Quantitative Analysis of Water Quality for Controlling Lake Pollution ― Case study of two lakes in Japan and China ―

Eutrophication is a major water environmental problem in many countries. This study intends to introduce a basic methodology to assess water quality in natural water sources. Lake Sanaru in Japan and Lake Dianchi in China were used as case studies. We studied the pollution mechanisms in these two lakes by analyzing data from aqueous environmental samples and other related factors. Then, a simple and practical simulation model was proposed to express material balances around the two lakes of several constituents present in water. Material balance equations were used to calculate the internal productions of several constituents. Finally, purification methods and pollution loading for water quality targets were discussed.


Introduction
The water qualities of Lake Sanaru and Lake Dianchi have had the worst COD (Chemical Oxygen Demand) values for their respective countries over the past several years.Lake Sanaru has a complicated water flow system since it is a brackish tidal lake affected by the back-flow water from a region downstream.Lake Dianchi has a simple system but has been subject to serious external pollution from rivers for a long time, and the internal pollution loads, which are mainly caused by the growth of phytoplankton, are extremely large.Both of them are recognized as to be cleaned and have distinct water quality targets.So their pollution loads need to be quantified for specific water improvement strategies to be developed.For the internal productions in the lake, there have been many previous researches aimed at measuring the nutrients that elute from the lake bottom sediment.However, it is difficult to measure the actual elution amount in the lake, which has a steady production of photosynthetic algae.In fact, along with algae carcasses subsiding in the lake bottom, a continuous pollution load is applied to the water body from the carcasses' decomposition.This decomposition is dependent on the temperature and DO (Dissolved Oxygen) of the lake bottom.If the overlying water is aerobic, the elution COD is easily suppressed and decomposed.In order to determine the internal production rate, firstly a model of the flow condition of the water body should be constructed, and then a material balance simulation for the target component should be made.By substituting the observed data quality, simple and practicable calculations could be performed.

The area description
Lake Sanaru is located in the western area of Hamamatsu City, Shizuoka Prefecture, Japan.The average water depth is about 2 m (the maximum depth is 2.5 m).The lake area is 1.2 km 2 .The Environmental deterioration began in 1960, which was caused by urbanization in the lake region.Water pollution increased rapidly in 1960s.Not only did a high concentration of nitrogen and phosphorus flow into the lake from upstream, but additionally, Lake Sanaru is a brackish tidal lake affected by back-flow water from a river 13 km downstream.The lake map is shown in Fig. 1.The Total-COD and Sol (Soluble)-COD concentrations are measured at the same time once a month as a water quality index.Fig. 2 shows the time course of annual average COD in Lake Sanaru.Over the years in Lake Sanaru, a dredging of the lake bottom and a renovation project of the river have been made as efforts for water quality improvement.It is contemplated that these projects can affect the annual fluctuation of COD with regard to the water quality.

The variation of water quality
The correlations between SS (Suspended Solid)-COD with Chl.a (Chlorophyll-a) during the years of 2001-2014 are shown in Fig. 3. SS-COD is calculated by subtracting the Sol-COD from Total-COD.The variation is large but certain correlations were seen in annual average values.In addition, there is a roughly proportional relationship between COD and Chl.a.One of the reasons is the dominant type of algae changes every season, so the ratio of COD / Chl.a is different.Fig. 4 shows the transition of SS-COD concentration in the lake from 2005 to 2014 for each month.
It is impossible to view a particular seasonal trend according to different years change.However, looking at the monthly average line, from late autumn to early winter, during which there is little sunshine, SS-COD is small.Secondly, the seasonal change of water temperature in Lake Sanaru is shown in Fig. 5. From these two figures we can see, SS-COD didn't decrease conspicuously even in the lowest temperature period from January to March; on the contrary, average SS-COD shows the greatest value in March.So the generation rate of algae seems not to have been dominated strongly by temperature.Some phytoplanktons prefer cold-water temperatures.In summer also phytoplankton shows solid growth but the amount is not the greatest.

The lake model and simulation
Considering Lake Sanaru as a brackish lake, a simulation model of the mass balance in Lake Sanaru region was constructed, as shown in Fig. 7.According to Eqs. ( 1) and ( 2), firstly the coefficients F1, F2, which represent the backflow effect of seawater on the material exchange between the center of the lake (water system X1), downstream Shinkawa River (water system X2), and Lake Hamana (water system X3) were decided upon using all the data of chloride ion concentration observed since 2001 to 2014 [1].For Cl -, the internal production rate is zero.So the coefficients F1 and F2 can be calculated in a stable state.The internal production rates of rx (COD), rx (TN:Total Nitrogen) and rx (TP:Total Phosphorus), expressed as the net differences in the amount of these constituents from the inflow and outflow, were estimated by the simulation model using the calculated F1 and F2 values.It was indicated that the effect of the F1 value on the calculation results was not as large as expected, meaning that the calculated internal production rate will show little change even if the value of F1 fluctuates widely.It was also presumed that rx (COD) has a large variation in downstream river than in the lake (Fig. 8).The rx (TN) and rx (TP) also showed a similar variation.In Fig. 9, the removal of total nitrogen from the lake appeared to continue, because the evaluated rx (TN) values were always negative.This phenomenon can be attributed to the denitrification reaction from NO3-N.On the other hand, the calculated rx (TP) values kept steady inclinations (Fig. 10).According to the simulation, we confirmed the tide has little effect on water quality because when F=0, there is no large difference between the two curves in the lake.(1) (2) Where the symbol explanation are shown as below: V1: the volume of Lake Sanaru (m 3 ), V2: the volume of Shinkawa River (from lake to Kuryou River)(m 3 ), F1: the coefficient of substance exchange in Lake Sanaru and Shinkawa River (m 3 /day), F2: the coefficient of substance exchange in Shinkawa River and Lake Hamana (m 3 /day), Roi: average flux in river i (m 3 /day), Xoi: the concentration of nutrient x in river i (kg/m), Roj : average flux in river j (m 3 /day), Xoj : the concentration of nutrient x in river j (kg/m).R1: outflow from Lake Sanaru (m 3 /day), R2: outflow from Shinkawa (m 3 /day), X1: the nutrient concentration in Lake Sanaru (kg/m 3 ), X2: the nutrient concentration in Shinkawa River (kg/m 3 ), X3: the nutrient concentration in Lake Hamana (Enden)(kg/m 3 ), rx1: internal production rate in Lake Sanaru of the nutrient x, the increased amount in unit time (kg/day), rx2: internal production rate in Shinkawa River of the nutrient x, the increased amount in unit time (kg/day)

