Environmental Engineering Reference
In-Depth Information
Chapter 23
How Valid Are Model Results? Assumptions,
Validity Range and Documentation
Hauke Reuter, Fred Jopp, Broder Breckling, Christoph Lange,
and Gerd Weigmann
Abstract In the previous chapters we have described a spectrum of different model-
ling techniques and application examples. Now we return to overarching aspects
which are relevant for model developments of different types. Model validity and
considerations which conclusions can be derived from model results are presented
jointly for different modelling techniques and application fields. Furthermore, we
discuss adequate means of communicating the models to others. The overall views
are largely theory-based and show that knowledge on the theoretical background can
present an important guidance to making the most appropriate use of ecological
modelling results.
23.1 The Last Stage of Development: Limits of Reliability
In the first chapter of this topic we started the introduction of ecological modelling
by pinpointing the possible discordance in the relationship of formalized models
and real-world ecological objects. When assessing the quality of models and
model results, we have to keep this in mind. The structure of an ecological
model does not represent the structure of the real ecological objects one-to-one.
It is a formalized extract to represent ecological interactions. When discussing the
limits of reliability for the developed construct, it is important to be aware of this
difference. A computer model is not a neutral, objective compound. It is based on
the underlying assumptions of model development (Chaps. 1 and 2). A relevant
but frequently forgotten part of model development derives from the abstraction
concept which underlies the modelling approach. The first step to evaluate the
