Estimating non-additive within-season temperature effects on maize yields using Bayesian approaches

Area of Science:

• Agricultural Science
• Climate Science
• Environmental Science

Background:

• Extreme heat events demonstrably harm U.S. crop yields.
• Limited research has explored the effects of intra-seasonal weather variability and temperature-precipitation interactions on crop production.

Purpose of the Study:

• To highlight the significance of analyzing temperature exposures at a finer temporal scale within the growing season.
• To quantify the impact of monthly temperature and precipitation fluctuations on maize yields in the U.S. Corn Belt.

Main Methods:

• Employed variable selection techniques for climate-yield relationship modeling.
• Utilized Bayesian estimation methods to model maize yield responses to climate variables.
• Compared models using disaggregated monthly weather data against those using aggregated seasonal data.

Main Results:

• Models incorporating within-growing season monthly weather variations outperformed those using aggregated seasonal data.
• Bayesian estimation demonstrated effectiveness in capturing complex climate-yield interactions.
• Predicted warming impacts on maize yields were smaller when using disaggregated monthly data compared to aggregated data.

Conclusions:

• Disaggregating temperature exposures within the growing season is essential for accurate crop yield impact assessments.
• Temperature effects on maize yields are non-additive across different months.
• Finer-scale climate data analysis provides more nuanced predictions of agricultural impacts under changing climate conditions.