Water quality model with multiform of N/P transport and transformation in the Yangtze River Estuary*

Biao WANG （王彪）， Shi-qiang LU （盧士強(qiáng)）， Wei-qing LIN （林衛(wèi)青）， Yi-fan YANG （楊漪帆），Dao-zeng WANG （王道增）

1. Shanghai Academy of Environmental Sciences， Shanghai 200233， China， E-mail: wangbiao5134@126.com

2. Institute of Applied Mathematics and Mechanics and Shanghai Key Laboratory of Mechanics in Energy Engineering， Shanghai University， Shanghai 200072， China

?

Water quality model with multiform of N/P transport and transformation in the Yangtze River Estuary*

Biao WANG （王彪）1， Shi-qiang LU （盧士強(qiáng)）1， Wei-qing LIN （林衛(wèi)青）1， Yi-fan YANG （楊漪帆）1，Dao-zeng WANG （王道增）2

1. Shanghai Academy of Environmental Sciences， Shanghai 200233， China， E-mail: wangbiao5134@126.com

2. Institute of Applied Mathematics and Mechanics and Shanghai Key Laboratory of Mechanics in Energy Engineering， Shanghai University， Shanghai 200072， China

As the Yangtze River Estuary and adjacent sea have been classified as a problem area with regard to eutrophication， it is important to explore the spatial and temporal variations of nitrogen and phosphorus （N/P） nutrients in this area. Based on danish hydraulic institute （DHI）’s open platform Ecolab， a hydrodynamic and water quality model was developed for the Yangtze River Estuary， in which the transport and transformation processes of different forms of N/P nutrients were considered. Validations against measured data show that the model is overall reliable. Preliminary application of the model suggests that the model can simulate the characteristics of high phosphorus concentration area in the Yangtze River Estuary， and the high concentration area is closely related to the resuspension process of particulate phosphorus.

water quality model， Yangtze River Estuary， N/P transport and transformation

Introduction

Over the past decades， human activities have significantly accelerated the supply of land-derived nutrients to water. Consequently， eutrophication has become a major environmental problem in estuary and coastal water throughout the world［1，2］. For instance，the estuary of the Yangtze River， the third longest river in the world， is becoming one of the most seriously polluted areas in China and its eutrophication problem has drawn more and more attentions in recent years［3-6］.It is of great importance to understand the nutrient dynamics in the Yangtze River Estuary for better environment management and protection.

Numerical modeling is an effective approach in studying water quality problems and has been applied to estuaries worldwide［7-11］. In the Yangtze River Estuary， several researchers have used numerical models to study water quality problems. For example，Ko developed a hydrodynamic and water quality model based on the SMS model to simulate sewage transportation and diffusion［12］. Li［13］and Liu et al.［14］both developed water quality models with Delft3D to examine the pollution concentration distribution caused by the sewage discharge during a dry season and flood season. Li investigated the influence of a chemical leakage accident on the water environment based on the MIKE21 model［15］. However， water quality processes in the above-mentioned models are mainly based on a single convection-diffusion equation， which is not adequate for much more complex estuary systems.

In many estuaries， N/P nutrients are present in several forms （e.g.， dissolved and particulate， organic and inorganic）， and their composition structure is changing by human activities， which can affect estuary and coastal water［16］. N/P nutrients can be transformed between different forms through biogeochemical pro-cesses［17］that are difficult to simulate using the traditional convection-diffusion water quality models. It is necessary to develop a new model that includes multiform of N/P nutrients and multiple underlying processes to simulate their complex transformations.

Based on DHI’s Ecolab open platform， we developed a hydrodynamic and water quality model for the Yangtze River Estuary. The model includes the transport and transformation processes of different forms of N/P nutrient. The model was calibrated and validated against the measured data， and then applied in the preliminary study of the phosphorus spatial distribution in the Yangtze River Estuary.

1. Numerical model

1.1 Hydrodynamic model

DHI’s MIKE， as a commercial engineering software package with good stability， efficiency and accuracy， is widely used throughout the world［18，19］. The DHI’s MIKE3FM was used in the hydrodynamic and water quality model as its hydrodynamic module. The unstructured grid of MIKE3FM enables the model to fit the complex terrain systems such as the Yangtze River Estuary with multilevel branching.

Hydrodynamic equations of the model are as follows:

1.2 Water quality model

A water quality model including N/P nutrient transport and transformation was developed based on the DHI’s Ecolab open platform. To simulate the complex N/P nutrient cycles， the model developed in this study contains the variables corresponding to different forms of N/P such as ammonia （NH3-N），nitrate， nitrite （NOx-N）， dissolved organic nitrogen（DON）， particulate organic nitrogen （PON）， dissolved inorganic phosphorus （DIP）， dissolved organic phosphorus （DOP）， particulate inorganic phosphorus （PIP）and particulate organic phosphorus （POP）.

