Close Window

TG-04 Abstract

Radon-222 Tracing of Carbon Exchange and Trace Gas Fluxes in Old Growth and Selectively-Logged Amazônian Forests

Christopher S. Martens — University of North Carolina at Chapel Hill (US-PI)
Osvaldo Luiz Leal de Moraes — UFSM - Universidade Federal de Santa Maria (SA-PI)

Objectives





We will study the rates and mechanisms of processes controlling carbon exchange and

trace gas fluxes at both old growth and selectively-logged forest sites in the Tapajós

National Forest and surrounding areas south of Santarém, Pará, Brazil. The work will

directly address two theme areas in LBA-ECO: Carbon Storage and Exchange, and Trace

Gas Fluxes. The proposed work can be divided into two primary objectives:











  • Quantification of air exchange rates and gas fluxes between old growth and selectively

    logged forests and overlying atmosphere utilizing continuous, in situ radon-222

    measurements


  • Comparison of radon-derived exchange rate constants and fluxes with values derived from

    eddy correlation and flux chamber measurements by collaborating

    investigators.










Radon-222 measurements in forest canopy air, the atmosphere immediately overlying the

canopy, the soil atmosphere and in flux chamber samples will be performed in direct

collaboration with long-term tower flux and microclimate observations as well as automatic

flux chamber studies of trace gas exchanges by other LBA investigators. The most critical

assumption in the use of long-term eddy covariance to determine ecosystem CO2

balance is that the measurements are not biased from day to night (Goulden et al. 1996).

Because net carbon uptake reflects the difference between two larger fluxes, respiratory

efflux at night and photosynthetic uptake during the day, a small selective

underestimation of flux at night can cause a large overestimation of long-term

accumulation. Our radon work will be focused on and fully integrated with related studies

at two flux tower sites along the east side of the LBA transect proposed to be established

by a team of investigators. One of the towers will be located in undisturbed, old growth

forest (Wofsy and others) and the other in an area where selective logging will be carried

out during 1998 and 1999 (Goulden and others). The radon measurements, conducted in

collaboration with tower studies of heat, momentum, CO2, H2O and

other gas transport by Wofsy et al., Goulden, Fitzjarrald and Moore, and others, and flux

chamber studies by Keller, Crill, Trumbore and others, will provide direct quantification

of the physical exchange of CH4, CO2, N2O, and other

trace gases between soils, the forest canopy, and the overlying atmosphere in undisturbed

versus logged areas. The radon measurements will employ multiple flow-through detectors

whose use in continuous tower flux studies has been pioneered by our group. The

multiple-detector set has been successfully deployed in tower studies at remote locations

in collaboration with other investigators for continuous measurements every 30 minutes for

periods of up to 4.5 years and at altitudes above the ground ranging from less than 10 cm

to 496 m.





The following scientific questions will be addressed through collaborative efforts with

other LBA investigators: What are the rates of gas exchange between forest canopy air and

the atmosphere and how do these rates vary on daily and longer term temporal scales? How

does vertical gas mixing vary with elevation within the forest canopy? How do episodic

events (e.g. gust fronts) affect exchange between the sub-canopy layer and the free

troposphere? How does thinning of forest canopy caused by selective logging change the gas

exchange rate with the atmosphere and the relative importance of various canopy air

ventilation mechanisms? Does canopy structure affect the rate of episodic flushing of the

sub-canopy atmosphere?





Canopy/Atmosphere Gas Exchange Processes





Recent eddy correlation studies of carbon exchange over tropical forests suggest that

mature tropical evergreen forests may indeed be a significant carbon sink (Grace et al.

1995. The most critical assumption in the use of long-term eddy covariance to determine

ecosystem CO2 balance is that the measurements are not biased from day to night

(Goulden et al. 1996). Because net carbon uptake reflects the difference between two

larger fluxes, respiratory efflux at night and photosynthetic uptake during the day, a

small selective underestimation of flux at night can cause a large overestimation of

long-term accumulation. Grace et al. (1995) reported an annual uptake of 1 tC ha-1 that

reflected the difference between ~10 tC ha-1 uptake during daytime and ~9 tC ha-1 loss at

night. It is troubling that Grace et al's (1995) site was a carbon sink on days with calm

nights but not on days with windy nights. Perhaps there is a systematic underestimation of

C loss during stable periods particularly at night. This is a methodological concern for

all tower-based carbon measurements.





Because most trace gases emitted by the biosphere are either photo-chemically or

biologically reactive, they are potentially unsuitable direct tracers for trace gas

transport within, out of or into vegetated land surfaces. However, radon-222, a

radioactive natural gas, is ideally suited for studies of gas exchange in the tropical

forests of Amazônia for the following reasons:









1. it is emitted almost exclusively from the soil



2. it is a chemically inert gas, making it suitable for tracing physical exchange

between forests soils, canopies and the overlying atmosphere



3. the only sink for radon is radioactive decay which can be easily quantified using

its known decay constant



4. its 3.8 day halflife yields nearly conservative behavior in studies of

soil/atmosphere and canopy/atmosphere gas exchange



5. existing technology now allows for continuous, multi-altitude radon activity

measurements both within and above the forest canopy.





Trumbore et al. (1990; see equation 1) showed that radon-derived gas exchange rates

compared favorably with estimates for CO2 obtained by eddy correlation

techniques (Fan et al., 1990; Fitzjarrald et al., 1990). Radon-222 was shown to have the

potential to provide an independent and reliable measure of gas exchange rates between

soils, forest canopy and the overlying atmosphere.



Our group pioneered the deployment of continuous tower-based atmospheric radon-222

activity measurements during the 1988 ABLE 3A mission near Bethel, Alaska (Martens et al.,

1992; Martens, unpublished data) and began continuous, multi-altitude measurements every

30 minutes at a micrometeorological tower site as part of ABLE 3B in the open-canopy,

boreal forests of northern Québec, during the summer of 1990 (Ussler et al., 1994). The

rate of gas exchange can be computed using a simple inventory model:





where h is the canopy height, C is the spatial mean trace gas concentration within the

canopy, S is the soil flux, k is a canopy gas exchange coefficient, Ct is the

concentration of the trace gas in the overlying atmosphere, and P and L are production and

loss, respectively of trace gas within the canopy. In practice the term on the left is

evaluated by comparing concentration profiles at successive time points.

Evapotranspiration is not an important source of radon within the Amazônian or boreal

forest canopy, and radioactive decay loss of radon is insignificant as well, so the

integral term on the right side of equation 1 is negligible and can be eliminated.













Last Updated: May 18, 1998

Close Window