This year, World Food Day fell on October 16. The United Nations’ Food and Agriculture Organization (FAO) released a new report about world hunger, based on data from 2016. In that report, the FAO says it is replacing its previous measures of food waste and food loss with two new indicators, the Food Loss Index, and the Food Waste Index. Commentary: The challenges ahead for food supply chains ran on FreightWaves on July 10, 2019, and delved into data from the FAO.
In this new report, the FAO stated, “If we want to reduce food loss and waste, then we need to know where it occurs and where interventions will be the most impactful.”
Coincidentally, a team of researchers from the University of Illinois at Urbana-Champaign and Kansas State University has been working to understand how to solve that problem in the United States.
In Food flows between counties in the United States, published on July 26, 2019, the authors “estimate food flows between all county pairs within the United States.” Performing such a study for the United States is important because the United States is a major world producer, consumer and exporter of agricultural produce. The authors differentiate between food flows – “the movement of food through complex supply chains within a country” and food trade – the international buying and selling of food commodities between countries.
The data the researchers rely on is:
- U.S. Census Bureau and the Bureau of Transportation Statistics’ Commodity Flow Survey (CFS), produced every five years during years ending in ‘2’ and ‘7.’
- The Freight Analysis Framework (FAF) from Oak Ridge National Laboratory.
An undertaking such as this quickly becomes very complex since the number of directed paths or links they must assess is given by (n)(n – 1), where n is the 132 FAF zones – with 17,292 potential links, and the 3,142 counties and county equivalents of the CFS – with 9,869,022 potential links.
To conduct the study, the researchers developed a computational algorithm that incorporates machine learning, linear programming, network constraints and mass balance. Mass balance is a conceptual requirement they developed to ensure that food flows from counties within an FAF zone – CFS level data, sum to food flows from the FAF zone.
The images below are maps of food flow networks within the United States. Maps depict total food flows (tons) for the (A) FAF and (B) county scale. Links are shown for all FAF data and for the largest 5% of county links.
The researchers found that:
- California and the Great Lakes region are major food outflow locations.
- Several California counties are important food inflow and outflow locations.
- Self-loops are prevalent, especially in Los Angeles County.
- The Los Angeles FAF zone is second only to the New Orleans FAF zone by size.
Why research of this nature matters
Research of this nature can help federal and state government agencies assess the risk that the United States’ food supply faces. Understanding where food is produced, processed and consumed can help in the design and maintenance of processing, storage, transportation and distribution networks required to reduce food loss.
The researchers suggest that future work could build on this research by:
- Improving the realism of the algorithms they developed to create these maps.
- Using the actual roadway network to estimate distances between counties instead of using the shortest linear path between counties.
- More fully utilizing the FAF data on transportation modes between counties in order to more accurately reveal areas of weakness and strength within the United States’ food supply chain.
