ASSESSING METHODS FOR DISTINGUISHING BIODIVERSITY PATTERNS IN ALPINE GRASSLANDS USING SENTINEL – 2 DATA

Abstract

With climate change, many aspects of the natural state and environment of our planet is changing and will continue to change. Among these, the diversity of living organisms, or biodiversity, is affected. Biodiversity is a key factor for Earth and the living organisms’ survival; thus, it needs to be monitored to make sure that key species or not too many species are lost. In order to keep track of the changes in diversity a suitable and reliable method is needed to evaluate this. In this report, two data analysis methods using Sentinel-2 multispectral satellite images are examined, namely the spectral variation hypothesis (SVH) and seasonal maximum NDVI. The study has the aim of examining these methods’ ability to distinguish between two types of habitats, specifically calcareous and siliceous grasslands, with different ranges of species diversity at Mt. Nuolja near Abisko, Sweden. Both methods proved to be useful when comparing the species diversity between the habitats. The calcareous grassland habitat exhibited higher values in both SVH and maximum NDVI analyses, consistent with known biodiversity patterns in these areas. The SVH showed a potential to provide more detailed information about biodiversity patterns compared to the seasonal maximum NDVI. Statistical analyses confirmed a significant difference between the two grassland habitats from the SVH analysis, demonstrating the methods' applicability. Although the results from both methods varied, they complemented each other. Additionally, this study demonstrates that the choice of principal components matters for determining spectral diversity as incorporating more adds noise to the data causing reduced information quality. Overall, SVH proved to be a useful method for assessing biodiversity patterns in subarctic grasslands, potentially aiding in biodiversity conservation efforts by providing a method for the assessment of species diversity.