Validation of performance assessment models Page: 3 of 15
This article is part of the collection entitled: Office of Scientific & Technical Information Technical Reports and was provided to UNT Digital Library by the UNT Libraries Government Documents Department.
Extracted Text
The following text was automatically extracted from the image on this page using optical character recognition software:
dependence of the real system while still retaining what are believed to be salient features and
processes of greatest importance. Commonly, our objective is to develop and employ a
conceptual model that embodies a minimum of these simplifying assumptions while producing
economical and realistic simulations of the real system. Such a conceptual model is then
translated into a well-posed mathematical problem. To obtain a solution to the mathematical
problem, two approaches are commonly taken. The first approach can be pursued whenever the
mathematical problem is sufficiently simple to allow a closed-form analytical solution. The
solution is then programmed to create a computer code and obtain the analytical solution. In the
second aporoach, the mathematical problem remains too complex to allow an analytical solution,
and a numerical (computer) code using numerical methods (e.g., integrated finite differences,
finite elements, etc.) is created and a numerical solution is obtained. Both analytical and
numerical solutions are referred to as mathematical solutions.
In the context of these definitions, verification speaks to the consistency between a well-
posed mathematical problem and its solution or simulation by the computer code. Verification
begins with a conceptual model and only deals with our ability to accurately simulate a site
using a particular code. Thus, verification does not deal with the complexity of an actual site.
In the simplest terms, verification is confirmation that the computer code is properly encoded
and produces solutions consistent with those of other codes and previously obtained analytical
solutions that the numerical code is able to simulate.
Benchmarking is another term used in the IAEA (1982) definitions. “Benchmarking is the
comparison of solutions of the same problem between computer codes. Benchmarking is used to
gain confidence that the code being applied is correctly solving the equations embodied In the
code by comparison with an independent calculation or another code. Benchmarking is
considered part of the verification step.”
Given that one employs a verified code that has been calibrated with data from an actual site
to form a numerical model, validation addresses the much broader issue of the ability of a model
to represent the real system. A valid model is one confirmed to provide a good representation of
the real system. A more recent and subtly different definition of validation is provided by the
INTRAVAL Validation Overview and Integration Committee (VOIC). The VOiC defines validation as
a three-step process (SKI 1990):
1. understanding the active processes and geologic structure
2. resolving how well one can quantitatively simulate through comparisons to field
experiments
3. providing peer review and public scrutiny.
Note, this process provides a "good representation of the real system" as called for in the IAEA
definition; however, the VOIC has provided a more explicit statement of the process necessary to
ensure a good representation. Laboratory experiments and simulation of these experiments can
serve to identify and validate relevant physical processes. The VOIC has also made explicit
reference to the important role that an understanding of the geologic structure plays in the
validation process.
2
Upcoming Pages
Here’s what’s next.
Search Inside
This article can be searched. Note: Results may vary based on the legibility of text within the document.
Tools / Downloads
Get a copy of this page or view the extracted text.
Citing and Sharing
Basic information for referencing this web page. We also provide extended guidance on usage rights, references, copying or embedding.
Reference the current page of this Article.
Bergeron, M.P. & Kincaid, C.T. Validation of performance assessment models, article, November 1, 1991; Richland, Washington. (https://digital.library.unt.edu/ark:/67531/metadc1095123/m1/3/: accessed April 25, 2024), University of North Texas Libraries, UNT Digital Library, https://digital.library.unt.edu; crediting UNT Libraries Government Documents Department.