Wave parameter classification based on morphological changes around a small wave-dominated tidal-inlet using a schematized Delft3D model

Abstract

The investigation of long-term morphological changes of a tidal-inlet using process-based models is a complicated and resource dependant task. To date, a number of input reduction techniques have been utilised to reduce the complexity of the modelling tasks; amongst them the wave condition has been shown to be of significant importance. Rapid and significant morphological changes generally occur as a result of a highly energetic waves occurring over a short period, whilst low energy waves can only be influential if they occur for a longer period. Therefore, research has been focused on methods to categorize a continuous time-series of waves into discrete events which, in combination, have an almost identical morphological outcome as the full time-series. Nevertheless, there is no clear agreement on a preferred categorization method based on an overall frequency of occurrence of wave parameters. Additionally, for a re-arranged, categorized wave dataset, the selected chronology of the events is a matter of further consideration as research has shown that the morphological outcome would vary. Moreover, despite some noteworthy research for coastal areas, the current literature is quite limited in regard to morphological changes around small tidal-inlet entrances. The first of the above-mentioned gaps was considered in this study and as such this article aims solely to determine important classes of wave parameters based on their relative morphological impacts on a wave-dominated, small tidal-inlet entrance. The effects of chronology, as well as the cumulative influence of consecutive wave events are therefore not further considered here. For this study several categorization methods from the literature were investigated. The selected approach uses an offshore directional wave time-series to statistically classify wave parameters. Wave data is initially separated by wave period (e.g. Tp), to distinguish between seas and swell waves. Then, based on the frequency of the occurrence of particular wave height and direction ranges, data is classified into sub-groups. Later, the sub-groups with the combined total frequency percentage of occurrence of about 90% are identified to encompass the required classes of schematized wave parameters for modelling. The remaining sub-groups (i.e. with total 10% occurrence) are considered to be morphologically ineffective. Furthermore, for preparing the model boundary data, there needs to be representative wave parameters for each of the selected classes. For the representative wave direction and period, a simple algebraic average of data for each class could be used. However, the best practice to find the representative average wave height could be through averaging the energy of individual incident waves for each of the classes. All these classified parameters then are used as input for morphological model. For the case study of this research, the wave classification procedure resulted in 20 classes of wave data. An existing, calibrated, validated, coupled Delft3d Wave and Flow model was used along with a simple harmonic tide at the boundary, to perform the necessary morphological modellings. In this research, the inlet entrance, its channel and the flood shoal of the selected case study area were the main focus. The modelling results showed that the wave classes which have undergone less refraction (i.e. cases with shore-normal offshore wave direction) were more prominent in transport of sediment all around the inlet entrance; irrespective of the wave height and period. Conversely, waves with a more oblique angle of incidence showed less transport in to the inlet inner lagoon; even for very large wave heights. The influence of wave period was also shown to be noticeable, as extensive erosion/accretion events based on larger wavelength of swell in comparison to sea waves was existent. The variation of depth averaged velocity, erosion/accretion and wave parameters at selected locations revealed that locations inside the inlet lagoon and at the entrance channel were less sensitive to different wave conditions. In contrast, the observation point offshore of the entrance, directly followed the changes to wave parameters. By having an insight in to the importance of each individual class of wave data, the classification procedure could be re-assured. Moreover, a particular wave class could be statically important but morphologically ineffective. Therefore, the output could assist in selection of an appropriate chronology for a cumulative morphological study.