image
in print
Drs. Jeff Tschirley, Josef Cihlar,
Scott Denning and Rene Gommes *1
At its Fourth meeting (Stockholm, November 1999), the Integrated Global Observing Strategy Partners (IGOS-P) approved terrestrial carbon as a theme for which a systematic global observation initiative should be developed under the leadership of the Global Terrestrial Observing System (GTOS). They also approved a process whereby terrestrial requirements will be integrated with the ocean carbon observation requirements being developed in the IGOS ocean observation theme.
The need for systematic assessment of carbon pools has become more important
as countries realize that their status as a net "sink" or a "source" of
carbon visa vis UNFCCC and the Kyoto protocol has implications for their
future economic development. Policymakers are also understanding that a
large
portion of the global annual net primary ecosystems productivity is
found in economic spheres such as forests, cropland and pasture.
Although carbon pools and their long-term changes will be at the heart of ongoing negotiations under the UNFCCC, there is also need to estimate carbon fluxes which are of a more short-term nature. The reason is that the consistency and reliability of our data on carbon pools can only be validated using models which in turn highlight the gaps in our current knowledge. They are also useful for identifying new observation requirements that can lead to greater confidence in both the pool and the flux data.
Because many factors interact to affect the carbon cycle, both above and below the soil surface, information on terrestrial carbon must be obtained frequently and with a high spatial resolution. Given the limitations of measurement techniques, this has not been possible in the past. New methods, including observing techniques and process-oriented models, now make the problem more tractable. This is a major reason for the increasing research interest in the observation and quantification of terrestrial carbon.
To consolidate and systematise global observations of terrestrial carbon, there is a need to agree on the observation and modelling requirements; to harmonise the main relevant projects and activities that can contribute to a global observing 'system' and to identify gaps and ways to fill them. The Terrestrial Carbon Observation Initiative' (TCO) aims to meet these needs.
The number of programs and initiatives concerned with this topic is
steadily increasing, at national, regional and global levels. Some have
a strong
observational base and include satellite data, others are based on
modelling or on local measurements using a sampling strategy.
Some requirements for estimating terrestrial carbon are well understood although in certain cases the continuity of their measurement is of concern. They include: land cover and land cover change, leaf area, biomass burning (fire scars), solar radiation, atmospheric composition, surface fluxes, and crop and forest production. The observational requirements for these issues have been established through various experiments and regional studies.
Other observations are less well understood and techniques for their
measurement may not be sufficiently developed. While progress has been
made
in some cases, further work is needed before global observations can
be implemented. The most important remaining issues are: biomass and its
changes, canopy structure, atmospheric CO2 concentration at a micro-scale,
plant biogeochemistry, and meteorological parameters with a fine spatial
resolution.
The importance of these observations is generally accepted by scientists
and to varying degrees the observation techniques have also been developed.
However, there is lack of experience in dealing with these types of
data and therefore with the derivation of appropriate bio-physical variables
for use in carbon exchange models. Thus, further research and technology
development are also necessary.
There is a clear need to move beyond the current approach whereby process models are tested on a very small scale, pronounced to be valid and then extended to the global scale. An observation network is needed that allows the development and testing of both process-based ecosystem models, their extrapolation
*1 Respectively affiliated with: the Global Terrestrial
Observing System, the Canadian Centre for Remote Sensing,
Colorado State University, and the Food & Agriculture
Organization of the UN.
(to be continued on next page.)