Measuring GHG emissions - the use of eddy covariance techniques

23 September 2008



Eddy covariance techniques are being used at the Eastmain-1 reservoir in Canada for hydroelectric reservoir CO2 flux measurements. Marie-Claude Bonneville and Dr Ian B Strachan explain how this will help assess the impact of reservoir impoundment on GHG emissions


The scientific community represented by the Intergovernmental Panel on Climate Change is in agreement that the earth’s climate is warming. It is understood that the rise in the global average temperature is largely due to the increasing emissions of greenhouse gases (GHG), especially through the burning of fossil fuel, but also because of land conversion. GHG have the ability to absorb infrared radiation, thus warming the earth’s surface, and contributing to the greenhouse effect. The four major GHGs are carbon dioxide (CO2), methane (CH4), nitrous oxide (N2O), and water vapour.

One way of reducing GHG concentrations in the atmosphere is through sequestration of carbon in soils and vegetation. Carbon dioxide is naturally absorbed by terrestrial ecosystems through plant photosynthesis, and is released to the atmosphere through decomposition and respiration. The balance between CO2 uptake and emission is termed the net ecosystem exchange (NEE). The direction and magnitude of NEE is determined by a variety of environmental and ecological factors, such as vegetation type, soil moisture regime, meteorological conditions, and organic matter content among others. Usually, vegetated ecosystems, such as forests and peatlands, are net sinks for atmospheric CO2 because these systems absorb more CO2 from the atmosphere through photosynthesis during the growing season than they emit through respiration on an annual basis. When a forest is cut or burned, most of the CO2 returns to the atmosphere. However, what happens to the carbon when an existing boreal ecosystem gets inundated?

Answering this question is the goal of an innovative project being undertaken by researchers from McGill University, the Université de Québec à Montréal (UQAM) and Environnement Illimité Inc., in partnership with Hydro Québec of Canada. Identifying the impacts of the creation of hydroelectric reservoirs on GHG emissions is an issue that must be addressed, especially in northern Quebec, where there is a huge potential for the production of hydroelectricity.

The Eastmain-1 (EM-1) reservoir in the James Bay region provides a unique opportunity to perform a comprehensive assessment of such impacts, and to use the eddy covariance technique for the first time in such a large scale project on hydroelectric reservoir CO2 flux measurements.

Eddy covariance technique

Many techniques have been used to measure the quantity of greenhouse gases that are exchanged between the atmosphere and a given surface. In addition to the static chambers used to quantify GHG emissions in different locations of the EM-1 reservoir surroundings, the research team is utilising the eddy covariance (EC) technique. This is now widely used by environmental researchers to measure the NEE of CO2, since it provides continuous flux information integrated at the ecosystem scale (Baldocchi, 2003).

The general principle behind the EC technique for measuring GHG fluxes is that as the air flows over a surface, turbulent eddies are created. These eddies transport energy, water vapour and trace gases away from and towards the surface. The instruments used in an EC system are mounted on top of towers and allow the measurement of the concentration of the trace gas, as well as the vertical wind speed continuously over a large area.

Knowing how many molecules of a gas are moved away from or towards the surface by the eddies over a given period of time allows scientists to calculate the upward or downward flux (or more properly a flux density which simply refers to a quantity that is moved through a unit area per unit time). The vertical flux of CO2 (Fc), for instance, is obtained by the covariance between the instantaneous departure (w’) from the mean vertical wind speed and the instantaneous departure (C’) from the mean gas concentration as:


(1)


Here ? is air density and the overbar indicates a time average. Measurements are typically averaged over a 30-minute period. This time step allows all turbulent fluctuations (operating on frequencies less than 30 minutes) to be captured while the influence of longer time scale phenomenon (weather patterns) is excluded.

The storage flux (Fs), i.e. the rate of change in storage of CO2 in the column of air below the instrumentation height, can be calculated based on Morgenstern et al (2004) as:


(2)


where hm is the measurement height, ?a is the mean molar density of dry air, and c is the mean CO2 mole mixing ratio. ?Sc is the difference between Sc of the subsequent and previous half-hours. The NEE can then be computed as:


(3)


where Fc is the CO2 eddy flux measured by the EC system. Atmospheric scientists typically use the meteorological sign convention, where negative NEE values represent a net uptake of CO2 by the ecosystem, and positive values indicate a net release to the atmosphere.

