Geoscience Reference
In-Depth Information
Despite that methodological agnosticism, the desire for greater integration does
highlight the importance of modelling in order to render the problem of ESS
manageable:
For a situation as complex as the earth, model[l]ing has a critical role. Only by reduc-
ing qualitative perceptions to quantitative formulas is it possible to communicate ideas
effectively across disciplinary boundaries and to analyze the subtle interactions and
feedback loops which control the overall functioning of the system. It may well be that
a complete numerical model of the whole system is never constructed and, unless fi rmly
grounded in observation, the results of such a model are assuredly debatable. Neverthe-
less, experience shows that the attempt to identify its essential features can help focus
critical issues, and maintain the perspective and balance which are essential in a
program[me] aimed at overall understanding. (Bretherton, 1985, p. 1119)
This model is presented as a pair of systems diagrams (fi gure 10.2) that aim to repre-
sent a way of analysing global change on a decadal to centennial timescale by divid-
ing the Earth system into a physical climate system and a biogeochemical cycling
component. Bretherton notes that these two components are actually relatively
weakly coupled in the model and goes on to point out four major caveats with
respect to the systems diagrams. First, the diagrams - and hence model of ESS - are
specifi c to research objectives, and it is these objectives that control the scale of
representation. In the particular case presented, there is a very explicit timescale as
well as a global spatial scale. Secondly, the representation is descriptive and not
functional and thus does not make claims to completeness. Thirdly, although
strongly affected by and having major impacts on the Earth system, humankind is
regarded as external to it. Fourthly, there is an assumption that the Earth system can
be defi ned in terms that are deterministic, and thus, predictable, even though parts
of the system - e.g., weather and climate - are known to exhibit chaotic behaviour.
We will return to the implications of these caveats later in this chapter.
On the other hand, Bretherton's paper provides a methodological statement
about the need to employ remote sensing as a way of informing and testing the
suggested approach. He highlighted fi ve roles of remote sensing that needed to be
developed. First, it provides the necessary global synoptic coverage and shifts
emphasis from relatively disconnected point measurements that characterised a
number of scientifi c approaches. Secondly, ESS forces a rethinking of algorithms
employed in remote sensing because of the complexity of extracting a signal that
can be meaningfully used in the parameterisation and testing of models. Thirdly,
the emphasis within ESS is on change and therefore the need for ongoing measure-
ments, with remote sensing being the most cost-effective way of doing so. Fourthly,
ESS should promote better practice for integrated data management to understand
what is going on in different spectra and thus to characterise different parts of the
Earth system simultaneously. Fifthly, there was a need for more training to remove
remote sensing from the minority role it had at the time. The extent to which these
fi ve roles have been addressed will be considered later.
Of course, such developments could not occur in a vacuum. They would need
signifi cant funding initiatives, international cooperation and data exchange, and
mechanisms for linking research with governmental and industrial requirements.
Bretherton's paper refl ected a major US initiative that included efforts from NASA,
NOAA and the National Science Foundation, together with inputs from the US
