Environmental Engineering Reference
In-Depth Information
17
Finding Simplicity
in Complexity in
Biogeochemical Modelling
H ordur V. Haraldsson 1 and Harald Sverdrup 2
1 Naturvardsverket, Stockholm, Sweden
2 Department of Chemical Engineering, Lund University, Sweden
purpose, understanding, is lost through explanations and
comments such as
17.1 Introduction to models
Finding simplicity in complexity is the driving force
behind any scientific modelling process. What is regarded
as an achievement within research is the ability to test a
hypothesis successfully on any given problem by creating
simple models that can explain a complex reality. Sim-
plification is a process that is initiated by the desire to
capture the essence of a complex problem. The simplifica-
tion is formed either objectively or subjectively. But total
objectivity in research is a mere illusion. Modellers often
find themselves slipping into the practice of overcom-
plexity, or being locked into certain routines or subjective
opinions.
A model is a simplified representation of an observed
aspect of the real world. A model is any consequence
or interpretation taken from a set of observations or
experience. Many problems in natural systems are so com-
plex, nonlinear and multidimensional that they require
a nonlinear approach. Traditionally, simplification has
seldom been dealt with in nonlinear fashion. Rather,
linear correlation between different independent compo-
nents that have ill-defined causal relations has been used.
This requires complex explanations and reduces under-
standing of the fundamental dynamics behind complex
problems. The understanding is then not the focus of
the study but the constructed model itself. The original
the biological system is determined by unknown
force so we cannot understand it
...
...
There are thousands
factors affecting
...
It cannot be observed, but it is very
important for
Well, it is always different in the real
world, you know so it is no use to try to explain it
...
...
Models that require such explanations lack transparent
principles and processes, and are hard to communicate.
Validating models requires insight and an understanding
of processes - how the essential parts of the model are
constructed.
Models are important in research, not because they
produce results themselves but because they allow com-
plex and nonlinear systems to be investigated and data
from such systems to be interpreted. With models, the
interaction of several simultaneous processes in a single
experiment can be studied. Basically all models serve one
or both of two purposes:
testing the synthesized understanding of a system, based
on mathematical representation of its subsystems and
the proposed coupling of subsystems;
predicting what will happen in the future, based on the
ability to explain how and why things have worked in
the past.
 
Search WWH ::




Custom Search