Comparison of Rrs between satellite and in-situ measurements in Pacific and Southern Oceans

Abstract

Polar biodiversity should be monitored as an indicator of climate change. Biodiversity is mainly observed by field survey although this is very limited in broad inaccessible polar regions. Satellite imagery may provide valuable data with less bias, although spatial, spectral, and temporal resolutions are limited for analyzing biodiversity. The present study has two objectives. The first is constructing a first-ever vegetation map of the entire Barton Peninsula, Antarctica. The second is developing a monitoring method for long-term variation of vegetation, based on satellite images. Dominant mosses and lichens are distributed in small and sparse patches, which are limited to analysis using high-resolution satellite images. A subpixel classification method, spectral mixture analysis, is applied to overcome limited spatial resolution. As a result, vegetation shows high abundance along the southeastern shore and low-to-medium abundance in the nearly snowfree inland area. Even though spatial patterns of vegetation were almost invariant over 6 years, there was interannual variation in abundance aspects because of meteorological conditions. Therefore, extensive and long-term monitoring is needed for aspects of distribution and abundance. The present results can be used to design field surveys and monitor long-term variation as elementary data.