3Professor, Dept. of Civil Engineering, Kanazawa University

40-20 Kodatsuno 2chome, Kanazawa Ishikawa 920-8667, Japan, Tel: 81-76-234-4607,

E-mail: hishida@t.kanazawa-u.ac.jp

Abstract: In this study, characterization of dissolved oxygen variation in Lake Yanaka is conducted with the aim of elucidating the mechanism behind the severe dissolved oxygen depletion observed in the hypolimnion of the lake. The analyses suggest that the DO concentration in the bottom layer is well correlated with the five-day average value of solar radiation, and water intake could significantly influence the DO resource of the lake in two ways. As the thermal stratification develops, the DO concentration in the bottom layer is further deteriorated due to high values of sediment oxygen demand (SOD) and algal respiration, which are estimated to be 2.64 gO₂m^-2day^-1 and 4.8 gO₂m-³day^-1, respectively. Finally, a DO improvement measure is assessed.

 

Keywords: Lake Yanaka, dissolved oxygen depletion, water intake, thermal stratification, sediment oxygen demand, algal respiration, diurnal variation

The concentration of dissolved oxygen (DO) is a significant index in the evaluation of water environment and in the management of water quality. Low DO in hypolimnetic waters of lakes or reservoirs can reduce the usefulness of these water bodies as well as harm receiving streams. Its impact may be reflected in a number of ways such as fish mortality, odors, and aesthetic nuisances. Oxygen enters the water by absorption directly from the atmosphere or plant photosynthesis. It is removed by respiration of organisms and by organic decomposition. Factors affecting the amount of DO in waters in addition to water temperature, include light condition, meteorological phenomena, nutrient cycling, algal effects, benthic oxygen consumption, flow condition and so on. Many of the factors and their interrelationships have not been fully understood. Therefore, a better understanding of the environmental factors that cause dissolved oxygen concentrations to vary is essential to defining strategies to satisfy environmental standards for dissolved oxygen. The major sources of DO are reaeration from atmosphere, algal photosynthesis and inflow loading. The sinks include algal respiration, carbonaceous biochemical oxygen demand (CBOD), nitrogenous biochemical oxygen demand (NBOD), and sediment oxygen demand (SOD), and outflow transport. The present study is aimed at gaining in-depth insights into the character of the DO depletion in Lake Yanaka.

Lake Yanaka is part of the Watarase retarding basin. It is the first multi-purpose impoundment lake constructed in alluvial plain in Japan. The lake has a surface area of 4.5 km² and an average depth of six meters with seasonal changes of about three meters for flood control. It is divided into three blocks by levees and connected by gaps as depicted in Fig.1. Inflow and outflow are regulated at the pumping station located at the south tip of the lake. Till a year ago, approximately 80% of the annual inflow was taken from the Yata River (on the left), and the rest from the Watarase River (on the right).

In recent years, the eutrophication problem has surfaced up in Lake Yanaka. High inputs of nutrients led to excessive growth of phytoplankton. The concentration of chlorophyll a ranges from 50 μg/l to more than 250 μg/l. Huang and Tamai³⁾ investigated the mechanisms of phytoplankton blooms in the lake. It was suggested the high inflow load and bottom release of nutrients were responsible for the spring and autumn blooms respectively. Amitabu, Huang and Tamai¹⁾ did time series analysis to detect the trend of nutrient concentration of the lake. Nevertheless, the dissolved oxygen of the lake has not been studied comprehensively.

The characterization of DO variations is based on filed data described below:

• Continuous data obtained from the automatic monitoring station installed at the middle point of South Block. It records hourly data at two different depths, namely 50cm below the surface and 1m above the bottom. Measured Quantities are water temperature, EC, DO, turbidity and pH.
• Periodic data measured at one point in each of the three blocks at two depths usually, and more than two depths from time to time. The measured items include water level, water temperature, transparency, pH, DO, BOD, COD, SS, turbidity, EC, TP, PO₄-P, TN, NH₄-N, NO₂-N, NO₃-N, chlorophyll a, so forth.
• Hourly meteorological data such as air temperature, wind direction and speed were collected from a nearby meteorological station.

