Inversion of Forest Leaf Area Index Calculated from Multi-source and Multi-angle Remote Sensing Data

Expand
  • 1National Engineering Research Center for Information Technology in Agriculture, Beijing 100097, China;
    2Department of Photogrammetry and Remote Sensing, Shandong University of Science and Technology, Qingdao 266510, China

Received date: 2009-08-05

  Revised date: 2009-12-14

  Online published: 2010-09-20

Abstract

Beijing small satellite, named BJ-1, and Landsat TM data were used to construct multi-source and four-angle datasets for inversion of forest leaf area index (LAI). Taking into account the vertical and 3-D distribution of forests, the hybrid model, INFORM, combining the geometric optical model and radiative transfer model, was used to support the retrieval model of LAI. The clustered method of ANN was utilized to obtain the information from forward INFORM- model simulated data under different groups of input parameters. After these steps, the inversion model was applied in different combinations of multi-angle under different levels of noise. The accuracy of inversion of forest LAI can be improved by adding observations of angle data if the quality of data is considered. Our data analysis resulted in an accuracy of R2=0.713, RMSE=0.957, which was 20% greater than the average accuracy of mono-angle data for inversion of LAI.

Cite this article

Guijun Yang, Wenjiang Huang, Jihua Wang, Zhurong Xing . Inversion of Forest Leaf Area Index Calculated from Multi-source and Multi-angle Remote Sensing Data[J]. Chinese Bulletin of Botany, 2010 , 45(05) : 566 -578 . DOI: 10.3969/j.issn.1674-3466.2010.05.006

Outlines

/

674-3466/bottom_en.htm"-->