The United States is the world’s largest producer of beef. In 2015, the latest year data is available, the beef industry was valued at $105 billion. Protecting millions of cattle from potential disease outbreaks is thus a crucial part of our nation’s economic security, as well as a public health priority. Two new web-based tools funded by the Department of Homeland Security (DHS) Science and Technology Directorate (S&T) are making it easier for public officials and livestock farmers to predict cattle shipments and prepare for potential disease outbreaks: The U.S. Animal Movement Model (USAMM)-Shiny App and the CADENCE What-If Tool.
The USAMM-Shiny App, the first national-scale model of livestock in the US, is an interactive web application that creates heat maps to visualize the movement of cattle across the U.S. Users can select a county, see how many shipments of cattle are predicted to be transported to and from the location, and note the uncertainty level of the prediction. The visualization of movement data can help inform risk assessments, scenario development and traceback efforts.
While it currently only shows cattle movement, the project will expand in the future to include shipments of other animals. The app was developed by Colorado State University biologist Colleen Webb, in partnership with Linkoping University in Sweden and the University of Warwick in the United Kingdom, as part of S&T’s Agriculture Defense Program in the Homeland Security Advanced Research Projects Agency Chemical and Biological Defense Division.
“The detailed pattern of cattle shipments across the U.S. is not very easy to observe, but is important for understanding the potential for disease spread,” said Dr. Colleen Webb, the USAMM-Shiny App developer and a Biologist at Colorado State University. “We see different facets of U.S. cattle shipment patterns in different data sets, but these different data streams need to be combined in order to get the whole picture. Modeling allows us to combine knowledge from different data streams and also characterize the level of agreement across the data streams when estimating the uncertainty associated with the model predictions. The USAMM-Shiny App allows users to visualize more detailed cattle shipment predictions than have previously been available.”
Figure 1: USAMM Heat Map Showing Predicted Cattle Shipments Transported to Ford County, Kansas
To generate the heat maps, the USAMM-Shiny App uses a 10 percent sample of interstate movements from Interstate Certificates of Veterinary Inspection, and includes additional data streams like National Agriculture Statistics Survey results. The information is processed through a Bayesian hierarchical model, which includes statistical and probability methods to account for uncertainty, and is then scaled up to make predictions about movement.
The CADENCE What-If Tool, also part of the S&T Agriculture Defense Program, produces livestock disease outbreak simulations across eight scenarios tied to American states. Users select inputs from several categories, including disease type, spread and detection. The model then quickly produces a summary of predicted disease duration, animal and farm depopulation and number of recommended farm vaccinations. This information is essential to public health officials and livestock farmers, who can use it to better prepare for outbreaks of foreign and endemic animal diseases. The CADENCE What-If Tool helps users evaluate control options across a wide range of scenarios, which means response efforts can be coordinated more effectively.
“Simulation modeling is often used to plan for disease outbreaks,” said Dr. Shrideep Pallickara, the CADENCE What-If tool developer and Computer Scientist at Colorado State University. “The expressiveness of simulations generally comes at a price – timeliness. When planners create scenarios, they have to find computing resources, run the simulations, and then await results. Depending on the complexity of the scenario being modeled,” he explained, “that wait can be hours or even days. The CADENCE What-If Tool retains the expressiveness of the models while producing results in less than a second.”
Figure 2: CADENCE What-If Tool Interface
How does the CADENCE What-If Tool process so much information in a matter of seconds? Researchers first ran the North American Animal Disease Spread Model through numerous scenarios and banked the outcomes. They then built the tool to extrapolate on these known scenarios, which increased its processing speed. As a result, the CADENCE What-If Tool can model disease spread across long- and short-range transmission routes, focusing on a single outbreak or multiple simultaneous cases.
In the next phase, Pallickara is working in collaboration with epidemiologists and economists at Kansas State University to move the tool into single and multi-player gaming environments, which would allow public officials to test multiple outbreak scenarios in real time.
The USAMM-Shiny App and the CADENCE What-If Tool both provide users with detailed insights about key aspects of the U.S. livestock industry. In funding their development, S&T seeks to strengthen state and local response capabilities and improve security for the nation’s food economy.
Visit the Homeland Security Advanced Research Projects Agency Chemical and Biological Defense Division’s webpage to learn more about S&T’s work in improving transboundary animal disease vaccines, agricultural screening, detection and diagnostics.