The 1% AEP wave and storm-tide level values were determined independently, and brought together in a joint probability analysis to produce a 1% AEP event. The likelihood of the two extremes occurring at the same time is very low, far in excess of an event that would be statistically likely to occur with a 1% annual probability. Although this is not the most conservative case for a hazard assessment, part of the scope of this project was to provide a more refined estimate of conditions and resultant hazards compared to previous work (i.e. Future Coast). Also, the results of this study are for inundation and coastal erosion hazard management purposes, not detailed design, therefore the inherent uncertainties that are associated with bivariate analysis are considered acceptable.

A methodology similar to that of Galiatsatou and Prinos (2011) and Shand et. al. (2012) was undertaken. The long-term records of waves and water levels and residuals were used in the joint probability assessment, long-term concurrent data records were only available for the open coast. Two sets of design conditions were determined for consideration in subsequent assessments, a worst case condition for wave impact (Table 4-6) and a case for overtopping / inundation analysis (Table 4-7). A number of test cases and sensitivities were carried out for the erosion modelling. The wave impact conditions (a larger water level with lower storm-tide level) was not used due to the shallow nearshore bathymetry. The depth-limited waves appeared to be unable to penetrate inshore in the model, especially in combination with a lower joint probability storm-tide level. The most appropriate condition to be used along the open coast was the worst case overtopping / inundation condition. Only under these conditions were water levels of sufficient elevation to allow waves to reach the shoreline and have a significant erosional effect.

Similarly to the waves and water-level assessments, the joint probability assessment was undertaken on the Point Nepean waves, with the SWAN model results used to transform the conditions to a variable climate along the open coast. The transformed joint probability wave-height values are summarised in Table 4-8. There were 30 open coast wave extraction points and 60 Port Phillip Bay wave extraction points established to produce the variable wave climate. The raw values for all wave points are presented in Appendix B.

Table 4-6     Joint probability conditions (worst case wave impact) plus sensitivity values

Table 4-7    Joint probability conditions (worst case inundation / overtopping) plus sensitivity values

To determine variable storm-tide levels for Barwon and Point Lonsdale (Rip Bank), the differences in tidal ranges along the open coast were applied. The storm-tide levels (STL) within the study area are presented in Table 4-8.

For input into the models, tidal curves and wave curves were required for the design-storm period duration of 4 days. This was based on the Anzac Day storm of 2009, which is the largest most recent storm for which measured records at multiple locations exist. The actual tidal cycle data for the two days before the storm and two days after was used for each location, with the magnitude of the peak of the storm increased to the maximum STL and wave height (as a range for each area) values presented in Table 4-8. The design storm for the erosion assessment was this storm applied twice, to simulate the effects of two design-storms.

Table 4-8     1% AEP event wave and storm-tide joint probability conditions for the study area