Assessment of sensor-based technologies for monitoring crop growth and development in cereals

Abstract

A three year study was undertaken to identify sensor-based technologies that may enable the characteristics of a growing cereal crop to be monitored, inputs determined, and to establish ways in which such technologies could effectively be implemented.

Working with agronomists, a review of the current methods used to monitor soil, crop and weed factors was undertaken. The roles that these factors have in determining the required field and crop inputs for crop establishment, weed control, crop protection and crop nutrition were established. Canopy characteristics can have some influence on inputs of both crop protection chemicals and fertiliser which account for approximately 60% of the variable costs of winter wheat.

Previous research has commonly used spectral reflectance techniques to measure crop canopy characteristics. However, the information from this approach has been limited because of the complex nature of the crop canopy that has many interrelated properties.

Use of a number of sensing techniques working in combination could provide a better characterisation of the crop canopy. A canopy measurement system incorporating both spectral reflectance and ultrasonic height sensing techniques was designed, built and tested over two growing seasons on a range of winter wheat varieties planted at different seed rates on different soil types.

Results from the canopy measurement system indicated:

Normalised difference vegetation index (NDVI) values were useful up to growth stage 31 and beyond growth stage 59 whereas ultrasonic sensors proved useful for monitoring the canopy beyond growth stage 30 and up to growth stage 59. Combining these two measurements enabled the crop to be monitored over the complete growing season.

By using a combination of normalised difference vegetation index and ultrasonic crop height measurements

• Crop height was estimated to an accuracy of ± 0.09 m,
• Tiller numbers were estimated to an accuracy of ± 125 tiller m-2 and
• Leaf Area Index was estimated to an accuracy of ± 0.47 without the need for direct calibration using destructive sampling methods.

These values can be used directly with canopy management principles to aid the agronomic decision making process to determine the optimum level of inputs of fertilisers, fungicides and growth regulators based on the recommendations of the agronomist. The findings held true for all the variety, seed rate and soil type combinations used in this study.