Proximal active optical sensing operational improvement for research using the CropCircle ACS-470, implications for measurement of normalized difference vegetation index (NDVI)

Authors: Conley, Matthew; Thompson, Alison; Hejl, Reagan

Submitted to: Sensors
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 5/22/2023
Publication Date: 5/24/2023
Citation: Conley, M.M., Thompson, A.L., Hejl, R.W. 2023. Proximal active optical sensing operational improvement for research using the CropCircle ACS-470, implications for measurement of normalized difference vegetation index (NDVI). Sensors. 23(11). Article 5044. https://doi.org/10.3390/s23115044.
DOI: https://doi.org/10.3390/s23115044

Interpretive Summary: Being able to monitor and evaluate plant characteristics in field conditions to assess growth, yield, or stress adaptations is a very important process in plant research. High-throughput plant phenotyping (HTTP) is a new approach to accomplish this in which sensors are often mounted on a sensing platform to capture measurements. These sensors can be sensitive to temperature and since HTTP is utilized in areas were temperatures are extreme, performance and accuracy of the instruments can be impacted. The purpose of this study was to characterize a customizable proximal sensor available for HTTP research under changes in temperature and in field conditions, and to provide useful guidance for the use of this sensor to end-users. Sensor performance was measured using large titanium-dioxide white painted field normalization reference panels, and the consequent sensor detector values as well as sensor body temperatures were recorded. The white panel reference measurements illustrated that individual filtered sensor detectors subjected to the same change in thermal environment can behave differently. The enhanced protocol from this paper is expected to improve raw instrument signal quality by an average of 0.24% per 1 °C. These recommendations are provided to support ACS-470 by using white panel normalization and sensor temperature stabilization.

Technical Abstract: Active radiometric reflectance is useful to determine plant characteristics in field conditions. However, the physics of silicone diode-based sensing are temperature sensitive, where a change in temperature affects photoconductive resistance. High-throughput plant phenotyping (HTPP) is a modern approach using sensors often mounted to proximal based platforms for spatiotemporal measurements of field grown plants. Yet HTPP systems and their sensors are subject to the temperature extremes where plants are grown, and this may affect overall performance and accuracy. The purpose of this study was to characterize the only customizable proximal active reflectance sensor available for HTPP research, including a 10 C° increase in temperature during sensor warmup and in field conditions, and to suggest an operational use approach for researchers. Sensor performance was measured at 1.2 m using large titanium-dioxide white painted field normalization reference panels and the expected detector 1.0 unity values as well as sensor body temperatures were recorded. The white panel reference measurements illustrated that individual filtered sensor detectors subjected to the same thermal change can behave differently. Across 361 observations before and after field collections where temperature changed by more than one degree, filtered detector values changed an average of 0.24% per 1 °C. Recommendations based on years of sensor control data and plant field phenotyping agricultural research are provided to support ACS-470 researchers by using white panel normalization and sensor temperature stabilization.