How to Build Reliability in the Results of Numerical Modeling

Engineering designs and assessments in water and environmental projects are increasingly relying on computational models as an alternative, or supplement, to experimentation. Verification and validation (V&V) and uncertainty quantification (UQ) techniques constitute a set of tools to provide quantitative insight into the reliability of those developed computational models.

This workshop is designed to give the audience a broad view and hands-on experience of V&V and UQ in the context of numerical modeling in water and environmental engineering.

An in-depth introduction to the fundamentals of verification procedures in scientific-computing is the first part of the course. Techniques for validation of models based on measurements and validation experiment design are discussed next. Then, a brief review of techniques for the quantification of uncertainty due to model, numerical techniques and parameters is presented.

The workshop concludes with the emerging techniques of SQA for computational modeling such as “literate programming”, “automation of documentation”, “version control systems”, “reproducibility” etc.