Search the EPA Research Database

Project title

Artificial Intelligence-powered Forecast for Harmful Algal Blooms

Primary Funding Agency

Environmental Protection Agency

Co-Funding Organisation(s)

n/a

Lead Organisation

Technological University Dublin (TU Dublin)

Lead Applicant

Ahmed Nasr

Project Abstract

Eutrophication of water bodies in Europe is contributing to the increase of Harmful Algal Blooms (HABs) which poses serious risk to human health. To address this problem, the AIHABs project will develop an early warning system to forecast the occurrence, spread and fate of cyanotoxins caused by HABs in inland and coastal waters, using Artificial Intelligence (AI) and the latest innovations in mathematical modelling, nanosensors, and remote sensing. The novelty of this project lies in merging these tools with the joint purpose of providing an early warning system to decision-making authorities in terms of risk to the public. The model predictions will allow timely action to minimise the risks of consuming surface waters or using them as recreational resources when the waterbodies are prone to produce toxic cyanobacterial blooms. A team of researchers in the School of Civil & Structural Engineering, TU Dublin will lead the AIHABs project which includes researchers from other six European academic and research institutes. The six institutes are (i) Norwegian University of Science and Technology Computer Science (IDI); (ii) GFZ German Research Centre for Geosciences Remote Sensing and Geoinformatics; (iii) University of South Bohemia in České Budějovice; (iv) International Iberian Nanotechnology Laboratory (INL); (v) Universidad Autónoma de Madrid; and (vi) University of Santiago de Compostela. Besides leading the AIHABs research consortium, the TU Dublin research team will also undertake the role of developing the inland and the coastal water quality mathematical models of the forecasting system. The water quality models consist of two components (i) a suite of hydrological (inland) and coastal deterministic models to simulate the physical and the bio-chemical governing processes of the fate and transport of toxin indicators, particularly phycocyanin (PC) and chlorophyll-a (chl-a) in inland and coastal waters, and (ii) a statistical model to relate PC and chl-a concentrations to the toxin levels in the aquatic environment. The deterministic models will be built using the MIKE modelling package developed by the Danish Hydraulic Institute (DHI), in particular the MIKE11 model will be used for the inland water modelling whereas the MIKE3FM model used for the coastal modelling. On the other hand a newly developed computer code will be written for the statistical modelling component. Two types of data will be required to develop the models including (i) physical data describing the physical characteristics of the modelled water bodies; and (ii) meteorological, hydrological, and water quality data to calibrate and validate the models. The required data will be acquired from open public source databases and other databases owned by some partners in the project. A number of candidate sites with a history of HABs in the countries of the project partners will be evaluated using multi-criteria analysis in order to identify the most suitable inland and coastal water sites for use in the study. The main criteria for selecting the sites will be the availability of the required data for modelling and the strong evidence of historical HABs.

Grant Approved

�290,850.00

Research Hub

Healthy Environment

Research Theme

n/a

Start Date

01/09/2021

Initial Projected Completion Date

31/08/2024