University of Twente Student Theses


Shifting Sands : Developing new measurement methodologies in GIS to analyse the spatial variability of tidal sand wave migration on the Netherlands Continental Shelf

Meijden, Rens van der (2021) Shifting Sands : Developing new measurement methodologies in GIS to analyse the spatial variability of tidal sand wave migration on the Netherlands Continental Shelf.

[img] PDF
Abstract:Tidal sand waves are large-scale bed features that occur in sandy, relatively shallow coastal seas, such as the Netherlands Continental Shelf (NCS) and the South China Sea. They have typical heights and wavelengths of 1-10 m and 100-1,000 m, respectively. Furthermore, they migrate with typical speeds of 0-5 m/year. It is mainly the combination of their significant dimensions and dynamic behaviour which makes sand waves highly relevant to a variety of offshore activities. For the NCS, there is an abundance of observational studies that quantifies sand wave migration. However, these mostly consider a limited number of sand waves at small-scale survey sites. This way, areally averaged estimates for migration rate and direction are less robust and large-scale spatial patterns are not revealed. Moreover, commonly applied measurement methodologies only capture migration perpendicular to the sand wave orientation. The aim of this research is to gain insight in the spatial variability of tidal sand wave migration on the NCS. This is done by developing an efficient migration detection method in a Geographic Information Systems (GIS), which quantifies migration rate and direction from bathymetric timeseries (a combination of two unique bathymetric surveys in time) in two horizontal directions. Two potentially suitable methods were implemented in QGIS: Pairs of Source and Target Points (PSTP) and Spatial Cross Correlation (SCC). PSTP quantifies migration rate and direction by measuring the displacement of vectorised crest and trough lines, which are extracted from consecutive bathymetric datasets in time. SCC locates the two most correlated points in different bathymetric datasets and quantifies migration rate and direction by measuring the distance and angle between these points. To validate the output of PSTP and SCC, artificial bathymetric datasets were prepared for which migration distance and direction was pre-defined. When the migration distance is between the raster resolution (25 m) and half the wavelength, migration is measured in the correct direction and with sufficient accuracy. It was found that, when analysing field data, the accuracy of both methods is limited for migration distances that are small relative to the raster resolution. To account for this, the time in between two surveys was taken sufficiently long. The performance of PSTP and SCC was tested at four field sites: an offshore site, a site with multiple large sandbanks, and moderately dynamic coastal site and a highly dynamic coastal site. These sites were selected to cover a variety of different morphodynamics and underlying topographies. The obtained spatial and frequency distributions of migration rate and direction were qualitatively comparable for all locations. Quantitative differences mainly occur because the measurements of PSTP are restricted to integer grid cell positions whereas SCC can provide measurements in between these grid cell positions. Furthermore, using a single set of input parameters for SCC is not optimal at locations where the sand wave morphodynamics strongly vary. PSTP was selected as the most suitable method for a large-scale analysis of sand wave migration at the NCS. Its preferability is mainly based on the fact that, compared to SCC, the optimal input parameters are quickly found and need less changing in between different analyses. In addition, sources of errors and anomalies can be readily identified. Subsequently, PSTP was used to analyse over 300 bathymetric timeseries. The analysed timeseries cover virtually all sand wave fields on the NCS. To gain insight in the spatial variability of migration rate and direction, the resulting 1,900,000 data points were aggregated to an average value per square kilometre. from north to east on the seaward flanks of the Brown Ridge and the Breeveertien banks, which are sandbanks located roughly between 40-100 km offshore of Noord-Holland. Furthermore, clockwise bilateral migration was observed on the flanks of the Brown Ridge. On the flanks of the Zeeland Banks, a notable observation was that the sand waves migrate in a clockwise as well as an anti-clockwise direction. Finally, the sand waves in the south-western corner of the NCS migrate to the north-east and south-west on a scale smaller than one square kilometre. Migration rates show significant spatial variability over the NCS. Average migration rates at the main field, located between IJmuiden and Zeeland, typically range between 0-3 m/year, although migration rates of up to 12 m/year are observed at the Zeeland Banks. The fields at the Wadden Island are more dynamic, with average migration rates typically ranging between 2-8 m/year. The highest average migration rates of 10-21 m/year occur near the coast of Texel and Vlieland. Three spatial patterns were revealed. First, migration rates increase non-linearly from the North Hinder bank in the south-western corner of the NCS in north-eastern direction towards the Wadden Islands. Furthermore, migration rates decrease rapidly from the coast in offshore direction, which is especially clear at the Wadden Islands, IJmuiden and the Maasvlakte. Finally, migration rates are higher where sand waves are located on top of a shoreface connected ridge or a sandbank, as was seen at the Zeeland Banks, the Brown Ridge and the Breeveertien Banks. For the latter, the sand waves on the seaward flank migrate faster than the sand waves on the landward flank. The obtained results are scientifically and practically relevant in multiple ways. First, this research has shown that GIS are highly applicable for analysing the spatial variability of tidal sand wave migration in shelf seas. Present-day GIS packages come with a vast amount of data manipulation and visualisation tools, which gives them an edge over programming packages such as Python or MATLAB. Even more so because GIS packages such as QGIS and ArcGIS have an integrated Python environment in which processing tools can be automated. With PSTP and SCC, two methods have been developed that have shown great potential for future analyses on sand wave migration at different spatial scales. Furthermore, a great amount of migration data has been produced which can be the starting point for new research. Insight in the contribution of processes and parameters governing sand wave migration can be obtained by correlating the aggregated migration rates to environmental data. In addition, the migration data can be used for the verification and validation of future process-based models. Finally, the NCS-wide analysis provides insight in the spatial variability of sand wave migration rate and direction at an unprecedented coverage. These insights can be valuable for offshore infrastructure planning, optimising hydrographic re-survey policies, dredging and mine hunting strategies.
Item Type:Essay (Master)
Faculty:ET: Engineering Technology
Subject:56 civil engineering
Programme:Civil Engineering and Management MSc (60026)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page