A Simple Temperature Domain Two‐source Model for Estimating Agricultural Field Surface Energy Fluxes from Landsat Images
Yunjun Yao,Shunlin Liang,Jian Yu,Jiquan Chen,Shaomin Liu,Yi Lin,Joshua B. Fisher,Tim R. McVicar,Jie Cheng,Kun Jia,Xiaotong Zhang,Xianhong Xie,Bo Jiang,Liang Sun
DOI: https://doi.org/10.1002/2016jd026370
2017-01-01
Journal of Geophysical Research Atmospheres
Abstract:A simple and robust satellite‐based method for estimating agricultural field to regional surface energy fluxes at a high spatial resolution is important for many applications. We developed a simple temperature domain two‐source energy balance (TD‐TSEB) model within a hybrid two‐source model scheme by coupling “layer” and “patch” models to estimate surface heat fluxes from Landsat thematic mapper/Enhanced Thematic Mapper Plus (TM/ETM+) imagery. For estimating latent heat flux (LE) of full soil, we proposed a temperature domain residual of the energy balance equation based on a simplified framework of total aerodynamic resistances, which provides a key link between thermal satellite temperature and subsurface moisture status. Additionally, we used a modified Priestley‐Taylor model for estimating LE of full vegetation. The proposed method was applied to TM/ETM+ imagery and was validated using the ground‐measured data at five crop eddy‐covariance tower sites in China. The results show that TD‐TSEB yielded root‐mean‐square‐error values between 24.9 (8.9) and 78.2 (21.4) W/m2 and squared correlation coefficient (R2) values between 0.60 (0.51) and 0.97 (0.90), for the estimated instantaneous (daily) surface net radiation, soil, latent, and sensible heat fluxes at all five sites. The TD‐TSEB model shows good accuracy for partitioning LE into soil (LEsoil) and canopy (LEcanopy) components with an average bias of 11.1% for the estimated LEsoil/LE ratio at the Daman site. Importantly, the TD‐TSEB model produced comparable accuracy but requires fewer forcing data (i.e., no wind speed and roughness length are needed) when compared with two other widely used surface energy balance models. Sensitivity analyses demonstrated that this accurate operational model provides an alternative method for mapping field surface heat fluxes with satisfactory performance.