ccording to an analysis published by in-cites, the work of Dr.
Peter Rayner has entered the top 1% of the ISI
Essential
Science Indicators
Web product in terms of
total citations in the field of Geosciences, with 27 papers cited 538
times to date. Dr. Rayner is a researcher at Australia’s CSIRO
Atmospheric Research branch, where he is a member of the Greenhouse
Gases: Observations and Modelling Team, as well as the Coordinator for
both the CSIRO Biosphere Working Group and the Coupled Carbon-Cycle
Climate Model Intercomparison Project. In the essay below, Dr. Rayner
discusses his highly cited work.
|
There is a clear scientific and policy need to quantify and
understand the processes controlling carbon dioxide (CO2)
exchange between the Earth's surface and atmosphere. Routine
monitoring since the late 1950s has revealed an inexorable rise in
atmospheric CO2 concentration. This gas is currently the
single largest contributor to the enhanced greenhouse effect, a
contributor to climate change. However, much about the processes
controlling atmospheric CO2 concentration remains unclear.
Depending on how the anthropogenic inputs are defined, somewhere
between 50% and 60% of the CO2 added to the atmosphere each
year by human activity is removed, overwhelmingly by processes at the
Earth's surface. This removal rate is highly variable; in some years
the growth rate almost stalls, while in others there is almost no
natural removal. Further, there is the possibility that future climate
change itself will affect the removal processes.
|

“[Carbon dioxide] is currently the single largest contributor to the enhanced greenhouse effect, a contributor to climate change. However, much about the processes controlling atmospheric
CO2 concentration remains unclear.”
|
|
|
These exchanges are both large and ubiquitous, stymieing any
attempt at direct measurement on a global scale. We therefore resort
to either predictive modelling of the exchanges or the use of inverse
methods. In the latter, we combine knowledge of atmospheric
concentrations and atmospheric advection and diffusion to estimate,
statistically, compatible patterns of surface exchanges. This has been
the focus of my research over the last decade.
By the late 1990s there were records of CO2 concentration
from several observing sites stretching back at least a decade.
There were also two other vital clues to how CO2
concentration had changed. With remarkable prescience, scientists had
been collecting and storing atmospheric samples from Cape Grim in
Tasmania since the late 1970s, against the hope that atmospheric
species they could not measure at the time might later turn out to be
interesting. These samples provided a long and precise record of
atmospheric oxygen concentration and the isotopic composition of
atmospheric CO2. An inverse calculation using this oxygen,
isotope, and CO2 data formed the basis of a paper we
published in Tellus in 1999 (PJ Rayner, IG Enting, RJ Francey,
R Langenfelds, "Reconstructing the recent carbon cycle from
atmospheric CO2, delta C-13, and O-2/N-2 observations," Tellus
B-Chem. Phys. Meteorol. 51[2]: 213-32, April 1999). The
combination of data gave us our clearest window yet into the problem
that had occupied the carbon-cycle community through the 1980s and
1990s: How much of the natural removal of CO2 was due to
ocean uptake and how much was due to land? Oxygen concentration and
the isotopic composition of CO2 are affected differently by
the processes causing CO2 exchange in either land or ocean,
so we could use variations in these species to partition the CO2
exchange. The data suggested that the net removal was predominantly
oceanic, although the terrestrial uptake was about as large once we
considered the large source believed to come from deforestation. The
result has since been refined to account for some processes we
neglected, but it roughly stands, i.e., that similar amounts of CO2
are taken up by the ocean and the land.
The simultaneous use of the three sets of measurements also gave us
a hint of the year-to-year variability and spatial structure of the
uptake. We noted, for example, that the land was responsible for most
of the year-to-year variability and that most of this occurred in the
tropics. More recent results have increased the role of tropical land
in year-to-year variations in carbon flux.
Although we could make our conclusions on global carbon budgets
with some confidence, the extreme lack of data, especially over the
tropical continents, made our comments on the spatial distribution
highly uncertain. Also, the representation of atmospheric transport in
numerical models is subject to uncertainty and, with our one model, we
could not easily comment on this. Quite a bit of our work in the
following years has focused on possible improvements in the spatial
coverage of data, for example by the use of remote sensing of
atmospheric composition, or checking the robustness of atmospheric
inversions using many different numerical models of atmospheric
transport. There are now satellite missions under development which
may provide the dense coverage of measurements needed to reduce the
spatial uncertainties in our inverse calculations and allowing us to
make clear divisions, for example, between the continents in the
northern hemisphere.
Dr. Peter Rayner
CSIRO Atmospheric Research
Aspendale, Victoria, Australia
|