No abstract

Predicting sea level is crucial since the rising of sea levels can cause flood, inundation and coastal erosion. This study investigates the potential of the chaotic model to forecast the future sea level in Malaysian Borneo (Sabah). The studied data is an hourly time series of water level collected at the particular benchmark location (station 74003) in the district of Sandakan, Sabah. This study employed 5136 time series data from June 2017 to November 2017. The research goals are to discover the existence of chaotic dynamics by reconstructing the multi-dimensional phase space and Cao method, and to predict future water level data using the local linear approximation approach. The outcome reveals that the correlation coefficient betwixt actual and forecasted data is 0.9228 which is close to 1. It thus reveals the reliability of the prediction method for forecasting the sea level time series data, as well as encouraging indication that the chaotic approach is appropriate to a time series of Malaysian sea level. In implication, these findings are expected to help agencies, particularly the Malaysian Department of Survey and Mapping (JUPEM), organise better sea-level management.