Confidence in Coastal Forecasts

Confidence in Coastal Forecasts

Baart, F.

Stive, M.J.F. (promotor)

Civil Engineering and Geosciences

Hydraulic Engineering

2013-06-05

This thesis answers the question "How can we show and improve our confidence in coastal forecasts?", by providing four examples of common coastal forecasts. The first example shows how to improve the estimate of the one in ten thousand year storm-surge level. The three dimensional reconstruction, based on paintings, shows that structural erosion allowed the sea to swallow the church of Egmond aan Zee in 1741. Storm surges as big as those in the 18th century have not been seen in the last century. Including the largest storm surges of the 18th century results in a 10cm higher storm-surge level estimate. The second example of a confidence interval shows that since the 1990 policy change (increased nourishments) the structural erosion is now reversed into structural accretion at the Holland Coast. The third example is the estimate of sea-level rise. Since the 1950s it was predicted several times that sea level at the Dutch coast will accelerate. So far, this did not happen. It is shown that the recent acceleration in Dutch sea-level rise is not due to climate change but due to the 18.6 year tide, reducing the confidence in these predictions. The final example is the extension of the operational forecast system with forecasts of coastal morphology, including confidence bands. To be able to warn citizens when a dune is about to fail, during a storm, it is important that we can predict coastal morphological changes several days ahead. The system was tested using four recent storms, which showed that the coastal changes can be predicted up to three days ahead with reasonable skill. For providing warnings of possible dune failure this is almost enough. To answer the question "how to show confidence?" a checklist for forecasts is presented. To answer the question "how to improve our confidence", it is discussed that from all the improvements in the scientific method, those that assume biased scientists have not found their way into the coastal forecasts. To improve confidence, coastal managers, the doctor of the coast, together with coastal researchers should further adopt the "evidence based practice”.

statistics
coastal morphology
coastal engineering
scientific method
research methods
forecasts

https://doi.org/10.4233/uuid:64161304-4714-4214-9790-e0da4a71399d

9789461917645

Institutional Repository

doctoral thesis

(c) 2013 Baart, F.