The advancement in satellite technologies has equipped the agricultural industry with tools to better understand the crop response to natural intra-paddock variability and management practices. Remote sensing devices such as satellite, airborne, or ground-based sensors can be used to measure a range of crop responses that can be correlated to a spectral signatures.
Stability analysis aims to quantify and automate a process that compares data between several seasons, based on the Normalised Difference Vegetation Index (NDVI). It calculates whether an area in a field is a stable or unstable production region. This is valuable for strengthening our understanding of crop variability, but also when this data is used in targeted applications. This project evaluated the use of NDVI derived from high resolution satellite imagery (IKONOS) in stability analysis over a three year period.
NDVI was derived from three IKONOS images captured between 2004 and 2006, in October and November; which correlates to when the crop is in peak biomass. Images were georeferenced to a master image that had been orthorectified. The three NDVI layers were combined into a multi-band raster dataset and the coefficient of variation of the intersecting pixels was calculated. Areas with high CV were deemed unstable.
Maps were derived that show the areas of the paddock that are unstable, and of the stable areas, whether the mean NDVI was high, medium or low. This paper will discuss the processes of stability analysis, and how issues when dealing with high resolution imagery were overcome.