Comparison between two vegetation indices for measuring different types of forest damage in the north-eastern United States

Publisher Website
Google Scholar
Abstract
The relative effectiveness of Landsat Thematic Mapper (TM) derived normalized difference vegetation index (NDVI) and short-wave infrared/near infrared index (SWIR/NIR) for measuring different types of damage in several forest communities was examined and compared. Study areas included: (1) Twin Mountain, New Hampshire, a site with well-defined fir waves; (2) Camels Hump mountain, Vermont, a site undergoing well-documented coniferous forest decline; and (3) a predominantly deciduous region seriously impacted by pear thrips in southern Vermont and north-western Massachusetts. Results indicated that the SWIR/NIR was excellent in distinguishing between high and low conifer damage for both fir wave and forest decline sites. The NDVI was less satisfactory in these conifer-dominated sites. High and low deciduous damage sites were easily separable using both NDVI and SWIR/NIR. The NDVI also allowed excellent separation between medium and low deciduous damage. Results from this study suggest that both indices are appropriate for global-scale monitoring of vegetation, with the SWIR/NIR being suitable for monitoring both deciduous and coniferous conditions and the NDVI being appropriate for monitoring broadleaf conditions.