Transpiration characteristics of different soybean varieties based on the Three-Temperature Model and thermal infrared remote sensing

Transpiration is an important process in the water cycle and is the key parameter to accurately quantify water use efficiency. Thus, it is of great importance for studying the relationship between the carbon and water cycles and for developing water-saving agricultural practices. The major objective of this study was to quantitatively study the transpiration rate of soybean plants of different varieties and under different water stress conditions, to identify differences in temporal and spatial characteristics, and finally, to provide a reference for the selection of drought-resistant and water-saving soybean varieties. Therefore, two soybean varieties (C08 and J21) were selected as the research objects and two water stress conditions (75%A0 and 37.5%A1 of the local empirical irrigation quota) were used for each variety. Based on the Three-Temperature Model (3T model) and using thermal infrared remote sensing, transpiration was quantified in the different soybean varieties under different water stress conditions. The diurnal variation in transpiration rate of the soybean plants under different water stress conditions was basically consistent with temperature, net solar radiation (R_n), and canopy temperature (T_c), showing a single-peak curve that first increased and then decreased, reaching a peak at value between 1.2 mm·h^-1 and 2.5 mm·h^-1 at noon. Moreover, the canopy temperature and transpiration rate of soybean plants under different treatments showed obvious spatial heterogeneity. Under different water stress conditions, C08 and J21 soybean varieties showed canopy temperatures in the order A0 < A1, with means of 6.55 K and 5.91 K, respectively. Transpiration rates were in the order of A0 > A1, with averages of 0.28 mm·h^-1 and 0.29 mm·h^-1, respectively. Transpiration rates were positively correlated with irrigation and negatively correlated with canopy temperature. Under the same water stress conditions, canopy temperatures were in the order of C08 < J21, with the mean canopy temperature of J21 1.83-2.47 K lower than that of C08. In addition, transpiration rates were of the order J21 < C08, with the mean transpiration rate of the J21 soybean variety 0.13-0.14 mm·h^-1 higher than that of the C08 soybean variety. Thus, the J21 soybean variety consumes more water than the C08 variety under the same conditions of water stress. In combination with crop growth indicators, such as leaf area index (LAI) and crop yield, these data provide an important reference for improving crop water productivity in the future. Compared with traditional methods, the method used in this study has some advantages. The 3T model requires fewer parameters which are easy to be measured through introducing the concept of reference soil. The high-resolution thermal infrared instrument used here can reach the millimeter scale and meets the accuracy requirements of crop transpiration rate measurement in the farmland microclimate environment. Therefore, crop transpiration estimation based on the 3T model and thermal infrared remote sensing technology is convenient and accurate and is of scientific significance in promoting efficient agricultural water use and selecting water-saving crop varieties.