Predicting shoreline response to wave and sea level trends.

Abstract

In March 2007 the KwaZulu-Natal coastline was devastated by an extreme storm event. There is international concern that such events are associated with climate change. There is evidence of global changes in climate but there is still uncertainty as to whether they are anthropogenic or part of natural decadal (or longer) cycles. The increase in frequency and intensity of extreme storm events will impact on the sediment dynamics of coastlines and the associated risks need to be modelled and quantifed so that they can be included in coastal planning and management. Durban is a coastal city on the east coast of South Africa and has been used as a case study to identify trends in wave parameters and beach profile volumes. The correlation between profile erosion, waves and tides was explored using singular spectral analysis. The dependence between wave parameters was modelled using copulas. The decadal trends were introduced into these models using a nonstationary generalised extreme value distribution. Numerical models (SWAN, SBEACH, XBEACH) were used to transform the statistical model to near shore waves and estimate the associated erosion. The copula model was used to investigate the relationship between multivariate return periods and erosion return periods. Coastal defence options were reviewed and those appropriate for Durban were identifed. This study provides a review of Durban and Richards Bay's 18 years of Waverider data. It presents wave parameter exceedance statistics and wave height return periods for Durban. Durban's wave data showed increasing trends in maximum significant wave heights, peak wave period, storm event frequencies and a trend towards a more southerly mean wave direction. However, only the increase in peak period and wave direction was statistically significant. The trend in wave direction is considered a potential coastal hazard as it has the potential to increase the littoral drift by 1 % per annum. Durban's beach profiles have shown a long term erosion trend which is due to a combination of wave and sea level trends, and a reduction in sediment supply. The reduction in sediment supply from rivers was found to be both anthropogenic and natural. Storm, wave parameter and sea level trends were estimated to contribute more than 75 % to the total long term erosion. It was found that it takes an average of 2 years for a beach to recover to its pre-storm volume. Different types of coastlines recover at different rates and these recovery rates should be considered in risk assessments. A method for estimating future impacts due to storm and sea level trends has been proposed in the form of a non-stationary copula based statistical model. In general a bivariate return period of wave height and duration was found to approximate erosion return periods, while a method for estimating an analogous multivariate storm and erosion return period was developed. Geotextile sand filled containers were found to be a suitable coastal defence as they satisfy social, environmental and political pressure.