Seagrass meadows provide a variety of ecosystem services; the recent attention that seagrass systems are beginning to attract amongst the service includes a potential to store and sequestrate atmospheric CO2. Thus, we anticipate that the restoration and conservation of seagrass meadows in port and coastal areas can contribute to atmospheric CO2 reduction. Restoration and conservation efforts have been made to recover the lost seagrass areas and to sustainably receive the services that the seagrass systems provide. Accurate measurement of the coverage and distribution of seagrass beds is important to establish the restoration technology and the quantification of the ecosystem services.
　Here, we developed the measurement technology of the coverage and distribution of seagrass beds and applied it to a model site, using both GIS (Geographic Information System) and high-resolution multi-band satellite imagery, consolidated by field observation data.
　The estimation accuracy of the distribution was greatly improved when field observation data were incorporated into the image analyses. However, supervised classification is greatly affected by "training data". Thus, it is still necessary to develop a valid method for the choice of "training data" in order to improve the analysis results.