1
“Flood peaks in Mid-Atlantic U.S. watersheds show a V-shaped response to urbanization: they decrease at low urban development (below 10% PDAW), then rise sharply as urbanization increases further due to complex climate-landscape interactions. A neural network model confirmed this nonlinear behavior, underscoring the need to consider more than just urban area when predicting flood risk.”…..Hua et.al.(2025).
Click here to know more.
2
“Artificial neural network (ANN) models using climate indices outperformed ARIMA models for long-term streamflow forecasting at six Victorian stations, with the best ANN achieving Pearson R = 0.90, RMSE = 0.04, and MAE = 2.50 for six-month advance prediction. ANN’s superior accuracy highlights its suitability for operational streamflow forecasts in this region.ANN models using lagged climate indices significantly outperformed ARIMA models for forecasting streamflow at Victoria’s stations, especially at Acheron where the ANN achieved Pearson R = 0.90, RMSE = 0.04, and MAE = 2.50. These results show ANN’s strong capability for accurate six-month-ahead streamflow forecasts.”……Oad and Imteaz(2025).
Click here to know more.
3
“This research demonstrates that multi-site LSTM models with trainable site embeddings can accurately predict groundwater levels globally, learning site-specific patterns directly from time series data without relying on inconsistent external descriptors. The study also finds that, while predictive performance is strong in data-rich regions, simply increasing dataset size does not guarantee better predictions in data-sparse areas, highlighting both the potential and challenges of scalable deep learning for groundwater forecasting.”….Nolte et.al.(2025)
Click here to know more.
4
The Knowledge Transfer Partnership funded by Innovate UK helped University of Exeter develop “Ground Truth,” a digital twin combining physical and numerical modeling with machine learning to predict water pipe deterioration and failures under various conditions. The tool, now used by multiple UK and European utilities, enhances pipe condition assessment and maintenance efficiency, earning the Exeter Knowledge Exchange Award 2024 for its impactful innovation.
Click here to know more.
5
The University of Exeter has joined the ENFORCE project for the period from 2024 to 2028. The ENFORCE project promotes sustainable environmental practices by integrating citizen science, AI, and geospatial intelligence to improve data quality, monitoring, and regulatory compliance across Europe. It establishes a pan-European collaboration hub, tests innovative tools at eight pilot sites, and supports policy development aligned with the EU Green Deal and UN SDGs.
Click here to know more.