Analysis of the Effectiveness of the Red-Edge Bands of GF-6 Imagery in Forest Health Discrimination
Jiahui Chen, Hanqiu Xu
Abstract
The red-edge band is closely related to biochemical parameters that characterize the growth condition of green plants and is an important factor in monitoring vegetation health. Therefore, red-edge indices based on the red-edge band have been developed to measure vegetation health. However, due to the limited availability of satellites with a red-edge band, most existing red-edge indices were not developed based on satellite data. Fortunately, the launch of the GaoFen-6 (GF-6) satellite provides favorable conditions for monitoring vegetation health using satellite imagery, as it has two red-edge bands with spatial resolution of 16 m. To investigate the effectiveness of the red-edge bands on the GF-6 satellite in monitoring forest health, this study selected six red-edge indices and conducted tests in Zhangjiajie region in Hunan Province, China, and Hetian Basin in Fujian Province, China. The selected indices are the Normalized Difference Red-Edge Index 1 (NDRE1), the Modified Chlorophyll Absorption Ratio Index 2 (MCARI2), the Red-Edge Chlorophyll (CI <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">red-edge</sub> ), the Inverted Red-Edge Chlorophyll Index (IRECI), the Red-Edge Position (REP), and the Missouri emergency resource information system Terrestrial Chlorophyll Index (MTCI). The results showed that when applied to NDRE1 and CI <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">red-edge</sub> , the red-edge bands of GF-6 can effectively distinguish forest health conditions, with discrimination accuracy of 92.3% and 92.5%, respectively. However, the performance of the GF-6 red-edge bands with the other four indices yielded accuracy generally lower than 70%. Overall, the two red-edge bands added to the GF-6 satellite contribute to discerning forest health conditions, with NDRE1 and CI <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">red-edge</sub> being the preferred red-edge indices.