By Johnny Ramirez

An invisible infrastructure brings food grown in the countryside to the markets in cities where people can buy them. It goes unnoticed because it works so well, and it works so well because each step in that process happens at great efficiency. “We have to produce the food, then we have to move it around,” says Dr. Debra Peters, Senior Advisor for the Earth Observation in the Office of the Chief Scientist at the US Department of Agriculture (USDA), and Research Scientist at the Jornada Experimental Range. “We have to store it, we have to get it in the trucks, the trucks need gas. You just step back and there are all these processes that rely on how much food we have and what kind of food it is. You would be surprised by how much within our country is determined by the kinds and amounts of food that are produced. It is just amazing.”

For all of this to happen smoothly, the people and organizations involved have to know what food is growing, when it will be ready, and how much of it there is. “Every month the USDA releases a report for all commodities in the country. We do a projection on how much food we are going to have at the end of the season.” Everything from corn to peas, from cotton to cattle has to be estimated. An agency within USDA, the National Agricultural Statistics Services (NASS), has the responsibility of performing a nation-wide survey of food producers every single month. The answers provided by these surveys are integral for the food infrastructure and its smooth operation. This information helps ensure food shortages don’t empty our shelves, that farmers get a fair price for what they produce, and that railroads know where to place their cars and how many are needed to get corn and other foods to market.

The survey methodology used is a time consuming and labor intensive process. It has also become increasingly difficult to get busy farmers and growers to respond to these surveys. These current methods are challenging to implement, but they are also well-tested and standardized through time. However, advancements in technology are opening up new methods to acquire these data with less of a burden on farmers and growers, yet with similar accuracy as the previous methods.

For Dr. Peters the question is, “How would you actually create a continental scale monitoring system that relies less on obtaining information directly from people?” Modern technology such as satellite imagery, ground based sensors, and other data coming from a variety of sources, could be used to estimate food production. “But the other part is, how do we develop a system that pulls in all of these different data sources?” Creating something on such a massive and integrated scale is an incredibly complex problem.

Enter the JASONs

JASON is an independent research group that spends much of its time advising the US government. It began in 1960 as a group of atomic physicists to advise the US government on nuclear issues. They are completely independent which puts them in the unique position of performing cutting edge research and providing advice without having the results influenced by outside parties. Since then, they have broadened their scope to solve complex issues such as cyber security and creating defenses against biological weapons. It is this group that Dr. Peters and others from the USDA are seeking help from to solve the problem of modernizing data collection.

Satellite imagery can show where fields are being cultivated, but can we tell peas from strawberries? How about apricot from peach orchards? One important question the JASON group will help us answer is: how much will we be able to rely on imagery versus relying on people making on-the-ground observations or providing information on surveys?

The JASON group is expected to meet and work on the problem this summer. If all goes well, the USDA will have a framework to help build a new generation of data collection technologies that the agricultural sector of our economy depends on. “What we want is a framework that puts it all together”, says Peters, “brings in new sources of data and integrates all of the historic information that we have been collecting for many decades, and synthesizes all of the information in new ways to help the USDA create production and yield estimates across the country in a more efficient way.”