Highlights
High-frequency UAS thermal data can identify the temporal nature of the spatial canopy stress patterns for soybean.
Thermal indices were calculated using the statistical approach from the lower and upper bounds of confidence interval.
The CWSI Histogram Approach (UAS) was compared to the CWSI Empirical Approach (IRT).
The distribution of canopy temperature (using the inter-quartile range) may be useful for irrigation management.
Abstract. The use of unmanned aerial systems (UAS) in the field of irrigation management has been increasing rapidly. Due to their ability to capture multi-temporal data over the field, new techniques for the calculation of the crop water stress index (CWSI) and degrees above non-stressed (DANS) using UAS have been evolving. In this study, a statistical CWSI approach (canopy temperature histogram method) was used to identify the diurnal crop water stress patterns in soybean crop at three different study sites in Nebraska. Two study sites were located in the Eastern Nebraska Research and Extension Center (ENREC) at Mead, Nebraska, having multiple irrigation treatments; the third site was located in the South Central Agricultural Laboratory (SCAL), Clay Center, Nebraska, having one uniform irrigation treatment. Based on the results obtained, the CWSI and DANS maps exhibited a clear diurnal pattern of crop water stress response from morning to afternoon, and recovery from late afternoon to evening, with variations between the treatments at ENREC and a similar trend on SCAL. ENREC had a stronger correlation between CWSI and DANS due to the wider range in canopy temperatures from having both irrigated and rainfed plots. When compared between deficit plots at ENREC and the irrigated treatment at SCAL, the study showed that the statistical approach was more reliable when there were differences in crop water stress among different treatments. The main advantage of using the statistical CWSI histogram approach compared to the conventional empirical CWSI approach is the reduced requirement of additional meteorological parameters and faster automation time. CWSI histogram distribution graphs were created for each flight to understand the temporal changes and reveal the mean CWSI values (approximately 0.49, 0.51, and 0.49, for ENREC1, ENREC2, and SCAL, respectively) and interquartile (IQR) range for the soybean crop. For a given field site, temporal changes in IQR were greater than temporal changes in mean CWSI. Besides the mean canopy temperature, the distribution of canopy temperature (using the IQR) may be useful for irrigation management. Keywords: Irrigation Management, Precision Agriculture, Python, Remote Sensing, Thermal Imagery, Unmanned Aircraft Systems.