Blog

How Fluvialycs Predicted Floodings during Storms Ciara and Dennis

In February 2020, the UK faced a formidable challenge as Storms Ciara and Dennis unleashed their fury, leaving a trail of destruction in their wake. With wind speeds reaching unprecedented levels and incessant rainfall causing widespread flooding, these extreme weather events posed a significant threat to communities and critical infrastructure.

Fluvialycs, armed put the to the ultimate test its AI predictor model, using real-world data from Storms Ciara and Dennis. To properly simulate Fluvialycs capabilities, we used AI models that encountered these storms for the first time and therefore were not trained using any data relatd to these extreme storm events, showcasing its adaptability and forecasting power.

We evaluated the performance using three distinct predictors—1-day, 3-day, and 5-day—which forecasted river levels into the future based on present-day data. The results demonstrated the model’s effectiveness in accurately predicting river levels, providing valuable insights for real-time decision-making during extreme weather events.

Predictors in Action

We evaluated the performance using three distinct predictors—1-day, 3-day, and 5-day—each offering a unique perspective on river level forecasting. The 5-day predictor, while exhibiting slightly lower accuracy, impressively captured significant patterns in river levels that unfolded over the next 5 days. The 3-day predictor demonstrated enhanced accuracy, mimicking changes in river levels over the next 3 days. The 1-day predictor, almost perfectly predicting immediate events, emerged as a powerful tool for real-time decision-making.

5 Days Prediction

While the 5-day predictor exhibits lower overall accuracy compared to our other models, it impressively captures significant patterns in river levels that will unfold over the next 5 days. Notably, the model accurately predicted the peak level observed on the 17th with remarkable precision, showcasing its ability to anticipate upcoming extreme events even at an extended forecast horizon.

3 Days Prediction

The 3-day predictor demonstrates improved accuracy compared to the 5-day predictor, and the model can confirm what was predicted previously. The model excels in mimicking the expected changes in river levels over the next 3 days, offering a valuable tool for short-term forecasting. This enhanced accuracy contributes to more effective decision-making and preparedness.

1 Day Prediction

At the 1-day forecasting horizon, the model achieves nearly perfect predictions. With high precision, it anticipates and mirrors what will happen in river levels the very next 24 hours. This level of accuracy provides infrastructure managers, emergency responders, and other stakeholders with timely and reliable information for immediate action and risk mitigation. With Fluvialycs you will know what will happen tomorrow.

Predict and Manage Future Threats

In the face of extreme weather events, having a tool like Fluvialycs at your disposal can make all the difference. Our predictor model, proven in the crucible of Storms Ciara and Dennis, offers unparalleled accuracy in forecasting river levels. Whether you’re an infrastructure manager, emergency responder, or stakeholder in risk mitigation, Fluvialycs provides real-time insights for proactive decision-making.

Curious to learn more about how Fluvialycs can revolutionize flood prediction and risk management? Contact us at [email protected] or request a demo to explore the power of predictive analytics in safeguarding against extreme weather events. Join us in the journey to a safer and more resilient future.

Hi, I’m fluvialycs