Fig.1 Main processes of N/P nutrient cycles considered in the model

For the Nitrogen cycle， the model includes not only the convection-diffusion process but also settlement and resuspension， hydrolysis and mineralization，nitrification and denitrification processes， and plant metabolic process （Fig.1）. For the phosphorus cycle，the model contains adsorption-desorption process and the processes similar to the N cycle.

Fig.2 Unstructured model grid with resolution of0.5kmin the Yangtze River Estuary and30kmat the open sea boundary

1.3 Model domain and grid

The model domain and grid were shown in Fig.2. Three major rivers， the Yangtze， Huangpu and Qiantang， were included in the model. The upstream boundary of the Yangtze River in the model was set at the Datong Hydrological Station，upstream of the estuary and the east open sea boundary wasaway from the coastline. Model grid resolution was about 30 km near the open sea boundary and about 0.5 km in the Yangtze River Estuary.

1.4 Model boundary and pollutant sources

For the hydrodynamic boundary conditions， river discharge was set at the river boundary， and water level calculated from 16 tidal constituents was given at the open sea boundary. For the water quality boundary， pollutant concentration at the river boundary was determined by the river pollutant load， while at the open sea boundary it was set according to measured data.

In addition to the Yangtze， Huangpu， Qiantang rivers， the model included other pollutant sources along the estuary coastline in the form of point sources， such as small rivers （e.g.， the Liu， Baimao） and sewage outfalls （e.g.， Shidongkou， Bailonggang，Zhuyuan sewage plants）. The pollutant load and nutrient composition of these point sources were determined from the literatures and the field data. Approximately， the chemical oxygen demand （COD） emission from all the point sources was 359 000 t/a. Nutrient emissions were 46 000 t/a （NH3-N）， 92 000 t/a （total nitrogen， TN） and 14 000 t/a （total phosphorus， TP），respectively.

1.5 Model parameters

Based on the literatures and the model calibration，the values of the main parameters for the N/P nutrient transformation processes in the model were determined and given in Table 1.

Table 1 Values of major parameters in the N/P transformation processes

Fig.3 A map of the Yangtze River Estuary showing the sampling locations

2. Model verification

2.1 Hydrodynamic verification

Fig.4 Comparison of the calculated tidal levels with the measured data in the Yangtze River Estuary

2.2 Water quality verification

Water quality data along the Yangtze River Estuary in March and October 2012 were collected for the comparison with the model results. The locations of the water quality stations （WS1-WS6） were shown as stars in Fig.3. The modeled NH3-N， TN and TPconcentrations were generally in good accordance with the measured data （Fig.5， whereis the concentraction）. The deviation of the calculated NH3-N concentration from the measured data varied between 0.002 mg/L and 0.521 mg/L， with a mean of 0.088 mg/L. The deviation varied between 0.001mg/L and 0.813 mg/L （mean: 0.120 mg/L） for TN， and between 0.001 mg/L and 0.041 mg/L （mean: 0.016 mg/L） for TP， respectively. The average relative error for NH3-N was 31.9 % due to relatively large errors at WS4 and WS5 during October， which decreased to 15.1% if those two values were excluded. The average relative error for TN and TP were 7.0% and 7.7%， respectively. In general， the model performed well in simulation of nutrient concentrations in the Yangtze River Estuary.

Fig.5 Comparison of the modeled water quality results with the measured datain 2012 in the Yangtze River Estuary

2.3 Composition proportion verification

The composition data of N/P nutrients were collected at CS1， CS2 and CS3 stations （shown as triangles in Fig.3） in the Yangtze River Estuary to examine the precision of the model in simulating the proportion the N/P nutrient composition. The comparison of the calculated nutrient composition proportion with the measured one was shown in Fig.6. The calculated NH3-N proportion was higher while the NOx-N proportion was lower than those of the measured data. This might be caused partly by the uncertainty in the pollutant sources given in the model. It also indicates that the parameters used in the nitrogen cycle should be further optimized. The calculated proportions of DP and DIP were quite close to the measured ones. In addition， the spatial variation of N/P composition proportion among the three stations were well simulated by the model （i.e.， decreasing downstream along the estuaryfor NOx-N， DIP and DP and maximizing in the middle （CS2） of the estuary for NH3-N）. In general， the model calculated N/P composition proportions were in agreement with the measured data. In average， the deviation of the calculated composition proportion was about 4.5% for NH3-N，17.6% for NOx-N， 4.1% for DIP and 7.1% for DP，respectively.

