NSF R2I2: Developing Subseasonal to Seasonal Weather Prediction Tools to Meet Agricultural Needs in the Midwest

About This Grant

Extreme weather and related environmental impacts disproportionately affect the agricultural sector, even when such events are rare. Yet agricultural stakeholders lack access to skillful operational subseasonal to seasonal (S2S) forecasts that could inform and empower them to take timely mitigative actions. This is due in part to sharp recent decreases in the capacity of global prediction models to be accurate beyond a two-week forecast lead time. Weather extremes represent the tail distribution of a variable and are generally less predictable than its mean, i.e., typical weather can be forecast relatively accurately while the extremes are becoming more unpredictable. For the Midwest, this is partly due to the lack of significant collaboration between meteorologists and agricultural stakeholders in producing, formatting, and disseminating extreme weather prediction information. The overarching goal of the proposal is to develop skillful and actionable S2S predictions of weather extremes and crop yields to help mitigate negative impacts on Midwest agriculture. By addressing the impacts of extreme events, the project will improve agricultural production and enhance food security. It will also aid the adoption of best practices by helping farmers overcome information gaps and uncertainty barriers, thereby improving agricultural sustainability and resilience. The project will develop skillful S2S predictions of weather extremes by effectively utilizing predictability sources and leveraging machine learning methods. The latest advancements in prediction and AI methods will be combined with the complex and situation-specific knowledge of agricultural operations. In Phase I of the project, the team will: 1) thoroughly investigate the S2S predictability of various weather extremes and environmental impacts affecting Midwest agriculture; 2) develop prototype machine-learning-based models for skillful S2S predictions of extreme precipitation; 3) develop a scalable phenology-integrated crop yield forecasting model; and 4) develop and strengthen partnerships with stakeholders for the co-development of S2S prediction products to mitigate agricultural impacts. Phase II of the project would focus on implementation and dissemination of skillful and actionable S2S predictions targeting weather extremes that are predictable on S2S timescales and have high impacts on Midwest agriculture. The proposed work will fill a gap in operational forecasts and help advance stakeholder understandings of S2S variability and predictability of extremes. Additionally, the integration of S2S predictions and phenology-integrated crop yield forecasts will help quantify the agricultural impacts of weather extremes to overcome information barriers for agricultural stakeholders and policymakers. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.