CD-10 Abstract

Net Ecosystem Exchange of CO2 and H2O from Primary Tropical Forest in Central Amazônia

Plinio B. de Camargo, CENA/USP (SA-PI)
Volker W. J. H. Kirchhoff, INPE (SA-PI)
Antonio Donato Nobre, INPA (SA-PI)
Steven C. Wofsy, Harvard University (US-PI)


We have been making measurements of net ecosystem exchange (NEE) and concentrations of CO2 and H2O using eddy correlation methods, intended to continue for 3-5 years at a primary forest site located at km 67 on the Santarém-Cuiaba highway in the Tapajós National Forest in central Amazônia. Complementing these measurements is a comprehensive set of ecological observations, including tree speciation, stocks and growth, coarse woody debris stocks and dynamics, etc. We are also making measurements of CO2 and other greenhouse gases at a remote coastal site near Natal. 

We are collaborating with Dr. V. Kirchhoff to help install and operate a long-term coastal site for observations of important trace gases, starting with CO2. We are collaborating with Plinio B. de Carmargo on the ecological measurements. Our work with Antonio Nobre involves advanced analysis and synthesis of eddy flux data from his Manaus site and comparison with the Santarém site.

The objectives of the measurements are: 

1. Define the experimental protocols needed to measure accurately seasonal and annual net
fluxes of CO2, H2O, and energy at a primary forest in the Tapajós region of Amazônia;

2. Define the net source or sink of CO2 from the undisturbed forest; 

3. Determine the variations of net exchange of CO2 seasonally and inter-annually, and define
the response of carbon sequestration in the system to climatic and other environmental

4. Provide the experimental control for interpretation of the results obtained at a harvested site at km 83 by recording the CO2 exchange at a nearby undisturbed site; 

5. Provide the flux and gradient measurements for CO2, sensible heat and momentum needed to define the flux of N2O, CH4, and mineral elements from sub-canopy concentration changes or from above-canopy gradient measurements of these species; 

6. Determine CO2 boundary layer concentrations at mid-continent and coastal sites to test

Value to LBA

The proposed work focuses on defining the "undisturbed state" of the primary tropical forest, a requirement "to fill the most critical of gaps and/or to gain the most leverage on major scientific uncertainties" (NRA-97-MTPE-02). Studies of this type were specifically requested for LBA: ("Continuous observations of a core set of measurements (e.g., CO2 fluxes, trace gas fluxes, trace gas concentrations, micrometeorological conditions, radiation, aerosols) made at the primary field sites over a period of 3-5 years". 


Fluxes of momentum, CO2, H2O, sensible heat, net radiation, and PAR, atmospheric and soil profiles of temperature, CO2 and H2O, and wind profiles are being measured continuously using automated instruments on a 60meter tower. Fluxes will be measured using eddy correlation, with an infrared gas analyzer (closed-path) for CO2 and H2O mounted on the tower near the sonic anemometer. The data are being analyzed to determine net exchange of CO2 and H2O on hourly, daily, seasonal, and annual time scales. Environmental regulation of variations in net uptake are being quantitatively elucidated by measuring the response of the system to climatic variations and by comparison between flux measurements and ecological observations. 


The measurements are being made from a 65 m tall, small-cross-section tower of the type used to support radio antennas (Rohn 55G, Peoria IL), selected to minimize wind distortion and possible heating artifacts, placing the sensor well above the tallest emergent trees. The data acquisition system and most instruments are housed in a climate controlled hut 15-30 meters west of the tower base, accessible by a dirt road. The CO2-H2O sensor will be placed close to the sonic anemometer near the top of the tower to keep tubing short.

The site is in a protected primary forest reserve at km 67 south of Santarém. This site is extremely flat, an extensive planalto that drops about 30m to the level of the Tapajos river 10-15 km to the west. Soils are uniform yellow oxisols similar to soils at the site recently harvested. The nearest secondary road is 5 km to the east and the nearest urban area is
60 km to the North (Santarém); otherwise the area is quite isolated. 

Tower Measurements

Table 1 shows the measurements at the site in this and in collaborative studies. Eddy fluxes of sensible heat, latent heat, CO2, and momentum are measured at 60 m and independently at 47 m on the tower. The mixing ratios of CO2 and H2O are monitored by sampling 6 to 8 standard liter/min (slpm) from an inlet located directly behind the vertical axis of the anemometer into a CO2/H2O infrared gas analyzer (IRGA; Model 6262, LiCor, Lincoln NE). Errors due to separation of the inlet from the anemometer should be small (1 to 2%, Lee and Black 1994). Wind and temperature will be measured with a 3-axis sonic anemometer pointed into the dominant wind direction (east) 

Table 1. Tapajos Forest Measurements


inlet/instrument alt(s)

determined quantity (data rate)

Sonic Anemometers

60m and 47m (3-axis),

u,v,w,T Fmom, Fheat (10 Hz, 30-min avg.)

high-speed CO2-H2O (IR-Absorbance)

60m, 47m

CO2 and H2O fluxes (30-min avg.)

slow CO2-H2O ( IR Absorbance)

8-10 levels

CO2, H2O vertical profiles (2 /hour)

Thermistors, thin-film capacitors

5-6 levels

T and R.H. profiles (5,60-min avg.)


2 cm (6 rep), 20cm, 50cm

soil temperatures (5,60-min avg.)

Photosynthetically Active Radiation

60,24,2 m

(5 min avg.)

Net radiometer


radiant heat flux (5,60-min Avg.)

Relationship with Other Experiments

The experiment is closely coordinated with three other LBA projects, in addition to the Antonio Nobre measurements at Manaus noted above: 

M. Keller, P. Crill: Our flux measurements will help extend the spatial (canopy) and temporal (years) scales, and provide a quantitative framework, for integrating biogeochemical measurements to the ecosystem function. 

Crill, Silver,and Li and Goulden plan to work at a flux tower on a nearby primary forest that will be commercially harvested during the observations. Our measurements at an uncut site nearby will provide the control for this experiment, determining changes in carbon storage and ecological parameters in the absence of disturbance. 

We plan to combine our data with observations of canopy/atmosphere interchange and energy balance by Fitzjarrald, Moore, to help define the tower footprint and to wring out systematic errors from the observations. 

Collaboration is ongoing also with Martens (222Rn) and Trumbore and Crill. (13C/12C) isotopic ratios and turnover rates of soil organic matter and wood.