Eddy covariance (EC) is an established technique that measures carbon exchange across the terrestrial-atmospheric interface, and is used to evaluate climate- and management-driven changes to carbon cycling across landscapes. EC systems have been widely used across the globe for more than 20 years, for example as part of the international consortium called FLUXNET, and offer researchers invaluable measurements including net ecosystem exchange and ecosystem respiration, which includes the contribution from vegetation and soil.
While the EC technique itself measures the net fluxes of gas to and from an ecosystem, ancillary measurements are often needed to assist with data analysis and interpretation, and data quality control and quality assurance (Reichstein et al., 2005). In order to fulfill these needs, a variety of parameters, including temperature, moisture, canopy gas storage and soil gas flux (respiration) are commonly used alongside EC measurements. However, the various EC assumptions and technical obstacles may be causing biases in gas exchange estimates by creating issues including gaps in datasets due to technical failures or poor environmental conditions, or by introducing other systematic and random measurement error into the gas exchange data. Measurements of soil respiration (RS) at the ground level may help alleviate some of these issues, and may provide opportunities for rich information that may help in the interpretation of EC data and ecosystem carbon dynamics.
Here are the Top 5 reasons that measuring soil respiration (RS) can benefit EC research:
To obtain robust carbon flux estimates, EC data processing requires that the air movement is above a particular turbulence threshold (u* filtering). This can be particularly important at night when the air masses above a forest cool, or following a rainstorm, and can result in hours or even days of lost data. Gap-filling models are thus necessary to generate data during these periods, but need to be constrained by real world data including nighttime RS, when the carbon exchange is assumed to be only from the soil (Refer to Moffatt et al 2007 for a review of several gap filling techniques).
2. Data QA/QC
EC data is prone to uncertainty due to technical issues, including instrument malfunction. One method to ensure data quality is to compare computed temperature sensitivity (Q10) from RS and soil temperature to the computed temperature sensitivity of ecosystem respiration from the EC towers during the night when photosynthesis is not active. Deviation of Q10 from seasonal averages can help to determine poor quality data (Refer to a previous blog post: AGU 2015 poster: Nickerson et al., 2015), and this approach could be applied to EC data as well.
3. Using Nighttime RS to Constrain RECO
Nighttime exchange, as long as turbulent mixing is adequate, is assumed to be a perfect measurement of nighttime respiration. Therefore, these measurements of soil gas efflux are commonly used to evaluate the quality of nighttime gas flux measurements as made by the eddy covariance technique (Baldocchi, 2003). However, systematic errors in EC carbon fluxes have been consistently observed, but need to be corrected for accurate measurement of C exchange. For instance, Lavigne et al. 1997, noticed that nighttime EC-derived carbon exchange were 20-35% lower than nighttime soil respiration measurements. This systematic underestimation needs to be reconciled.
4. Scaling Up and Down
The mid-level spatial scale of continuous RS data is useful. It may be scaled up to the footprint level of the EC tower with adequate spatial coverage for the EC site, and may be also used to partition RS into autotrophic (root) and heterotrophic (microbial) components within the particular site, especially using isotopic and/or RS isolation approaches (girdling or trenching). These components are particularly important when trying to understand how particular components of ecosystem carbon exchange may be altered differently by climate or land management. Watch for an upcoming article by Phillips et al. (in prep) for an excellent review.
5. Catch the “Hot” Moments
The potentially high-temporal frequency of RS measurements (by automated systems, for example) allows for detection of rapid changes in respiration due to moisture and temperature variation within the EC footprint over heterogeneous terrain (i.e. non-ideal sites). At the synoptic or weather event-based scale, large shifts in moisture/temperature (e.g. events such as freeze thaw, dry-rewet) are especially important to consider at sites with large moisture and temperature variability (Savage et al 2009). These rapid changes are important to quantify as they may have a large influence on C exchange not explained adequately using the half-hourly temporal scale measured by most EC systems.
To sum it all up…
Overall, EC datasets could be improved by addressing gaps and validation issues, through the inclusion of continuous soil gas flux measurements. To date, RS measurements have been used sparingly alongside EC towers because of the large cost to scale chamber systems to the EC footprint, as well as large data integration and processing burdens. Eosense’s eosFD chamber allows for inexpensive and autonomous measurements of RS, providing a more scalable approach to matching the EC footprint when compared to other commercial RS systems.
A deployment of Esosense’s eosFD chambers is being carried out at Howland Forest, Maine, this summer. Howland is home to one of the longest running EC sites in the AmeriFlux network, and will be an important test case for integrating RS and EC data, hopefully leading improved estimates of carbon exchange.
Come visit our poster at ESA (morning of August 12) in Fort Lauderdale for an update of this work!