Correlation between the dynamics and spatial configuration of the circumarctic latitudinal forest-tundra ecotone

• Main Authors: Guo, W, Rees, G
• Format: Article in Journal/Newspaper
• Language: English
• Published: Informa UK Limited 2021
• Subjects: 37 Earth Sciences; 3704 Geoinformatics; Tundra
• Online Access: https://www.repository.cam.ac.uk/handle/1810/311691; https://doi.org/10.17863/CAM.58781

This study uses the MOderate Resolution Image Spectroradiometer (MODIS) Vegetation Continuous Fields (VCF) data to investigate the relationship between the dynamics of the circumarctic latitudinal forest-tundra ecotones (FTEs) and its spatial configuration. This study also explores the extent to which the Google Earth Engine (GEE) can enable efficient, large-scale and multi-temporal examination of the circumarctic FTE. FTE dynamics is represented by latitudinal interface movement over the 2000 to 2015 period. To represent spatial configuration of the latitudinal FTEs, we propose a continuous measurement of fragmentation derived from window spectral analysis, which builds on the theory of FTE ‘forms’ – a discrete categorisation of altitudinal FTEs primarily based on vegetation fragmentation. Additionally, continentality of the FTEs is calculated to investigate the impact of water variability. Statistical relationship between these variables are analysed in evenly divided longitudinal bands in the circumarctic region containing FTEs. More fragmented latitudinal FTEs are found to be more likely to shift northward in the study period, and FTEs in more continental areas are mostly more fragmented. These empirical results show that similar linkages exist between circumarctic latitudinal FTE fragmentation and dynamics to those found in altitudinal FTEs. This can potentially contribute to the optimisation of future climate modelling as well as modelling of vegetation reactions to climate change. The GEE platform enables the examination of latitudinal FTEs through efficient circumarctic-scale vegetation data assimilation and processing, and the methodological framework allows for future incorporation of additional variables of interest into the analysis. Cambridge Trusts, Trinity College Cambridge, Fitzwilliam College Cambridge