Fig.2 shows the periodically measured DO values both near the water surface and lake bottom in the South Blocks for 1996. It can be seen clearly that an anoxic hypolimnion developed in early summer, around the beginning of June. The difference in DO between the upper layer and bottom layer is also appreciable in August and November, but it is much smaller in magnitude as compared to the case in June. And the other two blocks exhibited quite similar patterns through the year.

Fig.3 demonstrates the temporal variation of the vertical temperature profile. A thermal stratification was developed from late May to early June. Figure 4 shows the vertical distributions of DO during the early summer of 1996. By comparing Fig.3 with Fig.4, it can be seen that as the stratification proceeded, the DO in the hypoliminion was quickly depleted. The anoxic period lasted for about two weeks. For the purpose of illustrating the connection between meteorological condition and DO variation, the five-day average of net solar radiation at the lake site during the stratified period is presented in Fig.5, together with the near-bottom DO and the difference in water temperature between the surface and bottom. It is clear that the DO is of opposite phase to the solar radiation, while the water temperature difference is in phase with the radiation. The correlation coefficient between solar radiation and near-bottom DO is found to be 0.79, and the linear regression gives:

DO| _1m = -0.443×Rad.+9.867

The daily mean values of DO 1m above the bottom (denoted as DMB DO) and 0.5m below the surface (denoted as DMS DO) for the latter half of May are shown in Fig.6, together with the storage variation. It indicates that as the storage increased, both DMB DO and DMS DO decreased at the first. Then with the onset of stagnation, DMB DO further deteriorated, but DMS DO picked up. Statistical analysis reveals that the temperature difference between the surface and bottom layers is correlated with the storage variation (r²=0.69) during this period of time. The storage increase was mainly due to water intake for flood control, which carried a lot of ammonia nitrogen in May ⁴⁾. As mentioned in sec.1, NBOD is a sink of DO, to see if nitrification events have something to do with the DO depletion in Lake Yanaka, temporal changes of ammonia in the lake were investigated, and it is shown in Fig.7. Ammonia concentration was high in the middle of May, then decreased rapidly, and remained more or less constant since the onset of anoxia on May 21. Meanwhile, the concentration of chlorophyll a was found to be under steady decline from the beginning of May to May 21⁴⁾, so the loss of ammonia during this period could not be attributed to algal utilization, but to nitrification. Since the stoichiometric oxygen equivalent for ammonia oxidation is 4.57 gOgN-1, the reduction of ammonia from 0.42 mg/l to 0.12 mg/l may consume oxygen by 1.37 mg/l.

Based on these results, the following may be stated. Oxygen demand exerted by carbonaceous and nitrogenous matters from inflow loading caused a reduction in DO level in the first half of May. With the increased water depth due to the water intake, and high level of solar radiation, the vertical temperature gradients were established quickly in the second half of May. The stratification limited the exchange of dissolved oxygen between epilimnion and hypolimnion that allowed the DO in the bottom layer to be depleted by sediment oxygen demand and algal respiration. Considering the fact that oxidation is an aerobic process, it might be reasonable to assume that CBOD and NBOD are close to cease following the onset of anoxia.

The loss of oxygen due to the stabilization of organic materials deposited at lake bottom is termed “Sediment Oxygen Demand”. The mass balance equation for DO in the bottom layer can be written as:

Here SOD and R stand for sediment oxygen demand and algal respiration, respectively. Under the anoxic condition during the period of stratification, the oxidation and nitrification may be assumed to be inactive, so that the DO mass balance equation for the bottom layer of fixed thickness may be further simplified as:

Here, h is the thickness of the bottom layer, given to be 1m in this study. This equation indicates that the volumetric oxygen demand R+SOD/h can be estimated from the slope of DO variation curve if the variation curve does not deviate significantly from the classical pattern ²⁾. Since the DO variation 1m above the bottom on May 23 is found to follow the classical pattern of variation, we use the night data of May 23 to estimate the volumetric oxygen demand. By linear fitting to the night portion of the measured DO curve on May 23, (R+SOD/h) per hour is obtained to be 0.31. The daily mean value of R+SOD/h is then 7.44 gO₂/m³/day.

