New AI system can forecast ozone levels up to two weeks in advance

Ozone has very different effects, depending on where it is found. In the upper atmosphere (stratosphere), ozone shields us from the sun's harmful ultraviolet (UV) radiation, but high concentrations of ozone near the earth's surface (ground-level or tropospheric) is toxic to lungs and hearts and can trigger a variety of health issues.

Researchers at the University of Houston's Air Quality Forecasting and Modeling Lab have developed a new artificial intelligence (AI) system that is claimed to forecast ozone levels with accuracy up to two weeks in advance, compared to current systems that can accurately predict only three days ahead.

This advancement could lead to improved ways to control high ozone problems and even contribute to solutions for climate change issues, the researchers said.

Commenting on this development, Yunsoo Choi, professor of atmospheric chemistry and AI deep learning at UH's College of Natural Sciences and Mathematics, said, "This was very challenging. Nobody had done this previously. I believe we are the first to try to forecast surface ozone levels two weeks in advance."

Ozone has very different effects, depending on where it is found. In the upper atmosphere (stratosphere), ozone shields us from the sun's harmful ultraviolet (UV) radiation, but high concentrations of ozone near the earth's surface (ground-level or tropospheric) is toxic to lungs and hearts and can trigger a variety of health issues.

"Exposure can lead to throat irritation, trouble breathing, asthma, even respiratory damage. Some people are especially susceptible, including the very young, the elderly and the chronically ill," explained doctoral student Alqamah Sayeed, a researcher in Choi's lab and the first author of the research paper.

Conventional forecasting uses a numerical model that is slow and has limited accuracy. The new AI model uses a unique loss function - index of agreement (IOA) - over conventional loss functions. IOA is a mathematical comparison of gaps between what is expected and how things actually turn out.

Together the numerical model and the IOA enabled the AI algorithm to accurately predict outcomes of real-life ozone conditions by recognizing what happened before in similar situations.

There are still many commercial steps before the world can benefit from the discovery, the researchers said. The research findings are published online in the scientific journal, Scientific Reports-Nature.