HR: 0800h
AN: B11B-0144 [Abstracts]
TI: A Simple and Effective Image Normalization Method to Monitor Boreal Forest Change in a Siberian Burn
Chronosequence across Sensors and across Time
AU: * Chen, X
EM: xuchen@usgs.gov
AF: Institute of Atmospheric Sciences, South Dakota School of Mines and Technology,, 501 East Saint Joseph
Street, Rapid City, SD 57701
United States
AU: * Chen, X
EM: xuchen@usgs.gov
AF: Current address: SAIC,USGS/EROS Data Center,, 47914 252nd Street,, Sioux Falls, SD 57198
United States
AU: Vierling, L A
EM: Lee.Vierling@sdsmt.edu
AF: College of Natural Resources, University of Idaho, PO Box 441142, Moscow, ID 83844
United States
AU: Deering, D W
EM: Donald.W.Deering@nasa.gov
AF: NASA, Goddard Space Flight Center, Greenbelt, MD 20771
AB:
Satellite data offer unique perspectives for monitoring and quantifying land cover change, however, the radiometric
consistency among co-located multi-temporal images is difficult to maintain due to variations in sensors and atmosphere. To
detect accurate landscape change using multi-temporal images, we developed a new relative radiometric normalization scheme:
the temporally invariant cluster (TIC) method. Image data were acquired on 9 June 1990 (Landsat 4), 20 June 2000, and 26
August 2001 (Landsat 7) for analyses over boreal forests near the Siberian city of Krasnoyarsk. Normalized Difference
Vegetation Index (NDVI), Enhanced Vegetation Index (EVI), and Reduced Simple Ratio (RSR) were investigated in the
normalization study. The temporally invariant cluster (TIC) centers were identified through a point density map of the base
image and the target image and a normalization regression line was created through all TIC centers. The target image digital
data were then converted using the regression function so that the two images could be compared using the resulting common
radiometric scale. We found that EVI was very sensitive to vegetation structure and could thus be used to separate conifer
forests from deciduous forests and grass/crop lands. NDVI was a very effective vegetation index to reduce the influence of
shadow, while EVI was very sensitive to shadowing. After normalization, correlations of NDVI and EVI with field collected
total Leaf Area Index (LAI) data in 2000 and 2001 were significantly improved; the r-square values in these regressions
increased from 0.49 to 0.69 and from 0.46 to 0.61, respectively. An EVI øcancellation effectñ where EVI was positively
related to understory greenness but negatively related to forest canopy coverage was evident across a post fire
chronosequence. These findings indicate that the TIC method provides a simple, effective and repeatable method to create
radiometrically comparable data sets for remote detection of landscape change. Compared with some previous relative
normalization methods, this new method can avoid subjective selection of a normalization regression line. It does not
require high level programming and statistical analyses, yet remains sensitive to landscape changes occurring over seasonal
and inter-annual time scales. In addition, the TIC method maintains sensitivity to subtle changes in vegetation phenology
and enables normalization even when invariant features are rare.
DE: 1640 Remote sensing
DE: 0360 Transmission and scattering of radiation
DE: 0400 Biogeosciences
DE: 0840 Evaluation and assessment
DE: 0910 Data processing
SC: Biogeosciences [B]
MN: 2004 AGU Fall Meeting