On the importance of high-frequency air-temperature fluctuations for spectroscopic

corrections of open-path carbon dioxide flux measurements

OverviewThe IRGASON flux system consists of an open-path

infrared H2O and CO2 gas analyzer and a colocated 3D

sonic anemometer. The IRGASON measures CO2 mixing

ratio in situ by using the fast-response humidity corrected

sonic temperature measurement to account for the

temperature related density fluctuations. CO2 flux can be

computed directly using instantaneous CO2 mixing ratio,

eliminating the need for density (WPL) corrections in post

processing.

Instrumentation1. IRGASON – Integrated CO2 & H2O open-path flux system

with colocated sonic and gas sensing volumes2. EC155 – CO2 & H2O closed-path analyzer 3. CSAT3A – Stand alone sonic anemometer/thermometer4. Temperature probe in a five plates radiation shield5. Quad-Disk Pressure Probe (Nishiyama et al., 1989)Site DetailsLocation: +41° 46' 3.89" N, –111°51' 21.19" EMeasurement height: 1.65 m – IRGASON, 2.00 m – CSAT3Sampling rate: 20 Hz

LiteratureGrelle, A., Burba, G. (2007). Fine-wire thermometer to correct CO2 fluxes by open-path analyzers for artificial density fluctuations. Ag. For. Meteorol.,

147:48-57.Webb E., G. Pearman, and R. Leuning (1980). Correction of flux measurements for density effects due to heat and water vapour transfer. Quart. J. Roy.

Meteorol. Soc., 106: 85-100.Jamieson, John A., R. McFee, G. Plass, R. Grube, R. Richards. (1963). Infrared Physics and Engineering. McGraw Hill Book Company, Inc.Griffith, D. (1982). Calculations of carrier gas effects in non-dispersive infrared analyzers I. Theory. Tellus, 34, 376-384.

Acknowledgements: The first author thanks Campbell Scientific for the support of this study and Steve Sargent for the discussions and the review of the poster.

MotivationA growing number of studies report systematic differences

in CO2 flux estimates obtained with the two main types of

gas analyzers, open and closed path. Compared to eddy-

covariance systems based on closed-path (CP) gas

analyzers, systems with open-path (OP) gas analyzers

consistently overestimate CO2 uptake during daytime

periods with high positive sensible heat fluxes.

Ivan Bogoev(1)*, Manuel Helbig(2), Oliver Sonnentag(2)

(1) Research and Development, Campbell Scientific, Inc., Logan, Utah, USA; *ivan@campbellsci.com (2) Département de géographie, Université de Montréal, Montréal, Québec, Canada

Field experimentWe conducted a field inter-comparison with IRGASON

(open-path) and EC155 (closed-path) systems operating

side-by-side and using a common sonic anemometer.

Hypothesis: Inadequate correction of fast temperature related spectroscopic effectsThe strength and the half-width of the CO2 absorption lines are affected by temperature. Single-path, dual-wavelength,

non-dispersive infrared gas analyzers use air-temperature measurements to correct for these spectroscopic effects. The

probes used for the compensation have limited frequency response due to their thermal mass and are inadequate to

compensate CO2 absorption measurements for fast temperature variations.

1

4

3

2

5

Findings1. During periods of high sensible heat flux, the IRGASON measures more negative CO2 flux. Under negative heat flux

regimes, the IRGASON measures more positive CO2 flux (Figure 1).

2. Hourly mean CO2 concentrations measured with the EC155 and the IRGASON agree to within 3% (Figure 2).

3. During daytime under high sensible heat flux conditions, CO2 spectra measured by the two analyzers agree well in the 0

to 0.005 Hz range, but the IRGASON measures increased CO2 variations above 0.01 Hz (Figure 3).

4. Nighttime CO2 spectra agree well in the frequency range 0 to 0.1 Hz. The EC155 measured increased CO2 variations

above 0.1 Hz (Figure 4).

5. Compared to the EC155, under high sensible heat flux conditions, the co-spectral response of the IRGASON is

increased (Figure 5).

6. During nighttime the agreement of the co-spectral responses extends to higher frequencies up to 0.1 Hz (Figure 6).

-0.6 -0.4 -0.2 0 0.2 0.4

-0.6

-0.4

-0.2

0

0.2

0.4

Hourly CO2 Flux

EC155 Fc [mg m-2

s-1

]

IRG

AS

ON

Fc [

mg m

-2 s

-1]

1:1

Hs[W]

-50

0

50

100

150

200

Figure 1

400 420 440 460 480 500 520

400

420

440

460

480

500

Hourly Mean CO2 Concentration

EC155 CO2 [mol mol

-1]

IRG

AS

ON

CO

2 [m

ol m

ol-1

] slope: 1.0297 offset: -11.6571

R2 = 0.99404

1:1

-10

0

10

20

30

T [C]

Figure 2

10-2

100

10-2

100

102

Power Spectrum of CO2

Frequency [Hz]

S C

O2

[p

pm

2]

EC155

IRGASON

-5/3

Daytime

Figure 310

-210

010

-2

100

102

104

Power Spectrum of CO2

Frequency [Hz]

S C

O2 [

ppm

2]

EC155

IRGASON

-5/3

Nighttime

Figure 4

10-2

100

10-2

100

Cospectrum of CO2 and w

Frequency [Hz]

Co C

O2 [

ppm

m s

-1]

IRGASON

EC155

-7/3

Daytime

Figure 510

-210

010

-4

10-2

100

Cospectrum of CO2 and w

Frequency [Hz]

Co C

O2 [

ppm

m s

-1]

IRGASON

EC155

-7/3

Nighttime

Figure 6

𝐴 =𝑁

Δν 𝜈1

𝜈2

1 − exp −𝑆𝑖α𝑖cL

𝜋 𝜈 − 𝜈02 + 𝛼𝑖

2𝑑𝜈

A = Absorbed light energy in the spectral interval

N = Number of absorption lines in the spectral interval

ν = Wave number of the individual spectral line

c = Density of the absorbing gas

L = Path length

Si = Strength of the individual line

αi = Half-width of the individual line

Si = f(T,P) αi = f(T,P)

T = Air temperature, P = barometric pressure

-50 0 50 100 150 200 250

-5

0

5

FcOP

- FcCP

Hs [W m-2

]

Fc

OP -

Fc

CP [ m

ol m

-2s

-1]

=-0.014257*Hs+0.066828

=-0.0035314*Hs+0.06371

Slow Temp

fit

Fast Temp

fit

-0.6 -0.4 -0.2 0 0.2 0.4-0.6

-0.4

-0.2

0

0.2

0.4

Hourly CO2 Flux

EC155 Fc [mg m-2

s-1

]

IRG

AS

ON

Fc [

mg m

-2 s

-1]

slope = 1.086

R2 = 0.9371

slope = 1.026

R2 = 0.9771

Slow Temp

Fast Temp

ConclusionsUsing fast air-temperature measurements to correct for the absorption line broadening effects of open-path CO2 flux,

reduces the differences with fluxes measured by a closed-path analyzer.

Slow-response air temperature measurement

Fast-response air temperature measurement

Results of using fast-response air temperature in the conversion of absorption into CO2 densityThe difference between OP and CP flux measurements is reduced and less dependent on sensible heat flux.

2280 2300 2320 2340 2360 2380 24000

0.2

0.4

0.6

0.8

1

CO2 Absorption

Wave Number

Absorp

tion

T: 240 K

T: 330 K

Figure 7

Figure 8 Figure 9