Northeastern Section - 42nd Annual Meeting (12–14 March 2007)

Paper No. 3
Presentation Time: 8:15 AM-12:00 PM

COMPARISON OF AUTOCHAMBER SOIL EFFLUX MEASUREMENTS TO EDDY COVARIANCE TOWER ECOSYSTEM RESPIRATION AT HARVARD FOREST


PHILLIPS, Steve1, VARNER, Ruth K.2, FROLKING, Steve3, MUNGER, J. William4, CRILL, Patrick M.5 and WOFSY, Steven C.4, (1)Department of Earth Sciences, University of New Hampshire, 56 College Rd, Durham, NH 03824, (2)Climate Change Research Center, Institute for the Study of Earth, Oceans, and Space, University of New Hampshire, Durham, NH 03824, (3)Complex Systems Research Center, Institute for the Study of Earth, Oceans, and Space, University of New Hampshire, Durham, NH 03824, (4)Division of Engineering and Applied Sciences, Harvard University, Cambridge, MA 02138, (5)Institutionen för Geologi och Geokemi, Stockholm University, Stockholm, Sweden, steve.phillips@unh.edu

Soil carbon dioxide (CO2) emission has been measured semi-continuously using an automated chamber system at Harvard Forest from April through November during the years 2003-2005. There are eight automatic chambers on a slope from a beaver pond to upland, near the eddy covariance at the Harvard Forest Environmental Measurement Site.

We compared the soil flux (SF) as measured by the autochambers to both the nighttime net ecosystem exchange measurements (NEE) and gap-filled ecosystem respiration (ER) from the eddy covariance tower. Tower flux measurements are only available up through 2004; only the years 2003 and 2004 were included in the analysis.

Direct comparison of the soil flux to the tower nighttime NEE and gap-filled ER demonstrated a distinct seasonality of the ratio between SF and ecosystem respiration. During the spring and early summer, SF was less than ER; however, during the late summer and fall, the autochamber measurements were often larger than tower. In principle, for a given land area SF cannot exceed ER.

To observe variation along the moisture gradient, the autochamber data was grouped into three locations along the slope. The four autochambers closest to the beaver pond are referred to as the “wetland margin” chambers, the two partway up the slope are the “mid-slope” chambers, and the two at the top of the slope are the “upland” chambers. SF from the three sections of the slope were compared to the gap-filled ER from the tower for both 2003 and 2004. By plotting the three sections of the slope separately it is clear that the wetland margin, mid-slope, and upland chambers follow a different seasonal magnitude. The wetland margin chamber SF fluxes stay below the total ER values. The upland chambers approach values of about 100% of tower ER, while the mid-slope SF surpasses the tower, most distinctly in 2004. In any case, SF measured by the chambers comprises a larger fraction of ER at the tower during late summer/early fall.