with a standardized analysis procedure that has been developed by the Centre. The spatial resolution of freely available satellite sensors has improved from 30m to up to 10m with the installation of the Multispectral Imager (MSI) on board the Sentinel-2 satellite of the European Space Agency. Sentinnel-2 MSI images are available free of charge in the Google Earth Engine data catalogue and Amazon Web Services, and anyone can use their on-demand computing resources to perform analysis and create new products without incurring the cost and time required to download and use Landsat data.

New free platforms to analyze and visualize satellite images are also available from private companies such as Google Earth Engine that combines a multi-petabyte catalog of satellite imagery and geospatial datasets with planetary-scale analysis capabilities. The entire process of analyzing satellite images takes only a few days with the Google Earth Engine for which a single computer would have taken 15 years. 


The study also advises to develop an online tool to analyze satellite images, maintaining seagrass databases, and building capacities in scientific institutions and civil society to map seagrass distribution.

“Mapping distribution of seagrass is a tremendous effort,” says the study. The authors recommend collaboration among voluntary citizen groups and non-governmental organizations working on conservation and restoration of seagrass at international, regional, national and local levels. These include the world’s largest seagrass monitoring programme ‘Seagrass-Watch’ which runs ‘Project Seagrass’ devoted to the conservation of seagrass ecosystems through education, advocacy, research and action. ‘Seagrass Spotter’, a smartphone tool developed by Project Seagrass is available to the public to collect seagrass field data worldwide.

A prototype for the regional seagrass database, developed by the Special Monitoring and Coastal Environmental Assessment Regional Activity Centre - Learn more