AI will improve weather forecasting: but what about meteorologists?

Artificial intelligence might soon produce quicker and more accurate weather forecasts, meteorologists in Germany have said in a new research document. But aren’t they digging their own graves?

Researchers from Germany’s meteorological service Deutscher Wetterdienst (DWD) say they have relied exclusively on AI to successfully feed data from weather observation into forecasting models and analysis processes.

According to them, it’s clear that better speeds, improved analyses, and forecasts will benefit “anyone looking at weather forecasts” – flood protection specialists or people in the energy sector but also regular folk keen to know as precisely as possible what the weather will be like during their holidays, for example.

Of course, the process of data assimilation (DA) – the integration of observation data into computerized weather models – has been around the industry for decades. But now, AI can dramatically cut costs.

“Unlike previous hybrid approaches, our method integrates the DA process directly into a neural network, utilizing the variational DA framework,” said the researchers in the study, which is not yet peer-reviewed.

“This innovative AI-based system, termed AI-Var, employs a neural network trained to minimize the variational cost function, enabling it to perform DA without relying on pre-existing analysis datasets.”

"This technology is a crucial building block for the current revolution in weather forecasting through AI," Jan Keller, Head of Data Assimilation at DWD and co-author of the research, told the DPA news agency.

However, as in most other industries, the dawn of AI could also signal bad news employment-wise for human meteorologists – if the new models can indeed replace traditional methods, reduce computing costs, and speed up processes.

For instance, in late 2023, Google introduced GraphCast, a new AI model allegedly able to make 10-day weather predictions at “unprecedented accuracy” in under one minute.

In a blog post, Google said that GraphCast can also offer earlier warnings of extreme weather events, and that could save lives: “ It can predict the tracks of cyclones with great accuracy further into the future, identifies atmospheric rivers associated with flood risk, and predicts the onset of extreme temperatures.”

Finally, in early June, Microsoft also introduced its latest innovation in AI modeling with Aurora – a breakthrough foundation model that can more accurately predict weather patterns, ultimately helping decision-makers worldwide prepare for extreme weather events.