Google AI helps Indian farmers predict the beginning of the monsoon season

This summer, 38 million farmers in India receive a powerful forecast for the first season Monsoon, to help them make informed decisions about when they will invest their crops. This predicts partially empowered by Neultgcm, a Google study model, including traditional physics models are based on the accuracy of developed and efficient imitation system.

AI model to predict weather and weather

Age, weather and weather models called and complicated, often requires supercophomputer running. Our groups in Google Christian wants to see if we can build these models successfully and more accurately, leading to the construction of the Neullgcm. Unlike traditional models depending on the hard codes, this model driven by AI is trained for decades historical information and read past events, while using physics. Clearly, designed to be consistent and efficient – can run in one laptop, make high-forecasts easily accessible in scientific society.

Collaboration with the University of Chicago

When we open a wind, we hope society will use this new tool to enable their new programs. University of Chicago's Human Center The weather forecast did that. They realize that one of the most affected choices, but the most common, decisions of farmers in India is where they put seeds in the ground. Hundreds of millions of young farmers in all abortions rely on the details of the rainy season, known as Monsion, will come each year. However, accurate prediction of where the monsoon will begin, especially the long-term and location session, has remained an old adventurer of the century.

With a solid test of the weather models of AI weather, the University of Chicago Team found that Neultgcm, when it is associated with other developed models such as European Intelligence Confing / History Data, was the appropriate tool for work. It accurately predicts the start of Indian Kosoon until monthly, even a dried spelling of the monsoon continuity.