Award for AI Tool Predicting Posidonia on Dénia Coast

A UPV student has developed an application with artificial neural networks to protect Posidonia oceanica meadows with over 90% accuracy.

Image of a hand interacting with a holographic projection of a marine plant, with a blurred ocean background, representing artificial intelligence and sustainable development.
IA

Image of a hand interacting with a holographic projection of a marine plant, with a blurred ocean background, representing artificial intelligence and sustainable development.

An innovative tool using artificial intelligence and neural networks to predict the presence of Posidonia oceanica meadows, with over 90% accuracy, has been awarded by the Enia-UPV Chair.

The proposal, developed by Mishel Asparuhova Danova, a student at the Universitat Politècnica de València (UPV), won the award for the best Master's Final Project from the Enia-UPV Chair in AI and sustainable development, with Nunsys Group as the promoting company. This innovative and sustainable application can be used in coastal areas of the Mediterranean with limited ecological information.
Furthermore, the tool can evaluate solutions based on low-level dikes that protect the beach from erosion and also the Posidonia oceanica meadow itself by reducing wave energy. The main novelty is a formulation that allows estimating the upper limit of Posidonia oceanica distribution, i.e., the area closest to the coast from which it can be found. This advance facilitates decision-making in coastal protection actions, by allowing solutions to be adapted to the real conditions of the ecosystem.
Posidonia oceanica meadows are marine ecosystems formed by higher plants in the Mediterranean. They function as true lungs for this sea because they produce large quantities of oxygen and capture enormous tons of carbon dioxide, essential, therefore, in climate change. They also protect the coastline and are a refuge for biodiversity, acting as a natural filter by improving water purity and transparency, and a biomarker, as their presence indicates a healthy ecosystem of high environmental quality. Despite their importance, they are vulnerable to pollution, boat anchoring, and invasive species.
Therefore, this AI tool addresses an environmental problem of great relevance on Mediterranean coasts, such as the progressive degradation of Posidonia oceanica marine meadows and coastal erosion due to changes in maritime climate and human intervention on the coast. The model defines a procedure based on three axes: hydrodynamic and morphodynamic modeling; the application of artificial neural networks to estimate the probability of posidonia presence based on local wave energy and depth conditions; and a beach regeneration proposal, with ecological criteria for Blay Beach in Les Marines, Dénia. This includes the conservation of meadows, selective fillings, and dike design to promote natural circulation and wave attenuation.

"This work is directly framed within environmental and technological sustainability, and provides prediction, analysis, and design tools that allow integrating the ecological variable into coastal engineering objectively. In this way, it uses artificial intelligence to solve a sustainability problem, which is the objective of this Chair."

Vicent Botti · Director of the Enia-UPV Chair and Director of Vrain at UPV

"These types of habitats, considered priorities by the Habitats Directive (92/43/EEC), are essential for the balance of the coastal ecosystem. Therefore, their study and conservation are essential within the sustainable development and climate change adaptation strategies promoted by the European Union and the 2030 Agenda."

Mishel Asparuhova Danova · UPV Student