Since turbulent fluctuations occur at very high frequency, fast response instruments are required. The two main instruments used in the EC systems at EM-1 are:

• The three-dimensional sonic anemometer. This measures wind speed along three axes: in the mean flow direction, perpendicular to the mean flow, and vertically.

• An open-path infrared gas analyser (IRGA), which measures the concentration of CO2 and water vapour (Figure 1). These concentrations and wind speeds are recorded using fast-response data loggers, commonly at 10Hz (i.e. ten measurements per second).

Measurements at 10Hz generate a tremendous amount of data. Over 2GB are collected during a single month period. The processing of EC data is quite complex and requires a significant amount of programming using specialised software before the flux data can be used in a meaningful way. The two main steps in the EC data processing are the data quality control and gap filling of missing data.

As part of the data quality control, pre-established criteria following internationally agreed protocols help to tell if the data recorded are reliable or not. Data are first screened using the built-in diagnostic signal of the IRGA which indicates periods when the optical path was partially or fully obstructed – often corresponding to rain or snow events. Data are also rejected when the system indicates (usually small) physically improbable uptake of CO2 at night or during wintertime.

Fundamentally, the EC technique operates during turbulent conditions. On calm nights, the atmospheric conditions may be too stable for reliable data collection. These conditions are identified using a threshold value of the friction velocity – a measure of how turbulent the air flow is. Finally, instantaneous extreme values are eliminated based on the monthly averages and standard deviations of the CO2 fluxes. An additional step is performed for the reservoir flux data, which consists of removing the data when the wind is coming from the tower island in order to retain only the fluxes from the water surface itself.

Gaps in time series of EC data can occur due to a variety of reasons (e.g. following quality control, power failure, instrument maintenance and repair, etc.). Data gaps are usually divided into gaps of short and long duration. Short data gaps of less than four half-hours are filled by linear interpolation. For terrestrial ecosystems, longer gaps are filled based on empirical models specific for day and night with respect to daytime and night time dominant physiological processes (knowing that NEE = ER – GEP; with ER = ecosystem respiration and GEP = gross ecosystem productivity).

Respiration is driven by temperature so an exponential relationship of nighttime NEE vs. soil temperature is commonly used to fill gaps during the night and in the winter, i.e. when GEP is assumed to be zero so that ER is the only flux component. This relationship is also used to model daytime ER. During the daytime and during the growing season, photosynthetic uptake is dominant and is largely controlled by light (photosynthetically active radiation or PAR).

To fill daytime NEE gaps, a GEP vs. PAR hyperbolic relationship is used. The remaining NEE gaps are filled by subtracting GEP from ER. The gap filling procedure for the reservoir CO2 fluxes is more complex as the source and processes driving the CO2 emissions are disconnected spatially from the water surface and the transport processes in the atmosphere.

Preliminary work indicates that relationships between CO2 fluxes and water temperature and/or concentration of dissolved organic carbon, and wind speed would help us estimate the reservoir fluxes for any missing periods.

Why use eddy covariance in EM-1?

The creation of the EM-1 hydroelectric reservoir required the flooding of over 600km2 of the boreal ecosystem along the Eastmain river, of which 65% was occupied by forests, 14% by peatlands, and 21% by lakes and rivers. Such a disturbance obviously modifies many ecological, biological and physiological processes, which in turn affect the way GHG are produced and consumed in the flooded area and its surroundings. While CO2 is absorbed by trees and released through respiration and decomposition in forests, and accumulated in peatlands, it is mainly being emitted in aquatic systems. Quantifying the impacts of this landscape modification on the GHG emission patterns is an enormous task, but EC is one of the most promising measurement techniques available to achieve this goal.

Despite its complexity, the EC technique offers many advantages compared to other commonly used methods for measuring GHG fluxes. The main benefit of this method is the fact that it allows for the continuous and spatially-averaged measurement of the vertical exchange of carbon, energy and water between the atmosphere and a surface, with minimal disturbance to the environment.

