Using RapidEye imagery to identify within-field variability of crop growth and yield in Ontario, Canada

Abstract

Remote sensing has been recognized as a cost-effective way to detect the spatial and temporal variability of crop growth and productivity. In this study, multispectral RapidEye images were used to delineate homogeneous zones of soil and crop development in two fields in Ontario, Canada, one planted with canola (Brassica napus L.) and the other with spring wheat (Triticum aestivum L.). The two fields received different levels of nitrogen (N) treatments during the pre-planting land preparation phase. Soil textures, mineral nitrogen content and crop yield were used to interpret the results of zone delineation. The analysis of variance (ANOVA) tests revealed that the high-resolution RapidEye data, particularly the imagery acquired at peak crop growth stages (i.e. when leaf area index (LAI) is high), provided valuable information for delineating within-field variability of crop growth and yield. Further analysis showed that for both crops, the spatial patterns of crop growth condition varied throughout the growth cycle, revealing different impacts of soil properties and N fertilization on the crops. In particular, during peak growth stage, the within-field variability was most strongly affected by the pre-planting N application and had the strongest correlation with crop yield. These results suggest that high-resolution satellite data (e.g., RapidEye) could assist in making decisions on optimal N fertilization for enhanced crop productivity.