To extract quantitative biophysical parameters such as leaf biomass and leaf chlorophyll concentration from the remotely sensed imagery, the effect of atmospheric attenuation must be removed. The refined empirical lin...To extract quantitative biophysical parameters such as leaf biomass and leaf chlorophyll concentration from the remotely sensed imagery, the effect of atmospheric attenuation must be removed. The refined empirical line (REL) method was used to calibrate the IKONOS multispectral imagery. The IKONOS digital numbers (DN) were converted to the at-satellite re- flectance, then the linear relation between at-satellite reflectance and surface spectral reflectance (ρλ) was derived from six bright targets of known reflectance in the image, and modelled estimates of the image reflectance at ρλ=0. Validation targets were used to test the feasibility of REL method. The mean relative errors between ρλ retrieved from IKONOS image using REL method and ground-measured ρλ were 11%, 13%, 3% and 5% in the IKONOS blue, green, red and near-infrared (NIR) respectively. When dark targets are unavailable or measurement of dark target is inconvenient, the REL method was most crucial for retrieving surface spectral reflectance. The REL offers a simple approach for quantitative retrieval of biophysical parameters from IKONOS im- agery.展开更多
基金Project supported by the Hi-Tech Research and Development Pro-gram (863) of China (Nos. 2002AA130010-2-7 and 2003AA131020- 04-06) and the National Natural Science Foundation of China (No. 40171065)
文摘To extract quantitative biophysical parameters such as leaf biomass and leaf chlorophyll concentration from the remotely sensed imagery, the effect of atmospheric attenuation must be removed. The refined empirical line (REL) method was used to calibrate the IKONOS multispectral imagery. The IKONOS digital numbers (DN) were converted to the at-satellite re- flectance, then the linear relation between at-satellite reflectance and surface spectral reflectance (ρλ) was derived from six bright targets of known reflectance in the image, and modelled estimates of the image reflectance at ρλ=0. Validation targets were used to test the feasibility of REL method. The mean relative errors between ρλ retrieved from IKONOS image using REL method and ground-measured ρλ were 11%, 13%, 3% and 5% in the IKONOS blue, green, red and near-infrared (NIR) respectively. When dark targets are unavailable or measurement of dark target is inconvenient, the REL method was most crucial for retrieving surface spectral reflectance. The REL offers a simple approach for quantitative retrieval of biophysical parameters from IKONOS im- agery.
基金Project supported by the National Natural Science Foundation of China (No.40171065)the National High Technology Research and Development Program of China (Nos.2002AA130010-2-7 and 2003AA131020-04-06).
