Publications

Luo, ZL; Guo, MJ; Bai, P; Li, J (2022). Different Vegetation Information Inputs Significantly Affect the Evapotranspiration Simulations of the PT-JPL Model. REMOTE SENSING, 14(11), 2573.

Abstract
Evapotranspiration (ET) is an essential part of the global water cycle, and accurate quantification of ET is of great significance for hydrological research and practice. The Priestley-Taylor Jet Propulsion Laboratory (PT-JPL) model is a commonly used remotely sensed (RS) ET model. The original PT-JPL model includes multiple vegetation variables but only requires the Normalized Difference Vegetation Index (NDVI) as the vegetation input. Other vegetation inputs (e.g., Leaf Area Index (LAI) and Fraction of Absorbed Photosynthetically Active Radiation (FAPAR)) are estimated by the NDVI-based empirical methods. Here we investigate whether introducing more RS vegetation variables beyond NDVI can improve the PT-JPL model's performance. We combine the vegetation variables derived from RS and empirical methods into four vegetation input schemes for the PT-JPL model. The model performance under four schemes is evaluated at the site scale with the eddy covariance (EC)-based ET measurements and at the basin scale with the water balance-based ET estimates. The results show that the vegetation variables derived by RS and empirical methods are quite different. The ecophysiological constraints of the PT-JPL model constructed by the former are more reasonable in spatial distribution than those constructed by the latter. However, as vegetation input of the PT-JPL model, the scheme derived from empirical methods performs best among the four schemes. In other words, introducing more remotely sensed vegetation variables beyond NDVI into the PT-JPL model degrades the model performance to varying degrees. One possible reason for this is the unrealistic ET partitioning. It is necessary to re-parameterize the biophysical constraints of the PT-JPL model to ensure that the model obtains reasonable internal process simulations, that is, getting the right results for right reasons.

DOI:
10.3390/rs14112573

ISSN:
2072-4292