TERRA Research on AI-Powered River Contamination Prediction Presented at AIAI 2026
16 July 2026 | Chania, Crete, Greece
Research carried out in the context of the TERRA Horizon Europe project was presented at the 22nd IFIP WG 12.5 International Conference on Artificial Intelligence Applications and Innovations (AIAI 2026), held from 16–19 July 2026 in Chania, Crete, Greece.
The University of Thessaly (UTH) participated in the international conference with the paper “Satellite-Informed Machine Learning for River Contamination Prediction: A Time-Aware LSTM Approach,” presenting research directly linked to TERRA Demonstrator 1 and the project’s work on AI-enabled environmental monitoring and forecasting.
The presented research introduces a satellite-informed machine-learning approach for predicting river contamination using a time-aware Long Short-Term Memory (LSTM) model.
The methodology explores how Earth Observation data and temporally structured environmental information can be combined with deep-learning techniques to improve the monitoring and forecasting of water quality. By accounting for the temporal dynamics of environmental conditions, the approach contributes to the development of more effective predictive tools that can support environmental intelligence and informed decision-making.
The paper was authored by Dionysis Giannaropoulos, Minas Petridis, Panagiotis Oikonomou and Kostas Kolomvatsos, with Dionysis Giannaropoulos presenting the work during the conference’s session on Environmental, Energy & Sustainability Intelligence.
The presentation reached an international audience of researchers, academics and professionals working across artificial intelligence, machine learning, environmental monitoring and sustainability. It also provided an opportunity for scientific exchange on the methodology and its potential application in real-world water-quality monitoring scenarios.
Supporting TERRA Demonstrator 1
The research is closely connected to TERRA Demonstrator 1, which focuses on the monitoring and forecasting of water contamination through the integration of Earth Observation, environmental data and AI-based predictive methods.
Presenting these results at AIAI 2026 provided an important opportunity to communicate TERRA’s scientific work beyond the project consortium and demonstrate how the project’s AI-driven approaches can contribute to practical environmental challenges.
The participation therefore strengthens TERRA’s scientific dissemination and visibility, while promoting knowledge exchange with the wider AI and environmental research communities.
Extending TERRA’s scientific impact
Beyond the conference presentation, the research has been published in the official Springer proceedings of AIAI 2026, further extending its reach and making the results accessible to the international research community.
The paper appears in Artificial Intelligence Applications and Innovations: 22nd IFIP WG 12.5 International Conference, AIAI 2026, Proceedings, Part IV, within the section dedicated to Environmental, Agriculture, Energy & Sustainability AI.
Through activities such as AIAI 2026, TERRA continues to bridge Earth Observation, Artificial Intelligence and environmental science, transforming advanced research into approaches that can support more effective climate and environmental intelligence services.
Learn more:
AIAI 2026 International Conference:
https://ifipaiai.org/2026/
Published paper – “Satellite-Informed Machine Learning for River Contamination Prediction: A Time-Aware LSTM Approach”:
https://doi.org/10.1007/978-3-032-30809-2_8
