作者:Xu, BD (Xu, Baodong)[ 1,2,3 ] ; Li, J (Li, Jing)[ 2,4 ] ; Liu, QH (Liu, Qinhuo)[ 2,4 ] ; Huete, AR (Huete, Alfredo R.)[ 5 ] ; Yu, Q (Yu, Qiang)[ 6,7 ] ; Zeng, YL (Zeng, Yelu)[ 1,2,3 ] ; Yin, GF (Yin, Gaofei)[ 8 ] ; Zhao, J (Zhao, Jing)[ 2,4 ] ; Yang, L (Yang, Le)[ 2,4 ]

IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING

卷: 9  期: 7  页: 3267-3282  特刊: SI

DOI: 10.1109/JSTARS.2016.2560878

出版年: JUL 2016

摘要

Continuous leaf area index (LAI) observations from global ground stations are an important reference dataset for the validation of remotely sensed LAI products. In this study, a pragmatic approach is presented for evaluating the spatial representativeness of station-observed LAI dataset in the product pixel grid. Three evaluation indicators, including dominant vegetation type percent (DVTP), relative absolute error (RAE) and coefficient of sill (CS), were established to quantify different levels of spatial representativeness. The DVTP was used to evaluate whether the station-observed vegetation type was the dominant one in the pixel grid, and the RAE and CS were applied to quantify the point-to-area consistency for a given station observation and the spatial heterogeneity caused by the different density of vegetation within the pixel, respectively. The proposed approach was applied to 25 stations from the Chinese Ecosystem Research Network, and results show significant differences of representativeness errors at different levels. The spatial representativeness for different stations varied seasonally with different vegetation growth stages due to temporal changes in heterogeneity, but the spatial representativeness remained consistent at interannual timeframes due to the relatively stable vegetation structure and pattern between adjacent years. A large error can occur in MOD15A2 product validation when the representativeness level of station LAI observations is low. This approach can effectively distinguish various levels of spatial representativeness for the station-observed LAI dataset at the pixel grid scale, which can consequently improve the reliability of LAI product validation by selecting LAI observations with a high degree of representativeness.

通讯作者地址: Li, J; Liu, QH (通讯作者)

       Joint Ctr Global Change Studies, Beijing 100875, Peoples R China.

通讯作者地址: Li, J; Liu, QH (通讯作者)

Chinese Acad Sci, State Key Lab Remote Sensing Sci, Beijing 100101, Peoples R China.

地址:

[ 1 ] State Key Lab Remote Sensing Sci, Beijing 100101, Peoples R China

[ 2 ] Joint Ctr Global Change Studies, Beijing 100875, Peoples R China

[ 3 ] Univ Chinese Acad Sci, Coll Resources & Environm, Beijing 100049, Peoples R China

[ 4 ] Chinese Acad Sci, State Key Lab Remote Sensing Sci, Beijing 100101, Peoples R China

[ 5 ] Univ Technol, Plant Funct Biol & Climate Change Cluster, Sydney, NSW 2007, Australia

[ 6 ] Univ Technol, Sch Life Sci, Sydney, NSW 2007, Australia

[ 7 ] Northwest A&F Univ, State Key Lab Soil Eros & Dryland Farming Loess P, Yangling 712100, Peoples R China

[ 8 ] Chinese Acad Sci, Inst Mt Hazards & Environm, Chengdu 610041, Peoples R China