No abstract

The northern South China Sea (SCS) including the Luzon Strait, from around 19º to 22ºN latitude and from 114º to 122ºE longitude, is an ocean area where energetic internal waves occur frequently. In this study, satellite ocean color products from SeaWiFS (Sea-viewing Wide Field-of-view Sensor) and MODIS (Moderate Resolution Imaging Spectroradiometer) including true color and chlorophyll images are used for the statistical analysis of internal wave occurrence. For spatial distribution, fewer internal waves distributed in the east of 118ºE obviously originate from the Luzon Strait, and more internal waves in the west of 118ºE may propagate from the east or evolve into solitons originating from the eastern boundary owing to the fission effect of the shoaling thermocline. No internal wave is found east of the Luzon Strait. The lunar daily observed internal wave occurrence frequencies show that more internal waves are found after full moon and new moon, that is the spring tide. This indicates that the generation of internal waves in the northern SCS may relate to the internal tide. The monthly distribution of internal wave occurrence frequencies reveals that the high frequencies are distributed from May to August and reach a peak in July with a maximum frequency of 21.5%. The low occurrence frequencies are found in winter from November to February with a minimum frequency of 0.5% in January.

Multi-sensor data from different satellites were analysed to study the interannual variability of the cold domes northeast of Taiwan. From the empirical orthogonal function analysis of along-track satellite altimeter data and sea surface temperature, both revealed a cold dome phenomenon located at approximately 122°36´E and 25°24´N, and their principal components demonstrated a variation of 4-6 years. The cold dome variation correlates best with the Oceanic Niño index and poorly with Pacific Decadal Oscillation index. The forcing could be wind stress curl and Kuroshio volume transport, which produce an upwelling around the Men-Hua Canyon. Positive wind stress curl and the on-shelf water transport of the Kuroshio Current through the North Men-Hua Canyon increase the chance of the cold dome formation which may be responsible for the interannual variability of cold domes. ARTICLE HISTORY

Synthetic Aperture Radar (SAR) images from European Remote Sensing (ERS) satellites are used to investigate oil spill from ship navigations in the water adjacent to Taiwan. A total number of 136 images taken from 1993 to 1997 are used in this study. On the 136 images, only 46 images showing the possibility of oil spill which are based on the position and the shape of the discharge, the path of the ship, the sea characteristics of the area, and the weather conditions. The result shows that oil spill occurs most frequently in spring and least in winter. The sea area off eastern Taiwan has a probability which far surpassed other areas, followed by the middle sea area, the northern sea area, and the southern sea area.Regarding the oil spills at different areas with the distance to the shore, the oil spills at the middle area, with an average distance of 50 km (28 nautical miles), is closer than those at other areas. The statistical analysis demonstrates that the oil spill around Taiwan mostly occurs over 44 km (24 nautical miles) away from shore. Therefore, it is obvious that the probability of oil spills occurring as a ship leaves or enters the harbor is not high. Instead, the majority of oil spills takes place from middle to long distance navigating fishing boats as well as from oil and cargo freighters navigating international waterways.

The multilayer perceptron (MLP) neural network have been widely used to fit non-linear transfer function and performed well. In this study, we use MLP to estimate chlorophyll-a concentrations from marine reflectance measures. The optical data were assembled from the Sea-viewing Wide Field-of-view Sensor (SeaWiFS) Bio-optical Algorithm Mini-workshop (SeaBAM). Most bio-optical algorithms use simple ratios of reflectance in blue and green bands or combinations of ratios as parameters for regression analysis. Regression analysis has limitations for nonlinear function. Neural network, however, have been shown better performance for nonlinear problems. The result showed that accuracy of chlorophyll-a concentration using MLP is much higher than that of regression method. Nevertheless, using all of the five bands as input can derive the best performance. The results showed that each band could carry some useful messages for ocean color remote sensing. Only using band ratio (OC2) or band switch (OC4) might lose some available information. By preprocessing reflectance data with the principle component analysis (PCA), MLP could derive much better accuracy than traditional methods. The result showed that the reflectance of all bands should not be ignored for deriving the chlorophyll-a concentration because each band carries different useful ocean color information.