Calibration and Reliability
in Groundwater Modelling :
A Few Steps Closer to Reality
Edited by K. Kovar & Z. Hrkal
IAHS Publication 277 (published August 2003) ISBN 1-901-502-07-4; 526 + x pp; price £78.80
Financial sponsorship for this publication was provided by:
The International Hydrological
Programme of UNESCO
Environmental Modeling Systems Inc., South Jordan, Utah, USA
The abstracts of the papers in this volume can be seen at
www.iahs.info
together with information about other IAHS publications and IAHS activities
Groundwater systems are arguably the most difficult earth systems to model because the systems are inaccessible yet, as they are a source of drinking water, very detailed knowledge is required. The usefulness of predictive simulations obtained by groundwater models is often hampered by the inability to indicate and quantify the reliability of model results. Uncertainty in predictions primarily stems from errors relating to the model formulation, such as:
- inadequate conceptualization of processes and interactions,
- inadequate description of the state of the system, and
- inadequate description of spatial and temporal variability.
Recent research has resulted in a variety of approaches that can be used to incorporate the information about such errors into the modelling process and to establish the level of uncertainty in model-based decision making. Increasingly, these new techniques are being applied to a variety of problems and sites.
This volume comprises 68 papers selected from ModelCARE 2002 (the fourth of the ModelCARE conference series) and peer-reviewed; they address:
- Modelling Concepts
- Stochastic Modelling
- Parameter Estimation and Model Calibration
- Conceptual Model Uncertainty and Scale
- Reliability of Model Prediction
- Comprehensive Use of Field Information
- Methods and Tools for Assessment of Well-head Protection Zones
A major feature is the discussion of both deterministic and stochastic methods of investigation. For any given groundwater system, some things about the system are known relatively well and it is most advantageous to represent them in a model definitively, or deterministically. Other things are less clearly known or little known, and it is important to represent them using stochastic methods to obtain an honest evaluation of what can be said about the system. Thus, both deterministic and stochastic methods are important. Yet, progress in the two approaches most often proceeds independently, and sometimes divisively.