Document: ERI-3-59-71

Comparing estimates of soil respiration using chamber-based and tower-based methods.

DAVIDSON, E.A.* and K.SAVAGE

The Woods Hole Research Center, Woods Hole, MA 02543 ¹

Abstract:
Soil respiration can be measured directly using chamber techniques, and it can be inferred from tower-based eddy covariance estimates of total ecosystem respiration. Unfortunately, chamber-based estimates of soil respiration are often 50% higher than total ecosystem respiration estimated by eddy covariance. The causes of this discrepancy are not known. Chamber techniques underestimate gas fluxes when the natural concentration gradient is altered by the presence of the chamber, and this error can be minimized by using short measurement intervals. Variation of pressure within the chamber can also cause either over- or underestimation of fluxes. Pressure differentials should vary with the flow rates of gas circulated between the chamber and the CO₂ analyzer, but we have found no effect of flow rate on estimated CO₂ flux, indicating that this source of error is probably insignificant. Stable conditions at nighttime complicate eddy covariance estimates, possibly causing underestimation of respiration. In addition to possible systematic methodological errors of both methods, differences exist regarding the spatial and temporal variation over which fluxes are estimated. Chamber measurements are often made in the area around a tower, but the tower may be affected by soil emissions several hundred meters away from the tower. At the Howland Forest of central Maine, we estimated mean soil CO₂ flux from chamber measurements for each soil drainage class within the landscape and applied these to a fine resolution soil map covering 12 km² of the tower area. We then overlayed onto this map some plausible tower footprint areas that estimate the area sampled by the eddy covariance method. Depending on the wind direction and the assumed size of the tower footprint, the estimated soil source of CO₂ that would be detected by tower-based eddy covariance measurements varied by about 20%. Hence, spatial variation of soil drainage classes is an important factor for estimating soil respiration at the scale relevant to tower measurements, but spatial variation alone does not account for large differences often observed between tower-based eddy covariance estimates of respiration and soil chamber measurements commonly made near the tower base.

Keywords: soil respiration, eddy covariance, chamber methods, CO2, carbon

This abstract is being presented at: 9:30 AM in session:
Oral Session #25: Organic Matter Dynamics in Ecosystems.