We analysed the carbon dioxide fluxes from vegetation, measured by portable flux chambers, and several meteo-climatic and environmental variables obtained during six summer campaigns at two different sites in the High-Arctic tundra in the Bayelva basin near Ny Ålesund, Svalbard, during the period 2019–2025. The two sites have similar climatic conditions and vegetation cover, in terms of species distribution and abundance, but different distance from the glaciers. The measured fluxes provided information on Net Ecosystem Exchange ( NEE ), Ecosystem Respiration ( ER ) and Gross Primary Production ( GPP ) on 674 single measurement points. A single statistical, data-driven model on the whole dataset explained up to 53% of the variance of the data for ER , and 67% of the variance for GPP . Besides the classic dependences on soil temperature for ER and solar irradiance for GPP , the main drivers were found to be Green Fractional Cover and soil Volumetric Water Content. Site-specific modelling revealed limited differences between the two sites, while interannual variability was more consistent. • Arctic tundra ER / GPP fluxes are strongly affected by environmental drivers. • Green Fractional Cover and Volumetric Water Content are key drivers of the fluxes. • Empirical models describe 53–67% of the ER / GPP variability. • Nonlinear models provide better descriptions than linear ones for the ER and GPP fluxes.
Valdengo et al. (Tue,) studied this question.