Current measurements at 3 sections and numerical experiments were carried out in Masan Bay to understand the effect of the wind on the residual currents and pollutant transport. The vertical distribution of horizontal velocities were directly affected by the wind at the depths of 1m and 2m beneath the sea surface. Analysis of the velocity data suggested that changes in the vertical gravitational circulation contributed to the net circulation. The net transport of water through the northern part of the bay was observed to be landward, with wind-induced transport of about 100m super(3)/s. Hence, wind is concluded to be the dominant mechanism driving the net circulation in the northern area of Masan Bay. Numerical experiments are shown that when S wind with 5m/s blew, northern area of the bay was generated the horizontal circulation of clockwise and local gyre. On the contrary of those, N wind made her to the anti-clockwise. In the case of no wind, the tidal residual current(constant flow) is very small or neglected except the bay-mouth. The inflow or outflow pattern of the mouth is considered as the flows generated by tidal residual current only. The distance of wind-induced transport of pollutant was as long as 2 times of no wind during the one tidal period.

In order to grasp physical characteristics of Ganajin Bay receiving the effluents from land, a comprehensive field surve)'has been conducted at the north of Wando and Gogeumdo in 2002. Water temperature was most high in September while most low in December. A high temperature in September was inferred to have caused by the dispersion of warm fresh water with favor of a dominant wind in autumn. However, salinity and sigma-t in situ was most low in September while most high in December. A low salinity (or density) in September turned out to reflect the influence of a dense rainfall in summer. Water temperature, salinity and density at the surface layer were lower than those at the bottom layer, except for December. Their horizontal profiles suggested the influence of effluents such as Tamjin River. Particularly, time series of water temperature acquired near the sluice and at the north channel of Wando tended to rise at the flood flow but fall at the ebb flow in accordance with the tide. The form ratio of the tide in the study area was 0.31~0.32 and the amplitude of the tide appeared to increase towards the west. Northeastward or southwestward flows prevailed in this area but the residual flows were all northeastward with a magnitude of 3~4cm/s.

The dicriminant function was introduced to understand the cause and establish the prediction method of red tides occurring in Jinhae Bay, Korea. Two sea regions of Masan and Haengam Bays and Dangdong and Wonmun Bays had different types of causes and patterns for red tides. In Masan and Haengam Bays, the red tides concentrically occurred during June and September. For example, in June the red tides occurred from physical and meteorological factors, which are related to the stratification and the increase in planktons. However, in August the red tides occurred from the water quality environment, based on these conditoins. Futhermore, in September the red tides were caused by the balance between the meteorological and water quality environmental factors. In contrast to those, in Dangdong and Wonmun Bays, the red tides mainly occurred during July and October and the frequency of occurrence was not as much as Masan and Haengam Bays. Especially, in August and September most meteorological and physical factors or water quality environmental factors appeared to contribute to the occurrence of red tides. This indicates that red tides do not easily occur as they are controlled by various environmental factors particularly in these regions. The discriminant functions were applied to predict red tides which they were actually occurred in Masan and Haengam Bays in June. The results showed that they were successful for the prediction of red tide at Haengam Bay but not at Masan Bay. The reason for their discrepancy in Masan Bay could have come from using a slight higher value of pH or COD in May, instead of its value in June.