Assessment of Multi-Image Unmanned Aerial Vehicle Based High-Throughput Field Phenotyping of Canopy Temperature

Abstract

Canopy temperature (CT) has been related to water-use and yield formation in crops. However, constantly (e.g., sun illumination angle, ambient temperature) as well as rapidly (e.g., clouds) changing environmental conditions make it difficult to compare measurements taken even at short time intervals. This poses a great challenge for highthroughput field phenotyping (HTFP). The aim of this study was to i) set up a workflow for unmanned aerial vehicles (UAV) based HTFP of CT, ii) investigate different data processing procedures to combine information from multiple images into orthomosaics, iii) investigate the repeatability of the resulting CT by means of heritability, and iv) investigate the optimal timing for thermography measurements. Additionally, the approach was v) compared with other methods for HTFP of CT. The study was carried out in a winter wheat field trial with 354 genotypes planted in two replications in a temperate climate, where a UAV captured CT in a time series of 24 flights during 6 weeks of the grain-filling phase. Custom-made thermal ground control points enabled accurate georeferencing of the data. The generated thermal orthomosaics had a high spatial accuracy (mean ground sampling distance of 5.03 cm/pixel) and position accuracy [mean root-mean-square deviation (RMSE) = 4.79 cm] over all time points. An analysis on the impact of the measurement geometry revealed a gradient of apparent CT in parallel to the principle plane of the sun and a hotspot around nadir. Averaging information from all available images (and all measurement geometries) for an area of interest provided the best results by means of heritability. Correcting for spatial in-field heterogeneity as well as slight environmental changes during the measurements were performed with the R package SpATS. CT heritability ranged from 0.36 to 0.74. Highest heritability values were found in the early afternoon. Since senescence was found to influence the results, it is recommended to measure CT in wheat after flowering and before the onset of senescence. Overall, lowaltitude and high-resolution remote sensing proved suitable to assess the CT of crop

Current Practices in UAS-based Environmental Monitoring

“…Apart from the increased cost, these sensors are not always consumer-friendly, and can require significant additional resources to ensure calibration accuracy and sufficient user-training in order to extract useable, and scientifically relevant information. The standalone use of thermal sensors is adequate for the field monitoring of brightness temperature differences, especially for precision agriculture, as the canopy temperature is closely related to plant stress [41,42]. Furthermore, thermal sensors have proved to be useful in mapping urban heat islands [43], heat losses in large objects [44], soil water monitoring [11], and for animal detection [45].…”

“…Besides, the overlap also defines the angular composition of the data. Thus, a balanced across and along track overlap is beneficial [42,52].…”

“…Nevertheless, flight patterns should be developed in consideration of the weather conditions and especially the wind direction, as these can have a direct influence on retrieval quality (e.g., movement of vegetation affecting 3D reconstructions; impact of wind-speed on thermal measurements and sensors). If the area of interest has a particular structure, e.g., when the vegetation is planted in rows, it is beneficial to plan the flight path along these lines, to later use spatial correction algorithms to minimize the impact of fluctuating environmental conditions on the results [42]. The appropriate flight pattern significantly impacts the overall quality of the survey.…”

“…For imaging sensors, each pixel (due to its instantaneous field of view) has a different viewing geometry. Although the FOVs of 2D imaging sensors are rather small, it was shown that even small differences in the measurement geometry significantly affect the apparent reflectance [52,133,134] and thermal signal [42,135] when standard photogrammetric software is used. The reason for this is that when compiling an orthomosaic, the processing mode has a significant influence on which information from which pixel, and consequently with which measurement geometry, is used in the orthomosaic.…”

“…Moreover, when used as GCPs within the observed scene, these targets help to ensure accurate absolute geolocation of the thermal imagery [163]. For robust detection, composite materials made out of e.g., styrofoam and black metal, have been shown to work well with thermal data [42]. Alternatively, temperature monitored water baths (along with standard thermocouple assembly) can be installed within the observed scene to provide reference kinetic temperatures during the mission.…”