Analysis: AI improves flood warnings but can’t erase disaster risk

LONDON – When floods ravaged parts of Europe in September, the scale of the destruction surprised people. These intense rains should not have happened because they were predicted by sophisticated forecasting systems with artificial intelligence.

But being warned does not mean being warned. Although the rains were accurately predicted, their effects in flooded areas were not – a fact that highlights the difficulties of managing increasingly frequent extreme weather conditions.

AI has supercharged weather forecasting, using a range of statistical tools to analyze years of historical data and predict patterns, all at a lower cost than traditional numerical weather forecasts.

AI technology can create more specific forecasts ahead of events like urban flooding or in complex terrain like mountainous areas.

For example, Google-funded GraphCast, a machine learning-based method trained directly on reanalysis data, was found to outperform traditional models. Reanalysis data draws on past forecasts rerun with modern forecast models to provide the most complete picture of past weather and climate conditions.

But there are still gaps in knowledge, in how information is used and in investments to strengthen data collection models, experts say.

“In some cases and for some variables, AI models can beat physics-based models, but in other cases vice versa,” said Andrew Charlton-Perez, professor of meteorology at the University of Reading in the United Kingdom.

One problem is that the effectiveness of an AI model depends on the information it contains. If there are few input data or if extreme events occur more frequently at different times of the year or in different regions, weather disasters become more difficult to predict.

“A good use of AI-based weather forecasting would be to complement and improve our forecasting toolkit, perhaps allowing us to produce broader sets of forecasts allowing accurate assessment and interpretation of probability extreme events,” added Charlton-Perez.

COMMUNICATION IS KEY

Since January, the European Center for Medium-Range Weather Forecasts (ECMWF), an independent organization that provides forecasts four times a day to European countries, has been using artificial intelligence/integrated forecast system (AIFS).

This data-driven forecasting model quickly makes multiple forecasts and provides long-term predictions of weather events such as cyclones and heatwaves.

ECMWF readings before the September floods were accurate, experts say.

Thomas Wostal, press officer at weather observatory GeoSphere Austria, told Context/Thomson Reuters Foundation that their numerical models – including the ECMWF forecast – predicted 300 to 400 millimeters of rain locally, which happened.

But even with accurate forecasts, scientists say communication is key, especially at a time when climate change is leading to more extreme weather events.

“I think what’s happened with (the recent floods) … is that it’s so rare – an event that happens every 150 to 200 years – that even though the weather models pick it up, there has a reasonable degree of uncertainty,” Shruti said. Nath, postdoctoral research assistant in weather and climate forecasting at the University of Oxford.

“You have to produce the warning in a way that is communicative, in the degree of severity it could have on people, then people could see that the cost of inaction versus the cost of action is actually much higher. So they would actually put (in) more resources,” she said.

EUROPE BEHIND THE CURVE?

Europe faces urgent climate risks that outstrip adaptation policies and actions, warns a report from the European Environment Agency.

Extreme heat, drought, wildfires and floods will worsen in Europe even under optimistic global warming scenarios and affect living conditions across the continent, says the EEA.

After the floods, EU Crisis Management Commissioner Janez Lenarčič said the disaster was not an anomaly.

“These once-in-a-lifetime extreme weather events now occur almost every year. The global reality of climate change has crept into the daily lives of Europeans,” he said.

Some tech entrepreneurs say Europe is not ready.

Jonas Torland, co-founder of Norway-based 7Analytics, which develops flood and landslide forecasting models, said governments and companies in the United States have risk managers more accustomed to assessing risks. environmental risks, while in Europe, authorities lacked preparation.

“We often see substantial spending with minimal data for informed decision-making,” Torland, whose models are used in the cities of Oslo, Bergen and Kristiansand, told the Thomson Reuters Foundation.

“While AI is a crucial part of these models, unfortunately governments are not investing in or purchasing these advanced AI solutions,” he said, adding that he believes governments “s ‘stick to their old data providers and consultants’.

Data processing is also a challenge, as these complex AI models must be updated hourly as forecasts change.

This requires both a lot of computing power and a lot of time, especially at smaller scales.

A 1 by 1 meter grid, which 7Analytics uses for its predictions, is 100 times more detailed than a 10 by 10 meter grid, but requires more than 100 times more time to process.

High computing power also means that huge amounts of energy and water are required, making AI models part of the problem as they contribute to the planet’s heat emissions, origin of the climate emergency.

Some major technology companies, such as Microsoft and Google, are exploring the use of nuclear power to power their massive data storage centers.

Other scientists point out that beyond perfecting their forecasting capabilities, authorities need to invest in physical solutions, such as developing areas where floodwater can be safely stored and early warning systems .

They must also minimize development in flood-prone areas, given the likelihood of more intense flooding linked to climate change, and meet their commitments to limit emissions.

“It’s not a question of data, technology or knowledge. It’s a question of implementation, of political will,” Friederike Otto, a senior lecturer at Imperial College London, said in a email response to questions.

“As long as the world burns fossil fuels, the root cause of climate change, extreme weather events will continue to intensify, killing people and destroying homes. To stop this trend, we must replace oil, gas and coal through renewable energies.”