This study examines the atmospheric (synoptic) controls on snowmelt and snow-free season carbon dioxide (CO2) fluxes at Daring Lake, Northwest Territories. Atmospheric circulation patterns were derived from 500 hPa geopotential height and classified using the self-organizing maps artificial neural network. Snowmelt timing was not found to be influenced by atmospheric circulation patterns or large-scale teleconnection indices, but a shift from meridional to zonal atmospheric circulation marked the transition from pre-melt to melt period. Multiple linear regression identified heating degree days and incoming solar radiation as the most important meteorological predictors of melt length; however, the model would have benefitted from additional variables. Analysis of CO2 (net ecosystem exchange, NEE) during the snow-free season highlighted a strong correlation between NEE and temperature anomalies. Like the snowmelt period, no atmospheric circulation patterns were found to significantly influence NEE; however, these findings prompt further questions regarding snowmelt and CO2 fluxes in the Canadian low Arctic.
Author Keywords: atmospheric circulation, carbon fluxes, net ecosystem exchange, self-organizing maps, snowmelt, synoptic patterns