Remote Sensing for Site-Specific Crop Management: Evaluating the Potential of Digital Multi-Spectral Imagery for Monitoring Crop Variability and Weeds within Paddocks

This paper analyses the potential and limitations of airborne remote sensing systems for detecting crop growth variability and weed infestation within paddocks at specified capture times. The detection of areas of crop growth variability can help farmers become aware of regions within their paddock where they may be experiencing above and below average yields due to changes in soil or management conditions. For instance, the early detection of weed infestation within cereal crops is crucial for lessening their impact on the final yield. Transect sampling within a canola paddock of a broad acre agricultural property in the South West of Western Australia was conducted synchronous with the capture of 1m spatial resolution DMSI. The four individual bands (blue, green, red and near- infrared) of the DMSI were correlated with LAI and weed density counts collected in the paddock. Statistical analyses show the LAI of canola had strong negative correlations with the blue (-0.93) and red (-0.89) bands and a strong positive correlation was found with the near-infrared band (0.82). The strong correlations between the canola LAI and selected bands of the DMSI indicate that this may be a suitable technique for monitoring canola variability to derive information layers that can be used in creating meaningful "within-field" management units. Likewise, DMSI could be used as a non-invasive tool for in season crop monitoring. The correlation analysis with the weed density (e.g. self sown wheat, ryegrass and clover) attributed to only one negative weak correlation with the red band (-0.38). The less successful detection of weeds is attributed to the minimal weeddensity within the paddock (e.g. mean 34 plants m-2) and indistinct spectral difference from canola at the early time of imagery capture required by farmers for effective variable rate applications of herbicides.