Researchers have been working on an AI model that improves storm intensity predictions, and the results are promising. The model, which was developed by Google DeepMind, is able to make predictions about storm intensity using lower-resolution data, but the exact reason behind its success is still unknown.
According to Alet, the model is like a “black box” that provides physicists with a signal that something is happening that was not previously understood. The model produces a range of potential scenarios for a developing storm, which helps to capture any potential “butterfly effect,” where a small deviation from a trend could lead to much bigger changes down the line.
The AI model doesn’t just spit out one prediction.
It generates a range of possible scenarios for a developing storm. This allows forecasters to use the outputs, alongside those of other models, to inform their predictions about how a storm system will likely unfold. Last year, the AI model created 50 scenarios per storm; now, it generates 1,000.
Related: How early search engines built the web
Musgrave notes that the model’s ability to generate so many scenarios is something that can’t be done with existing numerical models, given current computing power.
Brennan emphasizes that the DeepMind model is just one tool in forecasters’ toolbox, and that it’s not a guarantee that it will be the best model for the next season or the next storm. The human element, he adds, is still critical in forecasting storms. A hurricane is not just a track or an intensity forecast, but also requires experts to translate that into what the impacts are going to be – and it’s the impacts that kill people.
They are not a replacement for human forecasters, but rather a tool to aid them in their predictions. By providing a range of possible scenarios, the model can help forecasters make more informed decisions about storm intensity and potential impacts.
As the AI model continues to improve, it’s likely that we’ll see more accurate storm intensity predictions. This could lead to better preparedness and response to storms, ultimately saving lives and reducing damage.
Related: Acer Chromebook defies low expectations
Google DeepMind has also announced that it is open-sourcing the WeatherNext models used during hurricane season, so that researchers can use and improve on them. Alet is hopeful that opening the models up to the research community could help uncover fresh insights into how cyclones work.
By making the WeatherNext models available to the public, Google DeepMind is allowing researchers to build upon their work and potentially make new discoveries. This could lead to a better understanding of how storms develop and more accurate predictions in the future.
This story originally appeared on wired.com.