Fig.6 Comparison of the calculated composition proportion of N/P nutrients with the measured data in the Yangtze River Estuary

3. Model preliminary application

3.1 Simulation cases

Compared to nitrogen nutrients， the phosphorus nutrient pollution draws more attention in the water of the Yangtze River Estuary. To examine the difference between the model and the traditional advection-diffusion model， two scenarios were set up to simulate the spatial distribution of phosphorus nutrients in the Yangtze River Estuary （Case 1 and Case 2）. Case 1 was carried out on the new model and Case 2 was on the traditional one. Considering the important roles of the particulate phosphorus （PP）， an additional scenario（Case 3） was set up with the new model but with a special treatment so that the resuspension process of PP was not considered in the model. All three cases were performed under the same condition of hydrodynamics and water quality， and their only difference was the choices of the models.

3.2 Results

In each scenario， the model was run for 60 model days， and the results of the last 15 d were averaged and used for further analysis. To save space， the analysis focused on TP. The spatial distributions of TP from Case 1 to Case 3 were shown in Fig.7 to Fig.9，respectively.

Fig.7 Averaged TP concentration distribution in the Yangtze River Estuary from Case 1

Fig.8 Averaged TP concentration distribution in the Yangtze River Estuary from Case 2

The TP concentrations in Case 1 （with the maximum of） were higher than those in Case 2 （with the maximum of） in general（Figs.7， 8）. Case 1 shows a low-high-low variation of TP concentrations along the Yangtze River Estuary，with the high concentration zone at the mouth of the estuary， while Case 2 exhibits a decrease downstream the estuary except a narrow zone of the high concen-tration along the coast outside the Huangpu River mouth， which may be caused by the input of the highly polluted water from the Huangpu River. For the mouth area of the Yangtze River Estuary， the TP concentrations in Case 1 were about 0.03 mg/L-0.06 mg/L higher than those in Case 2. A relatively higher concentration of TP toward the mouth of the estuary was detected in the water quality data in March and October 2012. In addition， Li et al.［21］pointed out that the higher concentration zone was located outside the mouth of the Yangtze River Estuary. These observations confirm that the calculated TP distribution result from the new model （Case 1） was consistent with the data.

Fig.9 Averaged TP concentration distribution in the Yangtze River Estuary from Case 3

As the resuspension was not included， the TP concentrations in Case 3 were lower than those in Case 1 and Case 2， with the maximum value ofnear the upstream of the Yangtze River Estuary. The spatial distribution of Case 3 was similar to that of Case 2， decreasing downstream along the Yangtze River Estuary， but the TP concentrations of Case 3 were about 0.02mg/L-0.03mg/L lower than those of Case 2 at the mouth area of theYangtze River Estuary. Compared with Case 1， the TP concentrations of Case 3 were about 0.04 mg/L-0.09 mg/L lower at the mouth area of the Yangtze River Estuary. The spatial distributions of TP from Case 1 and Case 3 were quite different， indicating that PP resuspension processes may play an important role in the TP concentration distribution in the Yangtze River Estuary.

4. Conclusions

Based on DHI’s Ecolab open platform， a threedimensional hydrodynamic and water quality model was developed for the Yangtze River Estuary， which considers the transport and transformation processes of different forms of N/P nutrients. Compared with the traditional convection-diffusion water quality models，the model developed in this study is more advanced and reasonable. The model verification results show that the model is of reasonable accuracy in simulating the hydrodynamics and the water quality in the Yangtze River Estuary.

Model preliminary application shows that the model can better characterize the distribution of TP in the Yangtze River Estuary， with a relatively high concentration zone of TP outside the river mouth. The high concentration zone is closely related with the PP resuspension process， indicating the significance of resuspension process in determining the distribution of TP in the Yangtze River Estuary.

Our model offers a good way to explore the compositions of N/P nutrients and their spatial and temporal variations in the Yangtze River Estuary， which is of great importance for the environmental management and protection.

References

［1］ CLOERN J. E. Our evolving conceptual model of the coastal eutrophication problem［J］. Marine Ecology Progress Series， 2001， 210（4）: 223-253.

［2］ TURNER R. E.， RABALAIS N. N. and JUSTIC D. et al. Global patterns of dissolved N， P and Si in large rivers［J］. Biogeochemistry， 2003， 64（3）: 297-317.

［3］ LI M.， XU K. and WATANABE M. et al. Long-term variations in dissolved silicate， nitrogen， and phosphorus flux from the Yangtze River into the East China Sea and impacts on estuarine ecosystem［J］. Estuarine， Coastal and Shelf Science， 2007， 71（1-2）: 3-12.

