Methane Fluxes Measured by Eddy Covariance at a Temperate Upland Forest in Central Ontario
Methane flux measurements were carried out at a temperate upland forest in Central Ontario, Haliburton Forest and Wildlife Reserve (45.28° N, 78.55° W) using the eddy covariance (EC) method. An off-axis integrated cavity output spectrometer (OA-ICOS) Fast Greenhouse Gas Analyzer (FGGA from Los Gatos Research, Inc.) operated at a sampling rate of 10 Hz allowed for simultaneous measurement of methane (CH4), carbon dioxide (CO2), and water (H2O) over five months from June to October in 2011. Air was pulled from the top of a 32 m tower, 8 m above the forest canopy, to the bottom of the tower through 40 m of tubing to the instrument. A sonic anemometer and a LI-7500 open-path sensor were also used at the top of the tower to provide high frequency wind data and comparative open-path measurements of CO2 and H2O. A nearby soil station measured soil water content and soil temperature at 0, 3, and 10 cm below the surface. The measurements were made in an uneven-aged managed forest last harvested 15 years ago containing sandy and acidic soils (pH 4.0 5.0).
The eddy covariance method was employed to calculate fluxes using an averaging time of 30 minutes throughout the measurement period. The lateral separation between the sonic anemometer and inlet of the FGGA was corrected for, and the tilt of the sonic anemometer was adjusted by the planar fit technique. For the closed-path OA-ICOS system, high frequency loss due to tube damping was corrected, and density fluctuations due to water vapour fluxes were taken into account using Webb-Pearman-Leuning (WPL) correction, using high frequency water concentration measurements also made by the closed-path FGGA. To ensure that turbulent conditions were met, only fluxes during night time with a friction velocity above 0.1 m s-1 were included in the data analysis.
Through the measurement period, the average mixing ratio of CH4 was 1.89 ppm with a mean flux value of -1.48 nmol m-2 s-1 (negative value indicates uptake). In the beginning of June when measurements commenced, the soil moisture was relatively high and CH4 flux values showed net emission. As the season advanced the soil became progressively drier, and there was an increasing trend in CH4 uptake, peaking in mid-September. There was also a diurnal trend in the CH4 flux, with increased uptake during the day, and decreased uptake between 0:00 and 09:00, shown in the figure. In the figure, the mean flux over all five months of sampling for each hour of the day is plotted, along with the 95% confidence interval for the mean value.
Although there was strong correlation between soil water content and CH4 flux, there was very little diurnal variability in the measured soil water content. There were diurnal trends in both soil temperature at 3 cm, and wind speed in the forest canopy, which matched well with the diurnal trend observed in the CH4 flux, but individual CH4 flux values did not correlate well with soil temperature. The CH4 flux values did correlate well with the horizontal wind speed measured within the forest canopy; as the wind speed increased from 0.1 to 0.9 m s-1, the average CH4 uptake increased from -1.25 nmol m-2 s-1 to -4.37 nmol m-2 s-1. We hypothesize that this may be due to a ventilation effect in which higher wind speed facilitates the introduction of CH4-rich air and removes CH4-depleted air near the methanotrophs in the soil. Alternately, the diurnal pattern in CH4 uptake could be attributed to increased daytime uptake of soil water by vegetation, which can lead to drier soil in proximity to root systems. While not reflected in the bulk soil moisture measurement, this could lead to localized increases in CH4 uptake due to dry conditions favouring aerobic methanotrophy near the roots.
Chamber flux measurements of CH4, CO2, and N2O were also performed at seven toposequences around the tower, every two weeks from June to October. The permanently installed PVC collars were capped and sampled every 30 minutes over a 90 minute period to calculate a flux. Along with observing a seasonal trend in methane fluxes that is similar to the tower measurements, the chamber measurements show that spatial patterns in soil moisture lead to differences in CH4 flux values. Excluding chamber measurements from significantly wet soils in low-lying areas, which showed very high emissions, the overall average from the chamber sites in the topographical gradients was -1.87 nmol m-2 s-1, which matches well with the eddy covariance flux value. This suggests that throughout most of the summer, particularly after the wetter month of June, only a small fraction of the landscape acts as a net source for methane. The consistency between canopy and soil-level methane flux measurements also indicates that there is very little probability of significant aerobic CH4 emissions from above-ground vegetation at this site.