Impact of Canopy Decoupling and Subcanopy Advection on the Annual Carbon Balance of a Boreal Scots Pine Forest as Derived From Eddy Covariance

Abstract

AbstractApparent net uptake of carbon dioxide (CO2) during wintertime by an ∼ 90 year old Scots pine stand in northern Sweden led us to conduct canopy decoupling and subcanopy advection investigations over an entire year. Eddy covariance (EC) measurements ran simultaneously above and within the forest canopy for that purpose. We used the correlation of above‐ and below‐canopy standard deviation of vertical wind speed (σw) as decoupling indicator. We identified 0.33 m s−1 and 0.06 m s−1 as site‐specific σw thresholds for above‐ and below‐canopy coupling during nighttime (global radiation <20 W m−2) and 0.23 m s−1 and 0.06 m s−1 as daytime (global radiation >20 W m−2) σw thresholds. Decoupling occurred in 53% of the annual nighttime and 14% of the annual daytime. The annual net ecosystem exchange (NEE), gross ecosystem exchange (GEE), and ecosystem respiration (Reco) derived via two‐level filtered EC data were −357 g C m−2, −1,138 g C m−2, and 781 g C m−2, respectively. In comparison, both single‐level friction velocity (u*) and quality filtering resulted in ~ 22% higher NEE, mainly caused by ~ 16% lower Reco. GEE remained similar among filtering regimes. Accounting for changes of CO2 storage across the canopy in the single‐level filtered data could only marginally decrease these discrepancies. Consequently, advection appears to be responsible for the major part of this divergence. We conclude that the two‐level filter is necessary to adequately address decoupling and subcanopy advection at our site, and we recommend this filter for all forested EC sites.