A satellite-based hybrid algorithm to determine the Priestley-Taylor parameter for global terrestrial latent heat flux estimation across multiple biomes

作者:Yao, YJ (Yao, Yunjun)[ 1,2 ] ; Liang, SL (Liang, Shunlin)[ 1,2,3 ] ; Li, XL (Li, Xianglan)[ 2 ] ; Chen, JQ (Chen, Jiquan)[ 4 ] ; Wang, KC (Wang, Kaicun)[ 2 ] ; Jia, K (Jia, Kun)[ 1,2 ] ; Cheng, J (Cheng, Jie)[ 1,2 ] ; Jiang, B (Jiang, Bo)[ 1,2 ] ; Fisher, JB (Fisher, Joshua B.)[ 5 ] ; Mu, QZ (Mu, Qiaozhen)[ 6 ] ; Grunwald, T (Gruenwald, Thomas)[ 7 ] ; Bernhofer, C (Bernhofer, Christian)[ 7 ] ; Roupsard, O (Roupsard, Olivier)[ 8,9 ]

REMOTE SENSING OF ENVIRONMENT

卷: 165  页: 216-233

DOI: 10.1016/j.rse2015.05.013

出版年: AUG 2015

摘要

Accurate estimation of the terrestrial latent heat flux (LE) for each plant functional type (PET) at high spatial and temporal scales remains a major challenge. We developed a satellite-based hybrid algorithm to determine the Priestley-Taylor (PT) parameter for estimating global terrestrial LE across multiple biomes. The hybrid algorithm combines a simple empirical equation with physically based ecophysiological constraints to obtain the sum of the weighted ecophysiological constraints (f(e)) from satellite-based normalized difference vegetation index (NDVI) and ground-measured air temperature (TO, relative humidity (RH), vapor pressure deficit (VPD) and LE for 2000 to 2009 provided by 240 globally distributed FLUXNET eddy covariance (ECOR) tower sites. Cross-validation analysis indicated that the optimization at a PFT level performed well with a RMSE of less than 0.15 and a R-2 between 0.61 and 0.88 for estimated monthly f(e). Cross-validation analysis also revealed good performance of the hybrid-based PT method in estimating seasonal variability with a RMSE of the monthly LE varying from 43 W/m(2) (for 6 deciduous needleleaf forest sites) to 18.1 W/m(2) (for 34 crop sites) and with a R-2 of more than 0.67. The algorithm's performance was also good for predicting among-site and inter-annual variability with a R-2 of more than 0.78 and 0.70, respectively. We implemented the global terrestrial LE estimation from 2003 to 2005 for a spatial resolution of 0.05 degrees by recalibrating the coefficients of the hybrid algorithm using Modern Era Retrospective Analysis for Research and Applications (MERRA) meteorological data, Moderate Resolution Imaging Spectroradiometer (MODIS) NDVI product and ground-measured LE. This simple but accurate hybrid algorithm provides an alternative method for mapping global terrestrial LE, with a performance generally improved as compared to other satellite algorithms that are not calibrated with tower. The calibrated f(e) differs for different PFTs, and all driving forces of the algorithm can be acquired from satellite and meteorological observations. (C) 2015 Elsevier Inc. All rights reserved.

通讯作者地址: Yao, YJ (通讯作者)

Beijing Normal Univ, Beijing 100875, Peoples R China.

地址:

[ 1 ] Beijing Normal Univ, Sch Geog, State Key Lab Remote Sensing Sci, Beijing 100875, Peoples R China

[ 2 ] Beijing Normal Univ, Coll Global Change & Earth Syst Sci, Beijing 100875, Peoples R China

[ 3 ] Univ Maryland, Dept Geog Sci, College Pk, MD 20742 USA

[ 4 ] Michigan State Univ, CGCEO Geog, E Lansing, MI 48823 USA

[ 5 ] CALTECH, Jet Prop Lab, Pasadena, CA 91109 USA

[ 6 ] Univ Montana, Dept Ecosyst & Conservat Sci, Numer Terradynam Simulat Grp, Missoula, MT 59812 USA

[ 7 ] Tech Univ Dresden, Inst Hydrol & Meteorol, Chair Meteorol, D-01062 Dresden, Germany

[ 8 ] CIRAD, UMR Eco&Sols Ecol Fonct & Biogeochim Sols & Agroe, F-34060 Montpellier, France

[ 9 ] CATiE Trop Agr Ctr Res & Higher Educ, Turrialba 7170, Costa Rica