The temporal variations of evapotranspiration (ET) and its controlling factors occur across time scales ranging from seconds to decades, with significant differences in the lag effects of ET drivers under varying water conditions. Therefore, identifying the dominant time scales of the relationships between ET and its controlling factors under varying water conditions is crucial for optimizing irrigation strategies of crops grown in a greenhouse. In our study, we utilized two years of continuous lysimeter observations of greenhouse tomato ET, and applied two water treatments: well-irrigated (0.9Epan, Epan is the cumulative pan evaporation) and deficit-irrigated (0.5Epan). Wavelet transform technology served as the core method to systematically examine the temporal variations of ET and its controlling factors. Observations indicated that the power spectra of ET featured pronounced peaks at daily and seasonal scales. The cospectra between ET and soil water content for greenhouse tomato revealed strong temporal correlation at 2~5 day scales, confirming the regulatory effect of irrigation cycles on ET. Moreover, ET variations were largely synchronous with net radiation, with ET lagging net radiation but leading vapor pressure deficit and air temperature at daily scales. In addition, significant disparities in phase angles between ET and individual meteorological variables were identified under 0.9Epan and 0.5Epan water conditions. Partial wavelet coherence revealed that net radiation was the primary meteorological driver of greenhouse tomato ET across multiple time scales, particularly at the daily scale, followed by vapor pressure deficit. These findings provide scientific evidence for selecting appropriate ET models at different time scales and offer valuable insights for optimizing water-saving irrigation for crops grown in greenhouses.
Gong et al. (Sun,) studied this question.