TY - CHAP
T1 - Chapter Two Good Modelling Practice
AU - Crout, N.
AU - Kokkonen, T.
AU - Jakeman, A. J.
AU - Norton, J. P.
AU - Newham, L. T.H.
AU - Anderson, R.
AU - Assaf, H.
AU - Croke, B. F.W.
AU - Gaber, N.
AU - Gibbons, J.
AU - Holzworth, D.
AU - Mysiak, J.
AU - Reichl, J.
AU - Seppelt, R.
AU - Wagener, T.
AU - Whitfield, P.
PY - 2008
Y1 - 2008
N2 - Models have become indispensable in environmental assessment, planning and management. However as models have increasingly been developed and disseminated, the risk of their misuse or misunderstanding of their capabilities has increased. Whether a model is used for simulation, prediction, decision making or communication of scientific analyses, it is important that its development and application conform to protocols or standards that help to maximise the scientific soundness, utility and defensibility of models and their outputs. The complexity and uncertainty inherent in environmental assessment make the pursuit of good modelling practice especially important, in spite of limited time and resources. This paper is an attempt to identify the key components of best modelling practice and our collective progress in its achievement, taking into account previous relevant reviews undertaken by several authors and agencies. The details are always likely to be the subject of lively debate, but the general components of 'good modelling practice' are probably not controversial. They are clear purpose, adequate reporting, and serious evaluation. Although these are common strands in the various definitions of good modelling practice the emphasis varies between different types of model application. For this reason it is important that good practice should not become overly prescriptive. We report a preliminary analysis which suggests that progress towards improving modelling practice is slow. This is despite very widespread agreement on what constitutes good practice. Why is this so? In the research community at least, the drivers for model development and evaluation are funding and publication. If modelling practice needs to be improved, and we think it does, sponsors and journals need to take a lead in creating an environment where developing a model requires that the work be undertaken under some system of good modelling practice. The suggestion has been made of a 'good practice check list' in the journal, Environmental Modelling and Software. While such a system would need to be flexibly applied, the principle is sound, and such steps should move us forward.
AB - Models have become indispensable in environmental assessment, planning and management. However as models have increasingly been developed and disseminated, the risk of their misuse or misunderstanding of their capabilities has increased. Whether a model is used for simulation, prediction, decision making or communication of scientific analyses, it is important that its development and application conform to protocols or standards that help to maximise the scientific soundness, utility and defensibility of models and their outputs. The complexity and uncertainty inherent in environmental assessment make the pursuit of good modelling practice especially important, in spite of limited time and resources. This paper is an attempt to identify the key components of best modelling practice and our collective progress in its achievement, taking into account previous relevant reviews undertaken by several authors and agencies. The details are always likely to be the subject of lively debate, but the general components of 'good modelling practice' are probably not controversial. They are clear purpose, adequate reporting, and serious evaluation. Although these are common strands in the various definitions of good modelling practice the emphasis varies between different types of model application. For this reason it is important that good practice should not become overly prescriptive. We report a preliminary analysis which suggests that progress towards improving modelling practice is slow. This is despite very widespread agreement on what constitutes good practice. Why is this so? In the research community at least, the drivers for model development and evaluation are funding and publication. If modelling practice needs to be improved, and we think it does, sponsors and journals need to take a lead in creating an environment where developing a model requires that the work be undertaken under some system of good modelling practice. The suggestion has been made of a 'good practice check list' in the journal, Environmental Modelling and Software. While such a system would need to be flexibly applied, the principle is sound, and such steps should move us forward.
KW - model evaluation
KW - model transparency
KW - modelling guidelines
KW - quality assurance
UR - http://www.scopus.com/inward/record.url?scp=51549098800&partnerID=8YFLogxK
U2 - 10.1016/S1574-101X(08)00602-9
DO - 10.1016/S1574-101X(08)00602-9
M3 - Chapter
SN - 9780080568867
T3 - Developments in Integrated Environmental Assessment
SP - 15
EP - 31
BT - Environmental Modelling, Software and Decision Support
A2 - Jakeman, A.J.
A2 - Voinov, A.A.
A2 - Rizzoli, A.E.
A2 - Chen, S.H.
ER -