Application of an open-path eddy covariance methane flux measurement system to a larch forest in eastern Siberia
Introduction
Monitoring of methane (CH4) flux dynamics at various terrestrial ecosystems is becoming important, as CH4 is the third most important greenhouse gas after water (H2O) and carbon dioxide (CO2). However, the role of forests in CH4 cycling has not yet been well understood, as the monitoring of CH4 fluxes at the canopy- or ecosystem-scale is still difficult (Shoemaker et al., 2014). Though previous chamber measurements at the forest floor show that many upland forest soils are a net CH4 sink (Megonigal and Guenther, 2008), the canopy- or ecosystem-scale CH4 flux over forests can be positive (i.e., net CH4 source), due to the presence of CH4 emissions from “hotspots” such as small wet areas within the micrometeorological flux footprint (e.g., Ueyama et al., 2018). Focusing on boreal forests, canopy-scale CH4 flux measurements have been attempted in Saskatchewan, Canada by a gradient method (Simpson et al., 1997), in Interior Alaska by a closed-path eddy covariance method (Iwata et al., 2015), and in central Sweden by Bowen ratio and combined eddy covariance and gradient methods (Sundqvist et al., 2015). All these studies have shown CH4 emission (positive flux) over forests at the canopy-scale, while Simpson et al. (1997) and Sundqvist et al. (2015) reported CH4 uptake (negative flux) at the forest floor by chamber measurements. Besides, even though approximately 44% of all boreal forests are located east of the Urals in northern Russia (Jarvis et al., 2001), canopy-scale CH4 flux over forests in such regions (e.g., eastern Siberia) has not been reported. Therefore, knowledge regarding canopy-scale CH4 flux over boreal forests in eastern Siberia can contribute to the further understanding of global CH4 cycling.
In the last decade, an open-path CH4 analyzer LI-7700 (LI-COR Biogeosciences, Lincoln, NE, USA) has been developed (LI-COR Inc., 2011, McDermitt, Burba, Xu, Anderson, Komissarov, Riensche, Schedlbauer, Starr, Zona, Oechel, Oberbauer, Hastings, 2011) which enables us to measure the variation in CH4 concentration, without the need for a power-consuming vacuum pump. As emphasized by McDermitt et al. (2011), the low-power characteristics of the LI-7700 will be a strong advantage when CH4 flux measurements are explored by solar-powered operation in remote areas without access to electricity, including in a vast portion of boreal forests. On the other hand, the open-path CH4 analyzer inevitably presents two errors: one is induced by the fluctuation of air density due to variations in temperature and humidity (Webb et al., 1980); the other is the spectroscopic effects induced by line broadening due to pressure, humidity, and temperature of the air (McDermitt et al., 2011). Though the methods for correcting these errors are theoretically grounded and robust, the extent of these corrections can be significantly large relative to the density-corrected (or ‘true’) CH4 flux, which sometimes changes the sign of fluxes from negative measurement (or raw, uncorrected) values to positive corrected ones, especially in the case of ecosystems where the ‘true’ CH4 flux is very small. For example, Chamberlain et al. (2017) evaluated the effect of these corrections on CH4 flux over sites with different flux strengths, and showed that corrections accounted for a greater than 100% increase in daily CH4 flux at “negligible-flux” sites (alfalfa and pasture). Chamberlain et al. (2017) also observed positive CH4 flux after correction in the daytime over a pavement airfield (zero-flux site), which was interpreted as an overcorrection. Since the absolute value of canopy-scale CH4 flux over forests is reported to be generally small compared to wetlands, we must pay especially close attention to instrument uncertainties when the LI-7700 open-path CH4 analyzer is applied to forest CH4 flux measurements.
When measuring eddy covariance fluxes, sampling (random) errors due to instrumental noise and natural variability (i.e., stochastic nature of turbulence) are unavoidable, resulting in uncertainties in the obtained data and finally calculated fluxes. Several methods for assessing such uncertainties have been proposed (e.g. Finkelstein, Sims, 2001, Langford, Acton, Ammann, Valach, Nemitz, 2015, Wienhold, Welling, Harris, 1995) and used for evaluation of random errors in fluxes (e.g. Iwata, Harazono, Ueyama, Sakabe, Nagano, Kosugi, Takahashi, Kim, 2015., Mauder, Cuntz, Druee, Graf, Rebmann, Schmid, Schmidt, Steinbrecher, 2013), and incorporated in a real-time flux monitoring and surveillance system (Kim et al., 2015). Random flux error due to instrumental noise has also been investigated (e.g. Billesbach, 2011, Mauder, Cuntz, Druee, Graf, Rebmann, Schmid, Schmidt, Steinbrecher, 2013); Rannik et al. (2016) reviewed these studies and suggested the method by Finkelstein and Sims (2001) is robust and accurate for estimating random flux error.