［4］ CHAI C.， YU Z. and SONG X. et al. The status and characteristics of eutrophication in the Yangtze River（Changjiang） Estuary and the adjacent East China Sea，China［J］. Hydrobiologia， 2006， 563（1）: 313-328.

［5］ WANG Bao-dong. Assessment of trophic status in the Changjiang （Yangtze River） Estuary［J］. Chinese Journal of Oceanology and Limnology， 2007， 25（3）: 261-269.

［6］ ZHOU M. J.， SHE Z. L. and YU R. C. Responses of a coastal phytoplankton community to increased nutrient input from the Changjiang （Yangtze） River［J］. Continental Shelf Research， 2008， 28（12）: 1483-1489.

［7］ XU J. T.， HOOD R. R. Modeling biogeochemical cycles in Chesapeake Bay with a coupled physical-biological models［J］. Estuarine Coastal and Shelf Science， 2006，69（1）: 19-46.

［8］ ARNDT S.， LACROIX G. and GYPENS N. et al. Nutrient dynamics and phytoplankton development along an estuary-coastal zone continuum: A model study［J］. Journal of Marine Systems， 2011， 84（3-4）: 49-66.

［9］ ZHENG L. Y.， CHENG C. S. and ZHANG F. Y. Development of water quality model in the Satilla River Estuary，Georgia［J］. Ecological Modelling， 2004， 178（3-4）: 457-482.

［10］ HU J.， LI S. Modeling the mass fluxes and transformations of nutrients in the Pearl River Delta， China［J］. Journal of Marine Systems， 2009， 78（1）: 146-167.

［11］ KUANG Cui-ping， CHEN Wei and GU Jie et al. Comprehensive analysis on the sediment siltation in the upper reach of the deepwater navigation channel in the Yangtze Estuary［J］. Journal of Hydrodynamics， 2014， 26（2）: 299-308.

［12］ KO K. Numerieal simulation of hydrodynamie characteristies and sewage diffusion in Yangtze Estuary［D］. Master Thesis， Nanjing， China: Hohai University， 2007.

［13］ LI Wen-ting. Numerical study on the impact of the sewage drainage on water quality of Shanghai water sources［D］. Master Thesis， Shanghai， China: Shanghai Ocean University， 2011（in Chinese）.

［14］ LIU Cheng， LI Xing-wei and WEI He-ping et al. Numerical simulation of the hydrodynamics and sewage diffusion in the Changjiang River Estuary［J］. Oceanologia et Limnologia Sinica， 2003， 34（5）: 474-483（in Chinese）.

［15］ LI Chi-hong. Numerical simulation of waterway planning influence to the water environment in the Changjiang River Estuary based on MIKE21 Model［C］. 2013 Chinese Hydraulic Engineering Society Annual Conference. Guangzhou， China， 2013， 1762-1766（in Chinese）.

［16］ ANDERSON D. M.， GLIBERT P. M. and BURKHOLDER J. M. Harmful algal blooms and eutrophication: Nutrient sources， composition， and consequences［J］. Estuaries， 2002， 25（4B）: 704-726.

［17］ BOWIE G. L.， MILLS W. B. and PORCELLA D. B. et al. Rates， constants， and kinetic formulations in surface water quality modeling［M］. （2rd Edition） Athens， USA: Environmental Protection Agency，1985.

［18］ EDELVANG K.， KAAS H. and ERICHSEN A. C. et al. Numerical modelling of phytoplankton biomass in coastal waters［J］. Journal of Marine Systems， 2005， 57（1-2）: 13-29.

［19］ FOURNIOTIS T. N.， HORSCH G. M. Three-dimensional numerical simulation of wind-induced barotropic circulation in the Gulf of Patras［J］. Ocean Engineering， 2010，37（4）: 355-364.

［20］ WOOL T. A.， AMBROSE R. B. and MARTIN J. L. et al. Water quality analysis simulation program （WASP）version 6.0 draft， user’s manual［M］. Atlanta， USA: Environmental Protection Agency， 2001.

［21］ LI Zheng， SHEN Zhi-liang and ZHOU Shu-qing et al. Distributions and variations of phosphorus in the Changjiang estuary and its adjacent sea areas［J］. Marine Science， 2007， 31（1）: 28-36（in Chinese）.

October 11， 2014， Revised February 3， 2015）

* Project supported by the National Natural Science Foundation of China （Grant Nos. 10972134， 11032007）， the Scientific research project of Shanghai Municipal Oceanic Bureau（Grant Nos. 2011-06， 2014-01） and the Shanghai Scientific Research Project （Grant Nos. 13231203600， 14231200104）.

Biography: WANG Biao （1983-）， Male， Ph. D.，

Senior Engineer