Location: Soil and Water Management ResearchTitle: Comparison of stationary and mobile canopy sensing systems for maize and soybean in Nebraska, USA
|BHATTI, SANDEEP - University Of Nebraska|
|HEEREN, DEREK - University Of Nebraska|
|Evett, Steven - Steve|
|MAGUIRE, MITCHELL - University Of Nebraska|
|KASHYAP, SURESH - University Of Nebraska|
|NEALE, CHRISTOPHER - University Of Nebraska|
Submitted to: Applied Engineering in Agriculture
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 1/27/2022
Publication Date: 2/10/2022
Citation: Bhatti, S., Heeren, D.M., O'Shaughnessy, S.A., Evett, S.R., Maguire, M.S., Kashyap, S.P., Neale, C.M. 2022. Comparison of stationary and mobile canopy sensing systems for maize and soybean in Nebraska, USA. Applied Engineering in Agriculture. 38(2):331-342. https://doi.org/10.13031/aea.14945.
Interpretive Summary: Freshwater available for irrigation is decreasing, especially in regions where water for irrigation is from aquifers with limited recharge like the Ogallala; therefore, there is a need to maximize the efficient use of irrigation water. There is growing interest in using feedback from plant sensors to monitor crop canopy temperature to aid irrigations scheduling, primarily infrared thermometers (IRTs) mounted on ground-based platforms. However, sensor use on such platforms is not widely practiced and there is some skepticism about the practicality of ground-based platforms relative to satellite or aerial platforms. In this study, researchers at the University of Nebraska and ARS-Bushland compared measurements of canopy temperature from IRTs mounted on the sprinkler lateral with measurements of IRTs that were stationary in the field, and with measurements from a thermal camera mounted on an unmanned aerial system (UAS). Results showed that canopy temperature measurements from the IRTs mounted on the sprinkler lateral were similar to those from stationary IRTs in the field when the sprinkler traveled in a dry mode viewing the same area of canopy. Canopy temperature measurements from the UAS disagreed with those from the IRTs by approximately 7 degrees F. Differences in canopy temperature between the UAS and the IRTs may be due to differences in pixel size and field-of-view; further investigation is required. Additional research is required on the application of spectral radiometers mounted on a moving ground-based platform for the purpose of aiding in irrigation scheduling; these results should instill confidence in researchers and others to utilize a moving sprinkler system as a platform for IRTs for the purpose of monitoring crop canopy temperature.
Technical Abstract: Irrigation can potentially be better managed by utilizing data collected from various sensors installed on different platforms. The accuracy and repeatability of each data source are important considerations when selecting a sensing system suitable for irrigation management. This study compared data from multispectral (MS) and thermal sensors mounted on different platforms to investigate their comparative usability and accuracy with regard to irrigation management. The different sensor platforms included stationary posts fixed on the ground, the lateral of a center pivot irrigation system, unmanned aircraft systems (UAS), and Planet satellites. The data from MS sensors were used to compute the Normalized Difference Vegetation Index (NDVI) and Soil Adjusted Vegetation Index (SAVI). Field conditions included rainfed and irrigated treatments. Irrigation was applied according to a spatial evapotranspiration model informed with Planet satellite imagery. The NDVI and SAVI curves computed from the different sensing systems exhibited similar patterns and were able to capture differences between the rainfed and irrigated treatments during senescence. Strong correlations were observed for canopy temperature measurements between the stationary and pivot-mounted infrared thermometer (IRT) sensors when scans were done with no irrigation application (dry scans). The best correlation was obtained for the maize irrigated crop, which yielded correlation coefficient of 0.99, RMSE of 0.4 degrees C, and MAE of 0.3 degrees C. The correlation between UAS and pivot-mounted thermal sensors for dry scan data was weak with coefficient correlation equal to 0.26 to 0.28, larger RMSE values of 3.7 degrees C and MAE values of 3.4 degrees C. Secondary analysis between thermal data from stationary and pivot-mounted IRTs collected during wet scans (during an irrigation event) demonstrated reduced canopy temperature from pivot-mounted IRTs by approximately 2 degrees C for soybean due to wetting of the canopy by the irrigation. Understanding the performance of these sensor systems is valuable in configuring practical design and operational considerations when using sensor feedback for irrigation management.