Publications

Jiang, YZ; Tang, RL; Jiang, XG; Li, ZL; Gao, CX (2019). Estimation of Soil Evaporation and Vegetation Transpiration Using Two Trapezoidal Models From MODIS Data. JOURNAL OF GEOPHYSICAL RESEARCH-ATMOSPHERES, 124(14), 7647-7664.

Abstract
Soil evaporation (ETs) and vegetation transpiration (ETv) are essential in evaporative demand analyses and agricultural applications. This study estimated ETs and ETv using two trapezoidal models, that is, the conventional trapezoidal model and the end-member-based soil and vegetation energy partitioning (ESVEP) model, at three AmeriFlux tower sites. First, the land surface temperature of a Moderate Resolution Imaging Spectroradiometer (MODIS) pixel was decomposed into soil and vegetation temperatures (T-s and T-v, respectively) which involve important information related to ETs and ETv and were compared with corresponding temperatures from the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER). Then ETs and ETv were parameterized, and the total evapotranspiration (ET) was validated with the eddy covariance system measurements after energy imbalance corrections via the residual energy and Bowen ratio methods. The results showed that the ESVEP model partitioned T-s, with a root-mean-square error (RMSE) of 1.56-2.70 K (4.98-10.16%), and T-v, with an RMSE of 2.54-2.80 K (9.73-11.11%), indicating a higher accuracy than that from the conventional trapezoidal model, where the RMSE was 2.26-3.64 K (7.31-13.40%) for T-s separation and 2.82-4.62 K (9.80-15.49%) for T-v separation. For the ET estimation, the ESVEP model still performed better. The estimation RMSE was 49.8-58.4 W/m(2) (22.8-27.2%) for the ESVEP model and 72.0-89.5 W/m(2) (29.8-41.1%) for the conventional trapezoidal model, compared to the RE corrected measurements; the RMSE was 42.6-53.3 W/m(2) (24.4-28.1%) for the ESVEP model and 57.5-78.1 W/m(2) (32.3-37.9%) for the conventional trapezoidal model, compared to the Bowen ratio corrected measurements.

DOI:
10.1029/2019JD030542

ISSN:
2169-897X