Morphological acceleration factor: usability, accuracy and run time reductions

Knaapen, M. and Joustra, R (2012) Morphological acceleration factor: usability, accuracy and run time reductions. In: TELEMAC-MASCARET User Conference 2012, 18-19 October 2012, Oxford.

Preview

Text HRPP543.pdf Download (541kB) | Preview

Abstract

Within SISYPHE 6.2, the option is created by BAW to use a morphological acceleration factor (MF) within the coupled TELEMAC2D-SISYPHE model. Multiplying both the evolution and the time by the same factor the model jumps forward in time, reducing the required computation time. This paper presents the usability of this approach, the gains in computation time and the loss of accuracy of this approach. Three different cases were used: a laboratory case of a trench in a constant flow, a river flood case and an estuarine test case. For the river flood case, a single event with varying water discharges, the approach is unsuitable. Using the MF implies that the water levels change too rapidly, altering the hydrodynamics. The same would be the case for tidal flow, but the morphological acceleration factor can still be used due to the repeating nature of the tide [1]. The downside is that strictly the results using a factor N are only valid after exactly N tides. For steady cases the approach can be used flexibly without limitations. Comparisons with the measured data the trench case show that the MF can be used with only limited loss of accuracy. The simulation time reduces rapidly, while the model skills only reduce marginally, up to the MF is 90. The simulation time initially reduces rapidly. For the larger model of an estuary, the gain using a MF of 20 reduced the run time by a factor 20. In this case however, the model does show some significant changes in the prediction.

Actions (for site administrators only - login required)

View Item