Predicting migration of sand waves based on shapes only, without temporal information

With the increasing demand on offshore space for human activity, the importance of sand wave dynamics grows. Sand waves are rhythmic bed level variations up to five metres high, with a spacing of 200 to 400 metres and migration rates up to tens of metres. Their presence reduces the navigation depth and their migration, is a danger to pipelines and cables buried in the seabed. Furthermore, their shape and migration rates are indicators for residual sediment transport patterns. Only recently, migration has been explained as a result of asymmetry in the tidal flow due to a residual current (Nemeth, 2003) and higher tidal components (Besio, 2004). However, the models behind this explanation require information on the residual flow and the phase differences between the tidal flow components, which are difficult to obtain. As a result, bathymetric time series are required to tune the models to local conditions. For practical purposes, this is a major disadvantage, as these time series are scares. Our research shows that sand wave migration can be predicted based on a single bathymetric observation. We analysed about 10 sets bathymetric field surveys. Each set contains either multi-beam echo soundings or single-beam surveys along pipelines, covering up to 7 years. For all surveys, we determine the crest and trough positions of the sand waves. These positions are then analysed to give wavelength height and lee-stoss asymmetry and migration rates of the sand waves. Comparison of the found sand wave characteristics shows a correlation between the migration rate and the lee-toss asymmetry. Using knowledge of the sand wave model, this correlation is translated to a migration rate predictor that is based on a single bathymetric survey, the mean tidal velocity and basic sediment characteristics. This information can be collected easily in a short period.