Data-Based Mechanistic Modelling and the Emulation of Large Environmental System Models - Environmental Modelling: Finding Simplicity in Complexity - page 107

7

Data-Based Mechanistic
Modelling and the Emulation
of Large Environmental System
Models

Peter C. Young 1,2 and David Leedal 1
1 Environmental Science, Lancaster Environment Centre, UK
2 Fenner School of Environment and Society, Australian National University, Canberra,
Australia; School of Electrical Engineering and Telecommunications, University of New
South Wales, Sydney, Australia

terms has permeated scientific endeavour over this period
and has led to a pattern of scientific investigation that is
heavily reductionist in nature. Such deterministic reduc-
tionism appears to be guided by a belief that physical
systems can be described very well, if not exactly, by deter-
ministic mathematical equations based on well known
scientific laws, provided only that sufficient detail can
be included to describe all the physical processes that
are perceived to be important by the scientists involved.
This belief leads inexorably to large, nonlinear models
reflecting the scientist's perception of the environment as
an exceedingly complex dynamic system.
Although deterministic reductionism still dominates
environmental modelling, there are some signs that
attitudes may be changing. There is a growing realiza-
tion that, despite their superficially rigorous scientific
appearance, simulation models of the environment based
on deterministic concepts are more speculative exten-
sions of our mental models and perceptions of the real
world than necessarily accurate representations of the
real world itself. The revived interest in the 'top-down' 1

7.1 Introduction

This chapter discusses the problems associated with
environmental modelling and the need to develop
simple, 'top-down' 1 , stochastic models that match the
information content of the data. It introduces the
concept of data-based mechanistic (DBM) modelling and
contrasts its inductive approach with the hypothetico-
deductive approaches that dominate most environmental
modelling research at the present time. The major
methodological procedures used in DBM modelling are
reviewed briefly and a practical example illustrates how
they have been applied in a hydrological context. The
chapter also shows how this same methodology can
be used as a basis for the simplification and emulation
of large, dynamic simulation models, including global
climate and hydrological examples (see also Chapter 26).
The environment is a complex assemblage of interact-
ing physical, chemical, and biological processes, many of
which are inherently nonlinear, with considerable uncer-
tainty about both their nature and their interconnections.
It is surprising, therefore, that stochastic dynamic models
are the exception rather than the rule in environmental
science research (see also Chapter 8). One reason for this
anomaly lies in the very successful history of physical
science over the last century. Modelling in deterministic

1 Note that the definition of 'top-down' and 'bottom-up' differs in
this chapter from elsewhere in the topic (e.g. Chapters 2 and 18),
and how the same terminology can come to mean essentially the
opposite idea to those working in different disciplines.