Seagrass detection in the Mediterranean: A supervised learning approach | Central Portal
  • Home
  • Seagrass detection in the Mediterranean: A supervised learning approach

Seagrass detection in the Mediterranean: A supervised learning approach

Democritus University of Thrace (DUTH) uses hydrodynamic model results to provide accurate forecasts, to implement and apply fish and mussel growth models and to employ optimization tools and services to managers and operators in the aquaculture sector. In parallel, DUTH uses wave data in relation to historic shoreline changes produced by the analysis of satellite images, to asses the wave climate, to describe extreme conditions and define wave erosion hotspots.  

In a recent research paper (Effrosynidis et al., 2018) DUTH aggregated CMEMS and EMODnet data to investigate the influence of environmental conditions on the presence-absence and the distribution of seagrass species over the Mediterranean Sea. To achieve this scope, DUTH leveraged on advanced machine learning techniques and combined a great number of morphodynamic, environmental and human impact variables. Seabed substrate, bathymetry and habitats were extracted from EMODnet Geology portal, distances from cities, ports and river mouths were derived from Desktop (2011) and temporally-changing parameters as water temperature, salinity, nutrients, chlorophyll-a, turbidity and wave height were deduced from the CMEMS database. DUTH performed comparative tests between the ML results when using the random forest and the maximum entropy algorithms.

The results suggest that the main environmental parameters affecting the distribution of seagrass at family and genera level are the winter (mainly December) chlorophyll-α and salinity levels, the autumn phosphate concentrations and bathymetry, expressing changes in temperature, pressure and light availability.

DUTH plans to continue to use EMODnet data in the future to inter-relate the abundance and diversity of fish species in relation to environmental variables, producing a significant impact to the society. 

About DUTH

Democritus University of Thrace was established in 1973 and is currently operating eight Schools and nineteen Departments in four cities of Thrace. It has 850 employees, 21,000 students, 2,300 postgraduate students and about 1,750 PhD students. Relying on its quality of teaching and research level, DUTH has secured a place among the best Greek Universities. DUTH is currently coordinating the H2020-funded project entitled ODYSSEA. In parallel, DUTH participates and initiates several actions towards the operational oceanographic monitoring and modeling in Thracian Sea. Hydrodynamic, wave and water quality models are currently in operation, simulating the conditions prevailing in the broader area, especially under the influence of Black Sea water outflow. DUTH has participated/ is currently participating in several EU funded programs, such as FP7 actions: (27; of which four are MSCA), H2020 actions: (12; two of which are MSCA and another one in the Grant Preparation Phase), JUST Action Grants: (2), Life actions: (1), Erasmus+ actions: (2), Jean Monnet Chair (1) etc. Furthermore, the University is participating in several transnational and national projects. The projects cover several scientific areas such as engineering, IT, environmental and agricultural sciences, medicine, molecular biology, law, economics, sociology, humanities, education and physical education.

Case Type:

Second Edition of the EMODnet Open Sea Lab

Are you a Coder? Communicator? Data enthusiast? Or Entrepreneur? Join us at the second EMODnet Open Sea Lab hackathon and develop innovative solutions with marine open data to address blue society challenges!
04 Sep 2019 to 06 Sep 2019
Zebrastraat, Ghent, Belgium

LATEST NEWS