The water quality is expressed as chemical oxygen demand (COD), whichhis calculated using the state quantities of phyto- plankton, zooplankton and detritus as follows;
COD = [COD:Cp]・ P + [COD:Cz]・Z + [COD:Cd]・D +QCOD (9)
where [COD:Cp], [COD:Cz] and [COD:Cd] are COD/ Carbone ratios of phytoplankton, zooplankton and detritus, respectively, and Qcod is COD load from the open water. The time histories of the calculated COD are shown in Fig. 3. The broken curves in these graphs represent the simulated results without considering the rock filter effect. The COD values with considering the rock filter effect are generally lower than those without considering the effect. When the rock filter effect is considered, the COD values in the inside (box 1,box 2) are lower than that in the outside (box 5, box 6), and slightly decreases with time even though it has seasonal fluctuations.
The time histories of the simulated nutrient concentrations are shown in Fig. 4. The nutrient concentrations at the surface layer in both inside and outside are lower than those at bottom layer. The values when the rock filter effect is considered also slightly decreases with seasonal fluctuations. These results suggest that the nutrients are effectively removed by the rock filter.
Fig. 3 Time histories of simulated COD
Fig. 4 Time histories of simulated nutrient concentrations
Fig. 5 Time histories of simulated DO
Figure 5 illustrates the time histories of the simulated DO. The values of DO at the surface layer increase in winter (espetially in the inside), when the rock filter effect is con- sidered. This means that the effectiveness of the breakwater on the reaeration is relatively large in winter.
Water purification process of a real artificial lagoon, where the water quality and the environmental conditions have been measured9), is also simulated using the proposed model. The artificial lagoon is located near the Kansai International Airport at Nishikinohania in Osaka Bay, and is surrounded by
