Journal of Remote Sensing & GIS

Abstract

Relationship of Vegetation Indices Derived from Sentinel 2A Image with Canopy Cover and Production

Jamal Imani

The present study was conducted to determine the best ground and spectral resolution (Sentinel 2A images) for estimating vegetation cover and production. Ground sampling was performed in three plant communities in two forms of six and three plots. Different dimensions of the nesting plot (including 1×1, 2×1, 2×2 and 3×3) were used to estimate production and cover. Samples were taken in each community within 30 pixels along three transects. Densities of dominant plants were calculated by counting the bases in 2×2 plots, of vegetation cover as estimated and production was measured by double sampling in relation to of cover. Also, the distribution of dominant species was determined by statistical tests. The results showed that in community 1, using 10 m resolution bands, NDVI, CTVI, MSAVI2, Ratio, RVI, SAVI and TVI had significant relationship with of cover and production. In community 1, 1×1 plot has no significant relationship and valid model and in other plots, three-plot sampling method has very low correlation and the resulting models are insufficiently valid while sampling method is insufficient but The six plots method has a significant correlation. In community 2, the indices of NDVI, Ratio, RVI and TSAVI1 have significant relationships. In community 3, except plots 1×1 and 2×2, other plots had good relationships in the six plots sampling method and NDVI, MSAVI2, Ratio and TVI indices had good relationships in this community. In using of Sentinel 2, with bands 60 meters, the results are somewhat different. In community 1, MSAVI2 and RVI indices, in community 2, TSAVI1 and RVI and in community 3, NDVI and Ratio have significant relationships. In using these bands (with a resolution of 60 m), both sampling methods are less correlated. In community 1, 2×1 plot with six plots method, in community 2, 3×3 plot with three plots method and in community 3, 2×2 plot with six plot method is suitable.

Published Date: 2021-09-10;