In the case of small CH4 flux measurements over forests, the signal-to noise ratio (SNR) of the open-path CH4 analyzer can be low, which is similar to cases of flux measurements of trace gases and volatile organic compounds (Langford et al., 2015). Langford et al. (2015) focused on eddy covariance data with low SNR, and showed procedures for quantifying random white noise and SNR for the measured data, based on the practical method for determining the components of variance attributed to a genuine (structured) signal and unstructured white noise using the auto-covariance function (Lenschow et al., 2000). They also assessed the random error and limit to the detection of fluxes using the root-mean-square error of the cross-covariance function within a time window defined well away from the point of zero time lag, a modification of the method used by Wienhold et al. (1995) and Spirig et al. (2005). Using these procedures, Langford et al. (2015) clearly showed large random errors and a large fraction of errors associated with instrument noise in the case of isoprene and acetone fluxes, and assessed the quality ensemble mean of diurnal variation in acetone and benzene fluxes, comparing different methods for determining time lag between vertical wind velocity and measured scalar concentrations. The procedure of Langford et al. (2015) is considered useful for the assessment of the applicability of the LI-7700 open-path CH4 analyzer to small CH4 flux sites.
The objectives of this study are to evaluate the applicability of the open-path CH4 flux measurement system to a larch forest in eastern Siberia using random flux error analysis according to Langford et al. (2015), and to analyze the characteristics of the CH4 flux measured in this forest.
Section snippets
Study site and instruments
Eddy covariance flux measurements were conducted in a larch forest in Spasskaya Pad Scientific Forest Station (62∘15′18″N, 129∘14′29″E, 220 m a.s.l.), which is located on the left bank of the Lena River, about 20 km northwest of Yakutsk, in the Republic of Sakha, Russia (Fig. 1), where long-term flux observation has been conducted since 1998 (e.g. Kotani, Saito, Kononov, Petrov, Maximov, Iijima, Ohta, 2019, Ohta, Hiyama, Tanaka, Kuwada, Maximov, Ohata, Fukushima, 2001, Ohta, Kotani, Iijima,
Auto-covariance and cross-covariance functions
Given the time series of variable x, the auto-covariance function fx,x(t) is defined as follows.where t is the number of data points associated with the time lag, and N is the number of data points in the time series. Note that where σx is the standard deviation of x. If the time series of x is affected by white noise, the contribution of this noise to fx,x(t) appears only at (Langford, Acton, Ammann, Valach, Nemitz, 2015, Lenschow,
Results
During the study period, the daytime latent heat flux λE gradually increased from about 50 W m to 150 W m where the energy balance closure was 79.9% (Fig. 4a). This might be due to the extremely dry atmospheric conditions with its water vapor pressure deficit (VPD) up to about 30 hPa (Fig. 4b). On the other hand, there were only three precipitation events, totaling 8 mm. These conditions resulted in a gradual decrease in volumetric soil water content at 10-cm depth (Fig. 4d). Air
Validity of positive daytime methane flux after corrections
The daytime positive Fm in this study was a consequence of the cumulative effects of high frequency, WPL, and SS corrections, whereas the uncorrected daytime Fm showed strong negative values. Since whether the daytime Fm is positive or negative is dependent on the magnitude of the cumulative effects of corrections, the sign of outcomes of Fm after corrections should be carefully validated. Such changes in the sign of Fm after corrections using the LI-7700 open-path CH4 analyzer have been
Conclusions
A canopy-scale CH4 flux over a larch forest in eastern Siberia in 2016 measured by an open-path eddy covariance flux system was shown here, together with random flux error and limit of flux detection (LoD). CH4 flux after corrections showed clear diurnal variation, showing emission in the daytime and near-zero in the nighttime, irrespective of wind direction. Though the calculated methane flux was nearly the same as the upper boundary of the LoD at the 95th percentile, this flux was considered
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
We thank Nate Bauer of International Arctic Research Center, University of Alaska Fairbanks for English proofreading of the manuscript. We are also grateful to anonymous reviewers for valuable comments. This research was supported by the Arctic Challenge for Sustainability (ArCS) Project of the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan. Methane flux data was provided by North-Eastern Federal University (NEFU), Russia, under the agreement between NEFU and
References (54)
Estimating uncertainties in individual eddy covariance flux measurements: a comparison of methods and a proposed new method
Agr. Forest Meteorol.
(2011)- et al.
Landscape-level terrestrial methane flux observed from a very tall tower
Agric. For. Meteorol.
(2015) - et al.
Comparing laser-based open- and closed-path gas analyzers to measure methane fluxes using the eddy covariance method
Agr. Forest Meteorol.
(2011) - et al.
Tools for quality assessment of surface-based flux measurements
Agric. For. Meteorol.
(1996) - et al.
Effects of water vapor dilution on trace gas flux, and practical correction methods
J. Agric. Meteorol.
(2015) - et al.
Productivity of Boreal Forests
- et al.
Fluxpro as a realtime monitoring and surveilling system for eddy covariance flux measurement
J. Agric. Meteorol.
(2015) - et al.
