Using and Verifying Hyper Local Model Forecast Data for Operational Agricultural Decision-Making

Farmers manage risk through a planning cycle that involves decision making at different time scales. Throughout this cycle, weather and climate information can play a critical role in effective risk management. There is frequently a mismatch between the spatial resolution of weather and climate data, and the field-scale decisions to be made. For agricultural operations such as irrigation, pest management, and weed control, high-resolution numerical weather prediction models assimilating both in situ and remotely sensed Earth observations have the potential to address this scale mismatch. Our current efforts focus on using and verifying precipitation, wind speed, and temperature forecasts from the U.S. National Oceanic and Atmospheric Administration (NOAA) High-Resolution Rapid Refresh (HRRR) model. We assess both deterministic and probabilistic HRRR forecast skill relative to other NOAA models including the North American Mesoscale Forecast System and Global Forecast System, as well as data from a commercial weather provider. Such verification based on direct comparison with observations allows us to quantify the accuracy of HRRR relative to other models at resolving weather variation for operational field-scale applications. The conference presentation will highlight results from multiple locations over the continental U.S.