Next, we try to estimate the algal respiration by examining the DO balance in the surface layer. Because the wind was almost absent before predawn on May 23, so that the reaeration could be expected to cease around this time. Consequently, one may assume that the respiration be the dominant term in DO balance in the surface layer in predawn hours of May 23. Then, from the predawn measurement, the algal respiration R is estimated from the DO balance to be 4.8 gO₂/m³/day. Correspondingly, SOD=2.64 gO₂/m²/day.

To improve the DO level near the bottom, a remedial measure is under review by the management authority of the lake. The idea is to increase the transfer of DO from the surface layer to the bottom layer by circulating the lake water. That is to pump the water out of the lake at the south gate, channel the flow to the north gate, then pump it back to the lake. To evaluate the effectiveness of the proposed circulation scheme, numerical simulations of DO response to flow circulation are performed with a two-layer mass balance approach which has been widely used in modeling efforts for lakes or reservoirs (for details, see 4), 5) ). In the simulations, an anoxic hypolimnion is given as initial state, solar radiation, wind speed, water temperature, phytoplankton, ammonia and BOD concentrations are specified according to filed data. SOD and circulation flow rate are varied to illustrate possible DO response to different scenarios. Fig. 8 shows the daily mean DO variations in the lower layer of the south block with different flow rates when the value of SOD is given to be 2.64 gO₂/m²/day. It demonstrates that both circulation rates may improve DO in the hypolimnion, and the percentage of increase in DO is 150% and 273%, respectively. However, if the value of SOD is raised to be 4.5 gO₂/m²/day, it is found that the two circulation rates could not bring about significant improvement.

The dissolved oxygen in Lake Yanaka was depleted in the early summer of 1996. The water intake played two roles in the process. Firstly, the inflow carried a large amount of oxidizable material into the lake that caused a reduction of DO by the mechanism of oxidation. Secondly, it increased the water depth. The deepened lake under high incoming solar radiation, facilitated the formation of thermal stratification that limited the DO transfer from the epilimnion to hypolimnion. Under such a circumstance, the dissolved oxygen in the hypolimnion was consumed by sediment oxygen demand and algal respiration, which are estimated by a simple mass balance approach to be 2.64 gO₂m^-2day^-1 and 4.8 gO₂m-³day^-1, respectively during the anoxia.

A circulation scheme, which has been under consideration by the lake’s authority, is evaluated in this study. Preliminary results indicate that the effectiveness of this scheme is dependant to a large extent on the level of SOD.

This study is supported by the Upper Tone River Working Office, Ministry of Construction of Japan.

[1]  Amitabu, R., Huang, G. and Tamai, N.: Lake Yanaka and its deteriorating water quality, Proceedings of the first International Summer Symposium, JSCE, 349-352, 1999.

[2]  Gunnerson, T. and Bailey, T.E.: Oxygen relationships in the Sacramento River, J. of the Sanitary Engineering Div., ASCE, Vol. 89, No.4, 95-124, 1963.

[3]  Huang, G. and Tamai, N.: Limnological studies in Lake Yanaka, Annual J. of Hydraulic Engineering, JSCE, Vol. 44, 1107-1112, 2000.

[4]  Huang, G. and Ishida H.: Application of WASP5 to Lake Yanaka - a modeling study, submitted to Int. J. of Lake and Reservoirs Management.

[5]  Thomann, R.V. and Mueller, J.A.: Principles of surface water quality modeling and control, Harper & Row Publisher, NY, 1987.

Fig. 1   Lake Yanaka

Fig. 2   DO variation in South Block in 1996

Fig. 3   Vertical temperature profiles

Fig. 4  Vertical profiles of DO

Fig. 5   Bottom DO, radiation,temperature difference storage

Fig. 6   Daily mean DO in the surface andbottom layers

Fig. 7   Depth-averaged concentrationof ammonia in the lake

Fig. 8   Response of  hypolimnion  DO,SOD=2.64