Satellites and grain
By: Alejandro Morales
In July of 1972 during the midst of the cold war Russia was experiencing severe drought and was in need of a foreign grain purchase. Taking the US by surprise, Russia bought 15 million tons of wheat, corn, soybeans and barley at subsidized prices causing instability in our grain stocks and a domestic food crisis. The soviets had anticipated that they and the US would be in shortage of grain through the use of satellite technology that could survey farmland and estimate crop yield to then predict prices. This method of technology has since been developed over decades of technological improvements to be one of the key factors that agricultural commodity firms, from large to small, have been using to anticipate the market.
Decades of technological advancement have turned satellite monitoring from a Cold War intelligence tool into a commercial necessity. Agricultural commodity firms, from large to small, are able to better forecast harvests, anticipate price swings, and navigate volatile markets. Today, satellites are just one of several data-driven tools shaping how firms predict supply, demand, and global trade flows.
Traditionally, specialists use complex models to predict crop yield and prices that are not always completely accurate and require input data that is often difficult to obtain prior to harvest. One example of this are process-based crop models that require data on a variety of agricultural factors such as climate, soil characteristics, and agricultural practices. These models however lack a consistent accuracy to annual yields despite their insight on various aspects of harvest. Another method are on-farm surveys that can gather information from those working directly on the field, offering accurate insight for estimates. However, gathering these surveys is tedious and resource intensive.
Satellite surveillance not only offers the most accurate crop yield, but is less resource intensive and can analyze large amounts of complex data without much physical input. The ability to accurately predict crop yield is essential to predict global crop prices. For example, since the US provides nearly 35% of the world’s corn production there is a direct correlation to the success of US corn yield on the global price of corn. This is also significant for substitute crops, where being able to predict the yields across various crops can allow for a prediction on the supply and demand of crops that can replace one another such as rice, wheat, soy, barley, etc.
With emerging AI technologies, the amount of information that can be computed using satellite imagery has not only expanded, but now no longer is subject to human error. AI has sharpened the human aspects of data processing and input previously required for data interpretation, optimizing crop yield predictions. This can be seen with the changing of constantly shifting factors in real time, such as soil moisture and current forecasts. Lastly, despite focusing on the commodity trading aspect of crop satellite imagery, this also has implications for each human on the planet. This technology allows farmers to more accurately analyze their crops, being able to invest in areas that will carry the most yield and which areas are not optimal. As AI continues to evolve, integrating it with satellite monitoring may redefine our understanding of agriculture, reshaping both global markets and how humanity feeds itself.
