Multi-scale spectral characteristics of latent heat flux over flooded rice and winter wheat rotation system

• Published in: Agricultural water management Vol. 288; p. 108471
• Main Authors: Zhang, Tao, Qiu, Rangjian, Ding, Risheng, Wu, Jingwei, Clothier, Brent
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01-10-2023; Elsevier
• Subjects: Continuous wavelet transform; Latent heat flux; Net radiation; Orthogonal wavelet transform; Partial wavelet coherence; Rice-winter wheat field; Net radiation; Partial wavelet coherence; Orthogonal wavelet transform; Rice-winter wheat field; Latent heat flux; Continuous wavelet transform
• ISSN: 0378-3774; 1873-2283
• DOI: 10.1016/j.agwat.2023.108471

The temporal variations of latent heat flux (λET) and its affecting factors vary from seconds to decades, and alter with crop species. Understanding the temporal coherence between λET and its influencing factors across different underlying surfaces is a crucial research topic with practical implications, and can enable better water management. To investigate this, we conducted a study in southern China measuring water and heat fluxes over flooded rice-winter wheat rotation fields from 2017 to 2021. Wavelet transform technology was employed to analyze the spectral properties of λET and its affecting factors. Results showed that the power spectra of λET exhibited different cascade laws for rice and winter wheat. The spectral variabilities of λET occurred at daily and seasonal time scales for both rice and winter wheat, and an additional weekly time-scale for rice. Furthermore, the cospectrum between λET and soil water content (SWC) for rice showed a significant temporal correlation at 2–4 days, indicating the period of water input. We found that the daily λET of both crops changed synchronously with daily net radiation (Rn), and preceded daily vapor pressure deficit (VPD) and air temperature (Ta) by 1.5–2.6 h. The phase angle between λET and Ta was significantly lower for rice (mean = 1.8 h) than winter wheat (mean = 2.3 h). Partial wavelet coherence revealed that Rn, followed by VPD, was the main meteorological factor affecting λET for the rotated flooded rice-winter wheat system at each time scale, especially at the daily time scale. Additionally, the effect of VPD on λET was lower for winter wheat than rice at scales below a month. These findings offered a useful insight into selecting models of λET for varying time scales and promoting better agricultural water management.
•Multiscale spectral characteristics of latent heat fluxes (λET) over rotated rice and wheat were analyzed.•The spectral peaks of λET occurred at daily and seasonal scales for both crops, and an additional weekly scale for rice.•The phase angle between λET and air temperature was significantly lower for rice than wheat.•The effect of vapor pressure deficit on λET was greater for rice than for wheat at scales below a month.