Can Google's AI Forecasts Improve Supply Chains for Farmers?

Google Research’s weather prediction tool now supports 38 million farmers in India.

Built on an AI model named NeuralGCM, the system delivers tailored weather forecasts directly to farmers, helping them prepare for monsoon rains.

With India’s farming sector heavily dependent on timely rainfall, accurate forecasts are critical for both agricultural production and supply chain coordination.

NeuralGCM – short for Neural General Circulation Model – combines traditional physics-based modelling with machine learning. The result is a tool that matches the performance of conventional systems but operates with less computational demand.

While standard forecasting relies on supercomputers, NeuralGCM works from a single laptop, making it easier to deploy in rural and remote areas.

The model’s rollout follows a collaboration between Google Research and the University of Chicago, in partnership with India’s Ministry of Agriculture and Farmers’ Welfare.

The work supports climate resilience by providing farmers with the data they need to make planting decisions, which directly affect India’s US$400bn agricultural economy and its supply chains.

Forecasting monsoons without supercomputers

Forecasting the monsoon has long been one of agriculture’s most persistent challenges.

Monsoon timing influences planting windows and crop cycles, particularly in tropical countries where rain-fed farming is widespread. Any deviation in rainfall can ripple through supply chains, affecting everything from seed orders to transport logistics.

Traditional models rely on detailed physics simulations. These are expensive and resource-heavy, limiting their accessibility.

NeuralGCM, by contrast, uses decades of historical weather data to learn patterns and make predictions. This enables faster, cheaper forecasting while maintaining accuracy.

Olivia Graham, Product Manager at Google Research, and Stephan Hoyer, Engineer at Google Research, say in a Google blogpost: “For years, weather and climate models have been costly and complex, often requiring a supercomputer to run.

"Our teams at Google Research wanted to see if we could build these models more efficiently and more accurately, leading to the creation of NeuralGCM.”

The model’s open-source design also removes licensing barriers, allowing institutions like the University of Chicago to adapt and deploy it as needed. That flexibility has been key to integrating it with existing systems like the European Centre for Medium-Range Weather Forecasts’ AI/Integrated Forecasting System.

In trials, this combined system successfully predicted monsoon onset as far as one month in advance. One test captured a dry spell during the monsoon’s progression – a disruption that typically delays planting, disrupts irrigation plans and complicates distribution networks.

Helping farmers make timely decisions

The University of Chicago’s research focuses on how AI weather tools influence farmer decisions. Their findings show that giving farmers forecasts about 30 days in advance allows for better alignment between weather and planting strategies.

In one pilot, early weather predictions nearly doubled the annual income of participating farmers, underlining the connection between climate data and economic resilience.

Visual comparisons during testing show how the model performs. One map displays the average expected rainfall from 120 years of data.

Another shows the actual rainfall recorded by the India Meteorological Department. The third displays NeuralGCM’s forecast, issued 15 days ahead of time – accurately reflecting the weather that followed.

These insights feed directly into decision-making. Farmers receive forecast updates via SMS, which include practical advice such as when to plant based on predicted rainfall. This not only helps reduce crop failure but also allows for better coordination with input suppliers, irrigation services and transport providers.

The messaging service is facilitated by India’s Ministry of Agriculture and Farmers’ Welfare, which manages national farming policies and support schemes. As half of India’s workforce relies on agriculture, distributing accurate forecasts at scale is essential for the country’s food security and economic stability.

AI opens doors to scalable climate tools

With 38 million farmers receiving forecasts during the summer growing season, the collaboration has already reached national scale. The forecasts are especially useful during years of weather anomalies, such as delayed monsoons, where traditional schedules no longer hold.

The integration of AI into climate services offers long-term benefits for agricultural supply chains.

By improving forecast accuracy, it reduces the uncertainty that often disrupts planting, harvesting and shipping. More consistent planning helps stabilise prices, reduce waste and support supplier networks across the sector.

As Olivia and Stephan explain, NeuralGCM’s success points to broader opportunities: “A powerful example of how foundational AI technology, born from research, can serve real-world use cases, ultimately helping communities around the world build climate resilience.”

By combining historical data with predictive modelling, NeuralGCM helps bridge the gap between science and practice.