“…Recently, the magnitude of soil CO 2 efflux has seemed to depend on tundra greening and boreal forest browning (Verbyla, 2008;Bhatt et al, 2010;Parent and Verbyla, 2010), as well as on the timing of both snow disappearance and the snow-covered period (Stone et al, 2002;McDonald et al, 2004;Sturm et al, 2005). During the seasonally snow-covered period, winter CO 2 efflux measurements have heretofore been gathered in tundra (Oechel et al, 1997;Fahnestock et al, 1998Fahnestock et al, , 1998Björkman et al, 2010;Kim et al, 2013), alpine and subalpine forests (Brooks et al, 1996;Mast et al, 1998;Monson et al, 2006a, b), and boreal forests Winston et al, 1995Winston et al, , 1997Kim et al, 2007, accounting for 10-30 % of the variability in annual carbon emissions. However, it is difficult to determine the timing of snow disappearance in the early spring season, due to shorter snow disappearance time spans, including changes of −0.13 day year −1 over 60 years in Barrow, Alaska, according to NOAA/CMDL historical data (Stone et al, 2002), and −0.94 day year −1 over 14 years according to microwave remote sensing (McDonald et al, 2004).…”