Influence of soil, crop residue, and sensor orientations on NDVI readings

Abstract

Site-specific in-season corn (Zea mays L.) nitrogen (N) rate recommendations based on remote sensing can increase nitrogen use efficiency (NUE) but most approaches require the corn to be at the V8 growth stage. Success in estimating N needs during early vegetative growth has been limited due to low plant biomass and background interference. The objective of this experiment was to measure the influence of soil series, soil moisture, surface crop residues, and sensor orientation on Normalized Difference Vegetation Index (NDVI) from soils prior to planting and corn from planting through the V6 growth stage. A controlled experiment was conducted in Virginia using four soil types commonly used for corn production. Spectral reflectance readings from the soils and four crop residues were measured using the GreenSeeker® sensor at four sensor orientations. Sensor orientation affected NDVI readings from bare soils before planting, with a difference of up to 0.16 units among orientations. The 15 cm mask generally resulted in the lowest NDVI. Wetting soils resulted in greater NDVI values from all soils with differences of up to 0.18 units between the same soil wet and dry. Values for NDVI were initially influenced by crop residue type but no differences due to type were detected once corn reached the V4 stage. Altering sensor orientation generally changed NDVI values but none resulted in NDVI that was similar across all soil and residue backgrounds. Background (soil and residue) influenced NDVI readings during early vegetative corn growth so the sensing background for the calibration reference areas should be uniform and similar to the larger field for implementation of the GreenSeeker® variable rate system.