How Google Earth Engine can inform more sustainable agriculture practices

Food and water security are among the most pressing problems facing the world today. According to the United Nation’s Food and Agriculture Organization (FAO), the world will see a 60% increase in demand for food by 2050. And according to the World Health Organization, since 2019, an additional 122 million people are experiencing hunger, due to multiple ongoing crises. On the 8th day of COP 28, the focus was on how to build resilience into food, agriculture and water systems.

Addressing the food and water security challenge under changing climatic conditions is critical, but it's also a complex challenge to solve given the scale and magnitude of the problem. The primary issue has been access to data and therefore insights on understanding both the impact of climate on agriculture, and vice versa.

At Google, our mission is to organize the world’s information and make it universally accessible and useful, so that decision makers can ask questions of increasingly large and complex datasets. At Google Cloud, we are partnering with organizations across the agricultural ecosystem to help solve the data and insight accessibility problem. Central to that endeavor is our planetary-scale remote sensing platform, Google Earth Engine.This pioneering analytics platform has 50 years of analysis-ready global satellite data, and the massive computational power to run machine learning (ML) models over the entire terrestrial surface of the globe, bridging the data and insight gap.

Earth Engine is already helping some of the largest organizations around the world, across many industries, including agriculture. For example, remote sensing startup Regrow, based in Australia, is focused on helping their customers transition to regenerative agricultural practices. Using Earth Engine and advanced machine learning models, they’re able to monitor 1.2B acres of land globally. By tracking, modeling, and monitoring agricultural practices, Regrow helps customers promote carbon sequestration in soil, reduce greenhouse gas emissions through low-till and no-till agriculture, and increase biodiversity and water usage. Earth Engine is also central to Unilever’s mission of sustainable sourcing, as part of their target of no deforestation in their palm oil supply chains. Leveraging Earth Engine and our partner solution TraceMark, which is built on Earth Engine, Unilever is able to trace its palm oil purchases back to the plantations, ensuring that they don’t source palm oil from deforested land.

In addition to working with enterprises, we’ve also been partnering with governments to drive adoption and acceptance of the technology. In India, Google Research and Google Partner Innovation have been working on AnthroKrishi, a project that allows data-driven climate interventions at scale by identifying precise agricultural landscapes and monitoring them via satellite imagery. In India alone, where farm sizes are much smaller on average than in North America and Europe, there’s a potential 750 million field parcels that would need to be monitored. Farm field-level data has been unavailable to developers/ecosystem until now. AnthroKrishi's vision is to provide a fundamental base layer with agricultural data which businesses, technical teams can build solutions on top of, and combine them with their own datasets so they can generate unique and useful insights. Different stakeholders can build bespoke agricultural solutions like access to finance for farmers to release loan disbursements, to tailor insurance needs to local variations in climate and soil quality, effective and efficient orchestration of the food supply chain, or to provide the right level of subsidies on inputs and to enable farm-level precision advisory and better access to markets.

We will be launching the AnthroKrishi API early next year for India, and late next year for other parts of the globe. If you are interested in early access, or would like to partner with Google Research to bring these sorts of insights to your country, please reach out via our website. If you’re interested in learning more about how to use Google Cloud’s advanced machine learning capabilities with Earth Engine to make a real climate impact, please reach out via your account team or contact us here.

^{Special thanks to Manish Gupta, Research Director for Google Research India and his team, for his invaluable contributions to this post.}