In‐Season Prediction of Corn Grain Yield Potential Using Normalized Difference Vegetation Index

Drastic increases in the cost of N fertilizer and increased public scrutiny have encouraged development and implementation of improved N management practices. This study evaluated the relationship between corn ( Zea mays L.) grain yield and early season normalized difference vegetation index (NDVI) sensor readings using the GreenSeeker sensor. The relationships between grain yield and several predictor variables were determined using linear and nonlinear regression analysis. Categorizing NDVI measurement by leaf stage indicated that growth stage was critical for predicting grain yield potential. Poor exponential relationships existed between NDVI from early sensor measurements (V6–V7 leaf stage) and grain yield. By the V8 stage, a strong relationship ( R 2 = 0.77) was achieved between NDVI and grain yield. Later sensor measurements (V9 and later) failed to distinguish variation in green biomass as a result of canopy closure. Normalizing the NDVI with GDD (growing degree days) did not significantly improve yield potential prediction ( R 2 = 0.73), but broadened the yield potential prediction equation to include temperature and allowed for adaptation into various climates. Sensor measurements at the range of 800 to 1000 GDD resulted in a significant exponential relationship between grain yield and NDVI ( R 2 = 0.76) similar to the V8 leaf stage categorization. Categorizing NDVI by GDD (800–1000 GDD) extended the sensing time by two additional leaf stages (V7–V9) to allow a practical window of opportunity for sidedress N applications. This study showed that yield potential in corn could be accurately predicted in season with NDVI measured with the GreenSeeker sensor.