Detection of damaged pine stands utilizing IRS satellite data
2001
Kishikawa, M. (Niigata Univ. (Japan). Faculty of Agriculture) | Abe, N.
Using LISS-3 and PAN data obtained from the IRS satellite, we attempted to detect pine stands damaged by pine wilt disease. Analysis of the LISS-3 data revealed that the normalized difference vegetation index (NDVI) of the pixels corresponding to the survey plot tended to increase with increasing degree of damage. The value of NDVI is considered to be high due to the vigorous growth of the broad-leaved trees invading the stand after the damage of pine trees and that of undergrowth in the open canopy. The satellite data were photographed in July when undergrowth was vigorously growing. The high NDVI value obtained may be related to this time of photographing. It was possible to detect the damaged pine stands from a simple equation between NDVI and the degree of damage. Since the pixel size of the satellite image obtained from the PAN data did not agree with the size of survey plot, we examined the relationship between the average digital number of pixels involved in the plot and the degree of damage. The average digital number tended to increase with increasing degree of damage. This is because the tone of the broad-leaved trees that had invaded after the damage of pine trees and of undergrowth in the open canopy was brighter than that of the pine trees as revealed from the LISS-3 data. The possibility of extraction of damaged stands by visual interpretation of such tone differences was suggested.
显示更多 [+] 显示较少 [-]