Estimating and validating the burial depth of conventional explosives : a hindcast of sandwaves on the Dutch Continental Shelf

This validation study builds on the Deltares study that estimated the burial depth of conventional explosives on the Dutch Continental Shelf (Deltares, 2023a). The report and summary below have been extended to include a validation of the methodology. Rijkswaterstaat and TenneT have requested Deltares to perform a seabed mobility analysis for the Dutch Continental Shelf (DCS). The study modifies and extends the analysis performed in the Deltares (2020a) pilot study for the DCS by applying a more detailed methodology and additional bathymetrical data. The overall objective of the Deltares activities is to generate one or more data products (i.e., maps and GIS files) that will indicate at what depth below the seabed objects can be expected and what the expected future seabed levels will be. A validation will assess the quality of both the methodology and the resulting data products. This information will simplify or preclude the choice of means for mitigation measures for risks associated with Conventional Explosives (CE) resulting from offshore activities in the subsurface – such as cable trenching or sand mining – depending on the location and situation. The aim of this morphodynamic assessment is to characterise the seabed dynamics and to quantify seabed level changes over the period 1945 to 2022 and over the period 2022 to 2050. Different levels will be provided describing the bandwidth in Lowest Seabed Levels for given years. This study focusses solely on the areas where sand waves are present on the DCS. In offshore areas without sand waves the sea floor dynamics are limited, hence resulting in relatively small seabed variations, and less predictable on this long timescale. The DCS is characterised by significant areas of sand waves particularly in the southern part of the DCS. Deepest water is found along the southwestern edge of the sand wave region. The water depth gradually decreases towards the coast. The area is furthermore characterised by tidal sand banks in the north and south, with the northern banks running from north to south and the southern banks facing the coast of Zeeland. In the southwestern part of the sand wave area, long bed waves with wavelengths of around 1.5 km can be seen. These rhythmic features are larger than sand waves, which typically have lengths ranging from 100 to 1000 m. To analyse seabed dynamics a detailed analysis is presented focussing on sand waves present on the DCS. Sand wave dynamics are determined by an assessment of historical bathymetric data, using a 2D cross-correlation technique. The sand waves typically migrate to the north-east, and locally to the south-west, with rates mostly ranging from 0 to 3 m/year. In the south-western part of the sand wave area, migration rates are low. Meanwhile, the north-eastern region shows higher migration rates which gradually increase closer to the Wadden Islands. The highest migration rates are found near coastal areas and on top of sand banks, which significantly redirect the migration direction of the sand waves in several areas. Based on the extrapolation of seabed trends, several seabed levels have been derived for the period 1945 to 2022 and for the period 2022 to 2050. These levels describe the hindcasted and predicted bandwidths in the Lowest Seabed Levels by means of a lower, mean and upper prediction (LSBLL, LSBLB, LSBLH). Furthermore Best-Estimate Bathymetries (BEB) are provided for the years 2022 and 2050. The analysis provides insight in the dynamics of sand waves on the DCS. The quality of the results is subject to a number of limitations. For example, limited spatial coverage of survey data, quality of available data and human interventions can significantly influence results. Quality checks resulted in exclusion of certain datasets because of limited quality. The predicted seabed level changes presented in this study follow from the applied morphological analysis techniques, describing the (uncertainty of the) physics and the natural variability of the analysed morphological system. No additional safety margins for design purposes have been applied. The presented hindcasted and predicted future seabed levels are a result of the applied methodology. This methodology was previously successfully applied in detailed morphodynamic assessments for wind farm zones on the DCS. Because of local characteristics, upscaling of the analysis to the entire DCS resulted in increased uncertainties of historic and future seabed levels as compared to detailed, local morphodynamic assessments. The presented seabed levels provide a good insight in historic and future bandwidths of seabed levels, however detailed assessments are recommended in case depths of buried objects and cable burial depths need to be determined with more accuracy. The presented seabed levels are a valuable dataset to (further) decrease the risks related to CE. Despite the limitations, as can be expected from a hindcast of 60 to 75 years, the resulting surfaces and maps will help in assessing the likelihood of presence of CE at and under the seafloor. Validation of the methodology using field-observed objects was limited by the number and location of objects encountered in the field. However, most burial depths fell within the predicted ranges. Cross-validation demonstrated that the methodology is reliable and consistent in areas with low sand wave migration rates. In regions with higher migration, performance improves when the interval between surveys is longer and the number of surveys is greater. Overall, the reliability of the generated data products depends strongly on the time span over which measurements are extrapolated. Additionally, any spatial or temporal inconsistencies in the bathymetric database used for this study can substantially reduce the quality of the resulting data products.