Artificial Intelligence (AI) has started leaving its mark in almost every industry. Many tech giants including Google are running the ongoing AI race and doing everything to integrate the tech into their products. Now in a groundbreaking development, Google’s AI research lab DeepMind has developed a new AI-based weather model called GenCast that provides 15-day forecasts with unmatched accuracy and speed.
Google’s new AI weather model “GenCast” largely beats the leading forecasting system
GenCast beats the world’s most accurate forecasting system developed by the European Centre for Medium-Range Weather Forecasts (ECMWF). ECMWF’s system currently provides weather forecasts for 35 countries.
In an official blog post, Google Deepmind notes that it trained GenCast on weather data from 1979 to 2018. Interestingly, the findings were beyond great. GenCast reportedly outperformed the ECMWF’s weather system with a precision of 97.2% when scientists tested it on 1,320 forecast metrics.
It further mentions that GenCast consistently outperformed the leading model when forecasting extreme heat/cold, and high wind speeds. However, Google’s new AI weather model is unlikely to replace the current forecasting system. Rather, it is likely to be used as an additional forecasting system to predict day-to-day weather including extreme events.
GenCast is the new one, but there are a few that already exist
GenCast isn’t the only AI weather model from the tech giant, there are a few others too. Some of the examples include Google DeepMind’s AI-based deterministic medium-range forecasts and Google Research’s NeuralGCM, SEEDS, and flood models.
These models are currently powering user experiences on Google Search and Google Maps with improved forecasting of precipitation, wildfires, flooding, and extreme heat.
In a statement, the ECMWF chief, Florence Rabier, says, “Any new method that can enhance and accelerate this progress is extremely welcome in the context of the extreme societal pressures of climate change.” She also praises the new weather model and says, “It is indeed a leap forward.“