Automated Sediment Volume Development Analysis with Uncertainty Propagation

More Info
expand_more

Abstract

Coastal regions change continuously due to natural processes. Human activities in coastal regions require a steady environment. Usually the dredging industry is contacted for coastal maintenance. Knowledge about the time development of sediment volume changes in specific regions is necessary for risk assessment in contracting. Nearly a century of available bathymetric data of the Dutch coastal zone can be used. Two different approaches of calculations of sediment volume changes are currently available from literature. Both are mainly manual calculations and they do not generate confidence bands. We set out to create two automated approaches, to analyse sediment volume changes with confidence bands. Development consisted of four steps: First, the two manual approaches were re-implemented as automated approaches. Unavailable data were calculated with a systematic and transparent algorithm filling the gaps in space and time. Second, the automated approaches were validated by 16 different unit tests with synthetic data. The difference between the calculations of the synthetic data, manual compared to automated-approaches, were acceptably small, demonstrating the validity of the automated approaches. Third, uncertainty was added to the automated approaches. Every step in the calculations to fill data gaps was identified and an uncertainty element added. This element merely points out the amount of uncertainty of the data. It is not a probabilistic uncertainty. Fourth, the automated approaches with confidence bands were used for various Dutch coastal regions as case studies to show differences with the two manual approaches from literature. These new automated approaches of sediment volume changes are validated, every step is transparent and reproducible and it generates confidence bands of the calculated data. Therefore it is a valuable additional input for risk assessment in contracting for coastal maintenance.