New algorithm can predict extreme environmental events

A new method for predicting extreme events has been developed by researchers from the Massachusetts Institute of Technology (MIT). Extreme events refer to incidents like a previously stable species going extinct, a wave rising from calm waters, or an instability inside of a wind turbine.

These random events happen all over but scientists have lacked sufficient data and technology to be able to predict them with any real accuracy until now.

Engineers from MIT have developed a framework that combines real-world data and already existing equations to predict extreme events.

The results of these calculations show that extreme events can in fact be predicted, and that the model is general enough to apply to many systems where events can occur.

“This happens in random places around the world, and the question is being able to predict where these vortices or hotspots of extreme events will occur. If you can predict where these things occur, maybe you can develop some control techniques to suppress them,” said Themistoklis Sapsis, an associate professor of mechanical and ocean engineering at MIT.

Previously, when attempting to predict extreme events, scientists relied on solving a series of complex mathematical equations that could indicate certain initial conditions were a precursor to an extreme event.

But these equations, as Sapsis noted, were not foolproof, and often resulted in inaccurate and unrealistic predictions.

Scientists who wanted to use real-world data to try and find some commonality or indicator that predicted the events also struggled because extreme events are random and rare.

Instead, Sapsis and fellow MIT collaborators created an algorithm that uses both the applicable equations and any known real-world data to create a more stable predictor of extreme events.

The researchers then simulated a turbine fluid flow and used the algorithm to identify the precursors to an extreme event. The method identified developing precursors 75 to 99 percent of the time.

Having an accurate model that predicts extreme events will be incredibly useful for future studies and for finding methods to prevent such events from occurring.

—

By Kay Vandette, Earth.com Staff Writer