Spin-up phenomenon in morphodynamic modelling

Abstract

Despite the
substantial progress in last decades in modelling short-term hydrodynamic and
sediment transport processes as well as long-term (centuries) morphological
development, the ability of morphodynamic models to simulate medium-term (decades)
development of estuaries is still very limited. One of the causes for the
limitation is the problem of spin up. Many application of morphodynamic models show
that at the beginning of the model simulation model results are unrealistic,
and only after a certain period the model gives meaningful results. A extreme example
is presented by Dam et al. (2016), who simulated the morphological developments
of the Western Scheldt estuary (The Netherlands) for decades to more than a
century. They concluded that the performance of the model, expressed in the
Brier Skill Score (BSS), first decreases and then increases. The model results
are "good" only after many decades. In this study, we aim to obtain
more insight into the problem. We try to answer the following questions: x What are the causes of the spin-up
problem? x What
is the time scale of the spin-up? x What
can we do to avoid or minimise the problem?