Estimation of purification measures
The purification measures applied to Lake Sanaru include sediment dredging, widening the drainage of the effluent river, trial of catalytic oxidation decomposition in the water, promotion of nitrogen removal by planting aquatic plants, and so on.Although, quantitative evaluations of the effects of these measures have not been performed, nevertheless COD represented in water quality has been improved gradually.The reason is the phosphorus concentration in inflow river water has been reduced annually.At present, it is considered that the speed of algae production is controlled by the supply of phosphorus from inflowing river water.This means that the load of phosphorus from the inflow rivers dominate the internal production rate of rx (COD) in the lake.Therefore, when the influx TP loads decrease continually, rx (COD) can be expected reduce in the future, too.
It is assumed that the value of the internal production rate of rx (COD) will reduce to rx = 151 (kg/day) by 2022.In this case, assuming the flow rate to be Roi = 45348 m 3 /day in a steady state (average value from 2001 to 2014), using calculations of the COD equation, the following result can be derived about the value of Cs: Cs = ( RoiCoi + rx ) / R = 4.3 mg/L However, in this case, the substance exchange capacity factor is assumed to be F = 0.The actual value may be slightly smaller than this value.In general, if the river inflow load of phosphorus continued to be reduced, the water quality target value of COD below5 mg /L in Lake Sanaru can be reached.

Lake Dianchi system
Lake Dianchi is located in Kunming, Yunnan Province of Southwestern China (Fig. 11).It is one of the Chinese government's key lake restoration priorities.The average depth is only 4.4 m at 1,887.4 m above sea level.The main portion of the lake area is 298 km 2 .Lake Dianchi is significantly affected by human activities as well as algal blooms breaking out almost every year.Fig. 12 illustrates the effect of these factors.Chl.a concentration is highest in summer, and has a separation phenomenon, Blue Greens remained as dominant spices almost all year around.The difference between it and Lake Sanaru is vast.In Dianchi, the dry season, wet season and level period are from February to May, June to September and October to January respectively.River runoff in Dianchi drainage changes dramatically according to the season.The climate of the drainage area is typically that of the monsoon region of low latitude and high latitude.As shown in Fig. 13, the microclimate of the region is very pleasant, averaging a temperature of 15℃ annually (summer: 23℃,winter: 8℃).Although Total-COD keeps the same value in different seasons (Fig. 14), we can infer SS-COD must change greatly in summer.And compared with Lake Sanaru, we can know that temperature is not the most important factor for phytoplankton growth.Precipitation seems to have more effect on algae growth and bloom.Waters coming from 6 treatment plants surround the lake, contributing high nitrogen and phosphorus loadings to Lake Dianchi together with numerous diffuse pollutants from the surrounding city, villages and agricultural surfaces.More than 30 rivers flow into the lake.Only one river flows out.The water system is simple.But limited by political and economic conditions, some data, like river flux, is difficult to acquire.So referring to some papers' data [2,3], we derived the average rate of internal production in Lake Dianchi shown in Table 1.The simulation formula is the same as shown in 2.3.

Conclusion
1) Lake Sanaru was affected by back flow water from downstream region, but the nutrient load in the lake showed little fluctuation even if the water flow amount changed greatly.The Shinkawa River helped to decrease the influence of pollution, which came from downstream, and its influence on the lake could be ignored.Therefore, water pollution management should focus on the pollution load of upstream rivers and internal production decrease, such as improving the sewer system and reducing phytoplankton in the lake directly.Using this simulation model of Lake Sanaru, if the internal production rate of COD decreases to half that of the current rate, the COD concentration can except to reach 5 mg/L as a water quality target, and will require 7 years.
2) In Lake Dianchi, the external pollution from poorly managed human activities and internal pollution load resulted in the combined effect of eutrophication.The discharge of pollutants into the lake in quantities seemed so far to exceed the natural self-purification capacity of the lake.The internal production rate was huge, which was the main reason for the algal blooms in the lake.If the current hydraulic condition is maintained, even if the external pollution could be controlled, the internal pollution load would affect water quality for at least 30 years.Water temperature is not an important factor for phytoplankton growth.Precipitation seems to have more effect on algae growth and bloom.

Fig. 13 .
Fig.13.Monthly change of average precipitation and temprature in Lake Dianchi Fig.14.Monthly change of Tol-COD concentration in Lake Dianchi

Table I .
Material balance characteristics of Lake Dianchi