Wavelet Analysis Identifies Carbon Processes in a Subarctic Stream During Snowmelt Spring Flood

Snowmelt spring floods dominate the annual carbon flux in Arctic streams. However, climate change is altering their timing and magnitude due to changes in snow conditions, further altering the processes controlling the carbon cycle at the catchment scale. Current knowledge is limited by a lack of high-resolution data from Arctic areas. In this study we combine high-resolution biogeochemical-hydro-climatological variables with spectral wavelet analysis for new insights into carbon processes.

This study was conducted during the snowmelt spring flood period in a sub-arctic headwater catchment in Pallas-Ylläs national park, Finland (68°02′N, 24°16′W). We collected in-stream dissolved organic carbon (DOC), carbon dioxide (C0₂), and terrestrial C0₂ flux alongside a suite of hydro-climatological variables measured at 30-minute intervals. Continuous wavelet transformations and wavelet coherence were produced to assess the relationship between hydro-climatological variables and carbon variables at different periodicities.

Wavelet transforms indicated that the onset of snowmelt caused the development of significant diel periodicity for in-stream DOC, CO₂ and terrestrial CO₂ flux, while substantial periods of significant periodicity were observed at multiple day periodicities. Wavelet coherence analysis identified that DOC was consistently lead by flow and conductivity across daily and multiple daily scales suggesting that transport of carbon from the surface and shallow sub-surface pathways to the stream were the predominant processes controlling in-stream DOC. Interestingly for in-stream CO₂, groundwater level showed periodic rather than consistent spectral coherence suggesting it is not a consistent control on CO₂ in the spring flood. The strongest coherence for in-stream CO₂ was with in-stream O₂, which may suggest the importance of in-stream metabolism as a control on in-stream CO₂ dynamics. Terrestrial CO₂ fluxwas controlled by notably different processes than in-stream Carbon and linked strongest to climatological variables. Photosynthetically active radiation (PAR) showed the strongest relationship with CO₂ terrestrial flux dynamics. 

Our study highlights the unique processes controlling different parts of the carbon cycle in a headwater arctic catchment during the snowmelt spring flood. We highlight in-stream DOC as particularly vulnerable to changes in spring flood magnitude and timing given the importance of snowmelt dominated transport processes to DOC flux. To identify future changes in the Arctic carbon cycle, wavelet analysis shows potential as tool to analyse changes in processes in high-resolution datasets.