1st International Congress on Environmental Modelling and Software - Lugano, Switzerland - June 2002
Keywords
value/risk management, uncertainty, parsimonious/detailed modelling, auditable modelling
Start Date
1-7-2002 12:00 AM
Abstract
It is increasingly the case that models are being developed as “evolving” products rather than one-off application tools, such that auditable modelling versus ad hoc treatment of models becomes a pivotal issue. Auditable modelling is particularly vital to “parsimonious modelling” aimed at meeting specific modelling requirements. This paper outlines various contributory factors and aims to seed proactively a research topic by inextricably linking value/risk management to parsimonious modelling. Value management in modelling may be implemented in terms of incorporating “enough detail” into a model so that the synergy among the constituent units of the model captures that of the real system. It is a problem of diminishing returns, since further reductions in the constituent units will create an unacceptable difference between the model and the real system; conversely, any further detail will add to the cost of modelling without returning any significant benefit. The paper also defines risk management in relation to modelling. It presents a qualitative framework for value/risk management towards parsimonious modelling by the categorisation of “modelling techniques” in terms of “control volume.”
Parsimonious Catchment and River Flow Modelling
It is increasingly the case that models are being developed as “evolving” products rather than one-off application tools, such that auditable modelling versus ad hoc treatment of models becomes a pivotal issue. Auditable modelling is particularly vital to “parsimonious modelling” aimed at meeting specific modelling requirements. This paper outlines various contributory factors and aims to seed proactively a research topic by inextricably linking value/risk management to parsimonious modelling. Value management in modelling may be implemented in terms of incorporating “enough detail” into a model so that the synergy among the constituent units of the model captures that of the real system. It is a problem of diminishing returns, since further reductions in the constituent units will create an unacceptable difference between the model and the real system; conversely, any further detail will add to the cost of modelling without returning any significant benefit. The paper also defines risk management in relation to modelling. It presents a qualitative framework for value/risk management towards parsimonious modelling by the categorisation of “modelling techniques” in terms of “control volume.”