Eddy-covariance data with low signal-to-noise ratio: time-lag determination, uncertainties and limit of detection
Atmos. Meas. Tech.
(2015) - et al.
A new low-power, open-path instrument for measuring methane flux by eddycovariance
Appl. Phys. B
(2011) - et al.
Correction of sonic anemometer angle of attack errors
Agric. For. Meteorol.
(2006)
Ultrasonic anemometer angle of attack errors under turbulent conditions
Agric. For. Meteorol.
Parameterisation of aerodynamic roughness over boreal, cool- and warm-temperate forests
Agric. For. Meteorol.
A comparison between various definitions of forest stand height and aerodynamic canopy height
Agric. For. Meteorol.
Seasonal variation in the energy and water exchanges above and below a larch forest in eastern Siberia
Hydrol. Process.
Interannual variation of water balance and summer evapotranspiration in an eastern Siberian larch forest over a 7-year period (1998–2006)
Agric. For. Meteorol.
Field intercomparison of four methane gas analyzers suitable for eddy covariance flux measurements
Biogeosci.
Quality control and flux sampling problems for tower and aircraft data
J. Atmos. Oceanic Tech.
Accounting for spectroscopic effects in laser-based open-path eddy covariance flux measurements
Glob. Change Biol.
Calculating CO2 and H2O eddy covariance fluxes from an enclosed gas analyzer using an instantaneous mixing ratio
Glob. Change Biol.
Evaluation of density corrections to methane fluxes measured by open-path eddy covariance over contrasting landscapes
Bound. Layer Meteorol.
Methane production and emissions in trees and forests
New Phytol.
Simplified expressions for adjusting higher-order turbulent statistics obtained from open path gas analyzers
Bound. Layer Meteorol.
Sampling error in eddy correlation flux measurements
J. Geophys. Res.
Applying a linear detrend to eddy correlation data in real time
Bound. Layer Meteorol.
A twenty-first century molar mass for dry air
HVAC&R Res.
Software bug affecting ‘w’ wind component of the WindMaster family
Technical key note series number kn11509v6
Summer soil CH4 emission and uptake in taiga forest near Yakutsk, Eastern Siberia
Agric. For. Meteorol.
Cited by (9)
Impact of filtering methods on ultrafine particles turbulent fluxes by eddy covariance
2022, Atmospheric EnvironmentLong-term dynamics of soil, tree stem and ecosystem methane fluxes in a riparian forest
2022, Science of the Total EnvironmentCitation Excerpt :Ecosystem-level studies using the eddy covariance (EC) technique suggest that tree canopies contribute to the ecosystem CH4 budget (do Carmo et al., 2006). Eddy covariance measurements of CH4 fluxes have been conducted in boreal forests (Iwata et al., 2015; Nakai et al., 2020), temperate forests (Wang et al., 2013; Hommeltenberg et al., 2014; Shoemaker et al., 2014; Flanagan et al., 2020), and tropical/sub-tropical forests (do Carmo et al., 2006; Dalmagro et al., 2019; Griffis et al., 2020; Liu et al., 2020). In addition, data are available from FLUXNET network sites in Finland, Switzerland, Russia and Indonesia.
Constraining models for methane oxidation based on long-term continuous chamber measurements in a temperate forest soil
2021, Agricultural and Forest MeteorologyCitation Excerpt :CH4 fluxes in upland soils decrease with increasing water-filled porosity, which determines the gas diffusivity, and has a low temperature sensitivity (Del Grosso et al., 2000; Hashimoto et al., 2011b; Smith et al., 2000). Canopy-scale tower flux measurements showed that upland forests are a net CH4 source with microspots for CH4 emissions, such as banks of streams and small ponds having a strong emission potential, within a forest (Sakabe et al., 2012; Sundqvist et al., 2015a; Ueyama et al., 2018; Nakai et al., 2020). Compared with studies on spatial variations, long-term observations are limited for CH4 fluxes in upland soils, except for a few long-term studies (Wu et al., 2011).
Lessons learned from more than a decade of greenhouse gas flux measurements at boreal forests in eastern Siberia and interior Alaska
2021, Polar ScienceCitation Excerpt :The clear diurnal variation of the CH4 flux, emission in the daytime and near-zero in the nighttime, was mostly explained by the atmospheric stability. The range of the daily mean CH4 flux from May 29 to June 12 was in net emissions of 4.9–13.8 nmol m−2 s−1 (Fig. 6; see also Figur 16 in Nakai et al., 2020). This was within the range of the CH4 emissions measured by static chambers on the forest floor and at a mesotrophic fen near the site, reported by a previous study (van Huissteden et al., 2008).
Persistent net release of carbon dioxide and methane from an Alaskan lowland boreal peatland complex
2024, Global Change BiologyMethane dynamics from a mixed plantation of north China: Observation using closed-path eddy covariance method
2023, Frontiers in Plant Science