A Machine Learning Prediction Model of Weather Extremes for U.S. Agricultural Industry

PI: Stephanie Henderson, assistant professor of Atmospheric and Oceanic Sciences

Co-PIs:

Christopher Kucharik, professor of Plant and Agroecosystem Sciences

Zhou Zhang, associate professor of Biological Systems Engineering

Fraser King, assistant professor of Atmospheric and Oceanic Sciences

Sara Gragg, professor of Animal and Dairy Sciences

Jason Otkin, research associate professor Space Science and Engineering Center

Steve Vavrus, senior scientist UW Extension/State Climatology Office

Co-Investigator:

Jim Luedtke, professor of Industrial and Systems Engineering

Description: Persistent high-pressure systems, known as atmospheric blocks, are associated with weather extremes such as drought, heat waves (“heat domes”), flooding and cold air outbreaks. Blocking extremes cost the agricultural industry billions of dollars, yet no forecast model has been developed specifically targeting blocking events.

The recent development of a blocking database at UW-Madison has opened the pathway for such a model, and this team will use machine learning (ML) to accomplish this feat. This project will provide preliminary results needed for a large extramural proposal, including (1) quantifying the impacts of blocking extremes on crops and livestock, (2) assess insurance losses due to blocking extremes, (3) develop a prototype ML model to forecast blocks, and (4) connect with the agricultural community to understand their decision-making timescales and needs. This will be a first of its kind model tailored specifically for farmers and ranchers based on their needs.