Assimilating MODIS data-derived minimum input data set and water stress factors into CERES-Maize model improves...

Area of Science:

• Agricultural Science
• Remote Sensing
• Agronomy

Background:

• Crop growth models and remote sensing offer valuable insights but face challenges in regional yield prediction.
• Accurate regional crop yield forecasting is crucial for food security and agricultural management.

Purpose of the Study:

• To enhance regional corn yield prediction accuracy by assimilating Moderate Resolution Imaging Spectroradiometer (MODIS) products into a crop growth model.
• To evaluate the performance of this assimilation strategy in Illinois, a major US corn-producing state.

Main Methods:

• Utilized the Crop Environment Resource Synthesis (CERES)-Maize model with minimal inputs.
• Estimated planting dates using a phenology model and MODIS-derived Leaf Area Index (LAI).
• Determined cultivar genetic coefficients by minimizing differences between observed and simulated LAI.

Main Results:

• The assimilation of MODIS-derived water stress factors and LAI under auto-irrigation conditions yielded the highest prediction accuracy (R² = 0.78, RMSE = 0.75 t ha⁻¹).
• The developed strategy successfully predicted yearly corn yields at a regional scale.

Conclusions:

• The integration of MODIS data into crop growth models provides a successful approach for regional yield prediction.
• Further research should explore the spatial portability of this strategy across diverse agro-climatic and agro-technology regions.