It also measures the NEE directly. As opposed to chambers that are useful in identifying discrete and small-scale variations in GHG emission patterns, GHG fluxes measured with the EC technique provide precise measurements over a relatively large area, or footprint (ie the area ‘seen’ by the instruments or the area contributing to the measured fluxes) upwind from the tower, which can be hundreds of square metres in area. The exact size of the footprint depends on the measurement height, the roughness of the surface and atmospheric thermal stability.

Tower platforms are used to mount the EC system many metres above the surface of interest both to increase the footprint area and to ensure that measurements are above the mean source/sink height in the case of a forested ecosystem. The EC systems are installed in areas that are largely homogenous and representative of the surface characteristics that need to be measured.

Furthermore, with the continuous and long-term functioning of the EC systems, it is possible to interpret and compare GHG emission patterns on multiple times scales (hourly, seasonally, and annually). Temporal variation in CO2 fluxes is common. Timing of the spring onset of photosynthetic uptake and the fall switch to net release, as well as ice formation and ice melt on the EM-1 reservoir, are key aspects in the study of NEE and could be significant in controlling the inter-annual variability in the forest, peatland, and reservoir net C balances.

The need to measure wintertime NEE is increasingly recognised since CO2 can be emitted through snow, and possibly through ice. Linking EC fluxes with meteorological and hydrological conditions is key in understanding the timing and magnitude of such variability. Although the use of the EC systems has CO2 as its primary focus, measurements of methane are being made on the reservoir and in the peatland using chambers and profiling systems. Previous studies have suggested that site productivity exerts a strong influence on CO2 emissions.

Therefore, used in combination with auxiliary measurements of meteorological and ecosystem (soil, plants) variables, the EC technique helps scientists to understand the gas exchange response to changes in environmental conditions. Additional variables measured at the EM-1 study sites include soil and air temperature, relative humidity, solar radiation, photosynthetically active radiation (PAR), wind speed and direction, precipitation, and snow depth; all of which are being recorded continuously. It is hence made easier to examine the dominant controlling factors on the CO2 fluxes at the different study sites, and thus to get a better understanding of the present and future contribution of hydroelectric reservoirs to the greenhouse effect.

Study sites in EM-1

In order to assess the impacts of the creation of the reservoir on the GHG emissions, the first challenge that the scientific team had to face was to find appropriate site locations that were representative of the pre-flooded (ie forest and peatland) and post-flooded (ie reservoir) environments. A mature black spruce-dominated forest having similar characteristics to the pre-flooded surface was chosen and became one of the two main study sites in which a flux tower was initially set up (Figure 2). The second main study site is located on the edge of an island on the reservoir, allowing the scientific team to acquire trace gas measurements from the reservoir itself, thus representing the post-flooding conditions (Figure 3).

These two main EC systems were installed and were operational by the end of summer 2006. The towers are 23m tall in the forest, and 13m above the mean water surface on the island, with the instruments mounted at the top of each tower. In addition to these, a third EC system was installed in spring 2008 on a portable 2.7m high tower in a peatland (Figure 4) to account for the second dominant terrestrial ecosystem of the pre-flooded environments. The results from these three towers will be used to evaluate the pre-flooding vs. post-flooding CO2 fluxes, and thus the net impact of the EM-1 reservoir creation in terms of CO2 emissions.

Trace gas measurements

One of the project goals is to compare the various techniques available for trace gas measurement from reservoirs. To provide an accurate assessment of the net GHG emissions, it is essential to compare the various techniques that have been used for trace gas measurement from reservoirs in the past. Gas flux measurements obtained in the EC tower footprints from smaller spatial and discrete temporal scales using chambers will be compared with the results obtained from the EC towers, which provide continuous and spatially-averaged measurement of the vertical exchange over a relatively large footprint.

The EC technique is also a useful complement to modelling studies, where the continuous flux measurements provide a good means of developing and evaluating better algorithms for scaling up from ecosystem to regional estimates of GHG fluxes. Therefore, the results obtained in this part of the project will be very useful in modelling efforts for predicting future impacts of inundation on carbon budgets within a changing climate, and following the creation of other hydroelectric reservoirs. Researchers in this project will integrate the EC results within a multidisciplinary modelling approach. The data obtained from EC systems will also be used in combination with the measured methane (CH4) emissions to get a more complete assessment of the net carbon fluxes from the EM-1 reservoir.

