Verification and Validation Practice in Astrophysics

Thomasz Plewa
The Center for Astrophysical Thermonuclear Flashes
University of Chicago

Astrophysics offers a number of challenges for observers, theorists and modelers alike. Due to its natural complexity, astrophysics has always been one of the most attractive fields for development and application of new numerical methods. It has also offered a great deal of quite unexpected if at times clearly suspicious results. Fame of some theoretical models was cut short by ever improving observations. For several applications, however, rich sets of possible "solutions" remains available. If the nature happens to act in more efficient way than theorists, it is quite conceivable that some of those solutions are potentially wrong either due to oversimplified or incorrect physics or perhaps buggy computer codes. There are also known cases of seemingly correct solutions obtained thanks to application of frivolous computational practices that escaped scrutiny of a peer review process. I will take a closer look at a few "best" V&V-aided studies in theoretical astrophysics. These will illustrate the current notion of V&V in astrophysics, typical approaches to code verification and an extreme challenge of astrophysical code validation.

Return to Program