This chapter investigates the impacts of sea-level rise on port hydrodynamics and navigational safety. In response to the growing risks posed by climate-induced sea-level change, three time-series analysis techniques—Simple Moving Average (SMA), Empirical Mode Decomposition (EMD), and Ensemble Empirical Mode Decomposition (EEMD)—were applied to water level data. These analytical results were integrated into a wave-current coupled hydrodynamic model to simulate tidal current patterns under different sea-level scenarios, and the model outputs were validated with field observations. The results show a consistent trend of slight increases in current velocity as the sea level rises, while the overall flow direction remains relatively stable. These findings highlight the importance of anticipating subtle hydrodynamic changes and supporting adaptive harbor planning, pilotage strategy, and maritime safety management under evolving oceanographic conditions.

Due to the rapid development of computers, researchers have made efforts since the 1990s to develop typhoon forecasting models and stochastic typhoon simulation models to assess typhoon disasters and risks. Typhoon forecasting models are primarily used to predict and track the movement of typhoons and provide warning information to the general public before landfall. Stochastic typhoon simulation models can assess extreme wind speeds and compensate for the limitations of current observations and simulation data length. Taiwan experiences approximately three to four typhoons yearly, of varying intensities and paths. Whether the marine meteorological data includes events of strong typhoon centers passing through will affect the results of frequency analysis. The development of offshore wind power in Taiwan is closely related to the unique marine meteorological conditions throughout the lifecycle stages, including wind farm site selection, feasibility studies, planning and design, construction and installation, operation and maintenance, and decommissioning. This study references relevant research and analyzes sixty-three scenarios using nine types of maximum storm wind speed radii and seven Holland-B parameters. The data from Japan Meteorological Agency Best Track Data (JMA BTD) is utilized, explicitly selecting 20 typhoon events after 2000 for wind speed simulation using a typhoon wind speed model. After validating the typhoon wind speeds with observation data from the Central Weather Bureau (CWB) in Hsinchu and the Longdong buoy, the technique of Monte Carlo simulation is utilized to generate synthetic typhoons randomly. The average of the relative absolute errors for the simulated maximum wind speeds is calculated, and through comprehensive evaluation, optimal parameter combinations ($R_m$, $B$) are obtained.

Considering practical and wide-ranging data acquisition, the development of image recognition technology can compensate for the shortcomings of traditional contact sampling methods. This study utilizes UAVs equipped with cameras to capture optical images of marine environments. It performs cluster analysis on the captured images using tricolor pixel values such as red (R), green (G), and blue (B) as clustering indicators. The research results demonstrate that in image recognition, the best image recognition can clearly distinguish various clusters of water bodies according to the colors corresponding to different turbidity levels, showing better recognition results compared to the highest value image recognition. Furthermore, by further calibrating with turbidity data measured by water quality meters, the analysis reveals a highly positive correlation between turbidity and chromaticity. This indicates that optical images captured by UAVs and classified by algorithms can quickly obtain the distribution characteristics of classified chromaticity groups

This study applies a coupled wave–hydrodynamic–water-quality model to quantify water exchange and BOD/DO dispersion in the Keelung Port–Tianliao River system. The model was validated using offshore Acoustic Doppler Current Profiler (ADCP) observations, showing consistent reproduction of waves and currents. Simulations were performed for normal inflow and for inflows with return periods of 5 years and 50 years. Two pumping mitigation options with the same discharge rate (90 m³/s) were evaluated at locations W16 (inner harbor/turning basin) and W18 (outer wharf). Results show limited dispersion and pollutant retention under normal inflow, while event inflows expand the impacted area and may reach the turning basin in the 50-year case. Pumping reduces BOD with clear spatial dependence: W16 improves the inner harbor/turning basin, whereas W18 benefits the outer wharf area. Under extreme inflow, mitigation effectiveness is constrained, suggesting the need for combined operational strategies.