Examples of results obtained from EC

As mentioned previously, one of the major advantages of the EC technique is the fact that it allows scientists to investigate the CO2 fluxes on different time scales, from minutes to years. Figures 5 to 7 provide examples of the different observable trends of NEE for the two main sites in EM-1 (forest and reservoir). In examining the measured fluxes over a complete year, it is possible to identify the time of the year when a given ecosystem, like a forest, switches from being a net CO2 source to the atmosphere to a net CO2 sink. The timing of this switch changes each year depending on climatic factors and therefore EC’s ability to measure continuously over long time scales permits such analysis.

In the example provided in Figure 5, the NEE of the EM-1 boreal forest over one complete year is shown. It is clearly seen that the forest is a net source of CO2 throughout the winter until spring when it switches to a net CO2 sink which remains through the summer growing season as the trees are actively accumulating CO2. In the fall, the forest ecosystem switches back to a net CO2 source because of the shorter day length and colder temperatures which decrease photosynthetic rates.

In Figure 6, the reservoir NEE is shown for the same time period. The reservoir is seen to be source of CO2 through time, which is expected since there are no plants to absorb the CO2 from the atmosphere. The diurnal variations of CO2 fluxes are shown in Figure 7 for three different months at the forest site. With such analyses, we can easily distinguish between the daytime net CO2 uptake and night-time net CO2 release.

Future work

Further analyses of the measured CO2 fluxes over time will allow the team to understand more clearly how the reservoir CO2 emissions evolve in the years following impoundment. It will hence be possible to verify if, as expected, the GHG emissions return to levels that are within the range of those measured over natural lakes within a few years following impoundment. Using EC systems in the dominant pre-flooded ecosystems of the EM-1 region, and within the reservoir itself, allows scientists to compute the net change in CO2 emissions resulting from the conversion of a boreal forest and wetland ecosystem to a hydroelectric reservoir.

Marie-Claude Bonneville, Project Manager, and Dr. Ian B. Strachan, Associate Professor, Department of Natural Resource Sciences, McGill University, 21 111 Lakeshore Rd., Ste. Anne de Bellevue, H9X 3V9, QC, Canada. Email: marie-claude.bonneville@mcgill.ca and ian.strachan@mcgill.ca



Eastmain-1 Reservoir Net Greenhouse Gas Emissions Project

The Eastmain-1 hydroelectric reservoir was created at the end of 2005 in Northern James Bay, Quebec, Canada. By May 2006, the reservoir had reached its maximal impoundment. The creation of this reservoir provided a great opportunity to study the impacts of reservoir creation on the environment, in particular on greenhouse gas emissions which are still largely unknown for this type of land conversion. As such, the Reservoir Net Greenhouse Gas Emission Research Project was put forward by Hydro-Quebec; this became the first research of its kind to be conducted in a northern hydroelectric reservoir.
While previous studies have measured the gross emissions from hydroelectric reservoirs, this study treats the creation of the reservoir as a land-use change problem. Correspondingly, the main objective of this unique project is to determine the net impact of the EM-1 reservoir creation on the emissions of GHG. In other words, scientists of various disciplines (aquatic, terrestrial, isotopes, modelling) use different techniques to measure GHG emissions before and after impoundment (2005-2009), the goal being to quantify and compare the reservoir carbon stocks and GHG exchanges with those of natural ecosystems.
Co-funded by Hydro-Quebec Production and the Canadian Foundation for Climate and Atmospheric Sciences (CFCAS), the measurement and modelling results from this large-scale project will have potential applications to other hydroelectric reservoirs and be useful for the study of future GHG emissions.



Figure 6 Figure 6
Figure 2 Figure 2
(3) (3)
Figure 5 Figure 5
Figure 1 Figure 1
(2) (2)
Figure 3 Figure 3
(1) (1)
Figure 7 Figure 7
Figure 4 Figure 4


Privacy Policy
We have updated our privacy policy. In the latest update it explains what cookies are and how we use them on our site. To learn more about cookies and their benefits, please view our privacy policy. Please be aware that parts of this site will not function correctly if you disable cookies. By continuing to use this site, you consent to our use of cookies in accordance with our privacy policy unless you have disabled them.