Hysteresis Between Canopy Temperature and Atmospheric Temperature and Their Drivers on Winter Wheat

Abstract Quantitative characterization of the time-lag effect between canopy temperature and atmospheric temperature and its controlling factors in the agricultural ecosystem may contribute to a higher inversion accuracy of soil water content using canopy-air temperature information. For this end, the canopy temperature (Tc) of winter wheat at four irrigation levels, W1 (field capacity of 95%), W2 (field capacity of 80%), W3 (field capacity of 65%) and W4 (field capacity of 50%), were continuously monitored, and the data of such environmental factors as solar radiation (Rs), atmospheric temperature (Ta), relative humidity (RH) and soil water content (SWC) were simultaneously collected. With the synchronous diurnal time series, the lag relationship between Tc and Ta was analyzed, and the lag time (LT) was calculated using time-lagged correlation analysis; multiple linear regression model was used to construct the time lag model based on factors of Rs, Ta, RH and SWC, and path analysis was used to study the interaction among the factors. The results showed: (1) hysteresis existed between Tc and Ta over the diel cycles, and different weather and irrigation levels did not change the direction of the time lag loop. (2) the key driver regulating the diel hysteresis pattern between Tc and Ta varied under different weather: on rainy days, key driver was Rs while on cloudy and sunny days, the key driver was RH. Meanwhile, SWC had some effect on hysteresis under a certain threshold.(3) the multiple regression model indicated that together Rs, Ta, RH, and SWC explained 68 ± 3, 48 ± 8, and 64 ± 3% of the variation of time-lag effect on rainy, cloudy, and sunny days, respectively. Path analysis showed that the key driver could enhance the time-lag effect through other indirect factors. These findings clearly indicated a dynamic process of time-lag effect between Tc and Ta with different weather and different irrigation levels. This study contributes to the understanding of the time-lag effect and its driving factors and this analysis provides the basis for further improvement on monitoring crop water deficit.