Opendata, web and dolomites


Teaser, summary, work performed and final results

Periodic Reporting for period 2 - ASICA (New constraints on the Amazonian carbon balance from airborne observations of the stable isotopes of CO2)


The ASICA project aims to better understand the carbon cycle of the Amazon rain forest. Thereto, air samples are gathered from light aircraft flying over the rain forest. These air samples are subsequently transported to a laboratory in Sao Jose de Campos, run by co-PI Luciana...


The ASICA project aims to better understand the carbon cycle of the Amazon rain forest. Thereto, air samples are gathered from light aircraft flying over the rain forest. These air samples are subsequently transported to a laboratory in Sao Jose de Campos, run by co-PI Luciana Gatti and her team. In this lab we can measure the concentrations of CO2, CH₄, N2O, CO, and H2 in the air that was collected. Through ASICA, we can now also measure the stable isotopes of CO2 on these samples: δ¹³C, δ18O, and Δ17O. To make these very difficult measurements possible was one of the main challenges for ASICA. Once successful, the program aims to build a long-term record of these compounds from a network of sites, and to analyze them together.

The analyses of ASICA focus on two scientific questions:

(1) What is the Gross Primary Productivity (GPP) of the Amazon rain forest?
(2) How does it change during extreme droughts?

The first question is addressed using the Δ17O of CO2, which is an isotope ratio that can be used as a proxy for GPP. Because plants take up CO2 very efficiently, they draw in large amounts of CO2 during sunlit hours, and this process very effectively changes the Δ17O of CO2 to a degree that we can measure. Gradients of Δ17O in CO2 over space (surface to middle troposphere) and time (wet season - dry season), when measured accurately, can thus be used to make an estimate of the GPP over the rainforest.

The second question is addressed using the δ¹³C of CO2. This ratio changes as a function of the opening and closing of stomata by plants. Plants tend to close their stomata when there is water-stress: they actively try to prevent water-loss this way, but also lose capacity to take up CO2. This leads to detectable changes in both CO2 and its isotopic ratios, and we are currently publishing the method to analyze this.

Note that both approaches require measurements over large areas of rain forest, and considerable temporal coverage to average noise and find signal. This is why now, mid-way into the program, we have prepared the analyses tools (numerical model), started the measurements, but not yet analyzed the data in great detail.

When successful, ASICA will give new insights into the functioning of the Amazon rain forest. This large forest provides very important services to society, not just in terms of products (wood, food, medicine, etc) but also as a source of water to the area (though evaporation, precipitation, and rivers) and as a sink of carbon dioxide. The continued existence of this carbon dioxide sink is questioned, as current climate models disagree whether ongoing warming will turn the area into a source of CO2 or not.

The overall objectives of ASICA are therefore to:

- Collect air samples from high aircraft flying twice per month out of four sites over the Brazilian Amazon
- Dry these samples effectively to allow novel isotope measurements on them
- Analyze these samples for greenhouse gases and isotope ratios in Brazil
- Develop new knowledge on the budget of δ¹³C and Δ17O through modeling and measurements
- Analyze the samples using the new model, to learn about carbon exchange
- Build capacity to sustain these measurements long-term

Work performed

Preparing to do measurements

The projected started in Sep-2015 with the first preparations to do the new measurements. We had two main challenges:

Develop the capacity to dry air samples during the flights, because Δ17O isotopes can only be measured on dry air
Develop capacity to perform high-precision, well-calibrated isotope measurements on the air samples

For (1), we have managed to develop a completely new air dryer (design is currently being published and illustrated in Figure 1) which is capable of drying the very wet air from the Amazon (>4% water vapor) down to a dew point of less than -5 degC. Building the first prototype was a huge challenge, and we finished it in March 2017 (month 18). The drying device for ASICA flights has been designed and tested in the Groningen University laboratory (by D. Paul and B. Scheeren), with a very satisfactory result. Completion of all tests including flight safety was done in March of 2017, six months behind schedule. Without using battery power, or dangerous chemicals, or large amounts of space, our new prototype dryer can extract nearly all the water from an 11L/min airstream that comes in at 32 degC and 85% relative humidity, which is a remarkable feat. This capacity was tested while using the dryer in a greenhouse from the Groningen university, which mimicked the most extreme conditions we expect near the surface in the Amazon. Further testing was performed in March of 2017 using an Aerodyne isotope spectrometer in Groningen.

Next, we took the dryer to Boulder, CO to be tested in the world renowned isotope lab from INSTAAR, where they also have the same sampling equipment as used in Brazil. In the full tests, the dryer performed extremely well (Fig 2). We thus decided to ship the dryer to Brazil in September 2017. There, it was slightly modified and awaited a change in sampling strategy at the sites, before being deployed for the first time in February 2018. Since then, the ASICA air dryer is functioning very well at Rio Branco, and we are currently in the process of building more of them.

For (2), we have taken two separate approaches. One is to design and build a CO2 extraction system for the lab in Sao Jose de Campos, so that each air sample coming through their operations can have its CO2 frozen out and stored in a small glass vial for later analysis in the Groningen isotope lab in the Netherlands. Specific steps we performed include:

Design and testing of a prototype CO2 extraction system by partner Groningen University, postdoc Dipayan Paul, in the period 01/09/2015 through 01/09/2016.
Training of personnel from partner INPE (C. Correia and L. Domingues) on the use of the system in Groningen in March 2016
Installation of the extraction system in INPE, Brazil (by D. Paul, W. Peters, and B. Scheeren) in June 2016
Testing and acceptance of the extraction system (by L. Gatti and Stephane) in July-Aug 2016
Collection of the first air samples from Amazon by ASICA-sponsored flights in Nov-2016.
Shipment of first set of 80 vials of CO2 from INPE to Groningen in Dec 2016
First analyses of δ¹³C in CO2 at the Groningen Laboratory in January 2017
Continued sampling of more than 300 vials of CO2 (late Feb 2017)
The first analyses have indicated some possible water-contamination in our samples related to the design of the extraction system, which we can remedy without major problems at no additional costs. Finding this solution did lead to some additional work on our extraction system in the spring of 2017 performed by INPE PI Luciana Gatti.
Since then, more than 1800 flasks have been extracted by the INPE personnel (see Figure 3), and are now awaiting further measurements in Groningen. The first few δ¹³C profiles shown were measured already and show good signals. A system to automate their measurement is under development.

The second approach taken towards high-precision measurements is to purchase an Aerodyne TILDAS dual-laser spectrometer for use in Brazil. This devic

Final results

Mid-way through the project ASICA has already gone significantly beyond the state-of-the-art. It’s measurement methodologies for airborne isotopes are unique in the world, and uses an air dryer that is custom built and not available yet to any other project. The equipment installed at INPE in Brazil is unique: only one similar instrument exists in the world. It allowed us to make the first tropical tropospheric measurements of Δ17O in CO2, at a precision that is only rivaled by a few labs in the world. Modeling of Δ17O isotopes has not been done on any smaller scales than full hemispheres, and the full TM5 model we built is the first one to go to regional gradients of this tracer. In data assimilation of δ¹³C in CO2 we are the first to estimate discrimination by vegetation and show its relation to droughts. The SIFTERv2 fluorescence product we co-developed during ASICA is unique in its view of the tropical land surface GPP.

Until the end of the project we expect to make further steps in data assimilation of δ¹³C, Δ17O, COS, and SIF using our frameworks around TM5 and SIB4. These efforts will use the Amazon isotope data that we now collect on a frequent basis. Expansion of this effort is opportune and planned. As part of a collaboration with another project at INPE, we also expect to co-measure COS in the flasks we collect.

Website & more info

More info: