Assessing Temporal and Spatial Variability of Algal Bloom in Three Gorges Reservoir Using
Assessing temporal and spatial variability of algal bloom in Three Gorges Reservoir using an extended WASP model
Li Jian1,2,a, Qin Xiaosheng2,b*, Chen Min2,c
1Yangtze River Scientific Research Institute, No.23 Huangpu Road, Wuhan China
2School of Civil and Environmental Engineering, Nanyang Technological University, 50 Nanyang Ave.,Singapore
, ,
Keywords: Three Gorges Reservoir; Xiangxi River; SELFE; WASP
Abstract: Patterns of temporal and spatial variability of algal bloom in Three GorgesReservoir (TGR) and one of its tributaries,Xiangxi River (XR), were examined using hydrodynamic model SELFE and the extended Water Quality Analysis Simulation Program (WASP) model. Dynamics of phytoplankton biomass characterized by chlorophylla, nutrients (including organic and inorganic nitrogen and phosphorus), dissolved oxygen (DO), biochemical oxygen demand (BOD), suspended and bottom sediments, and water temperature considering solar irradiance were modeled. The coupled SELFE-WASP model was driven by surface wind force, heat fluxes, oxygen exchanges at water-air interface, solar radiation and boundary conditions fluxes from Yangtze Riverand XR. The model was calibrated by the field observed data of algal blooms in TGP in June. The chlorophyll-a concentration characterized as algal biomass influenced by many complex factors was preliminarily simulated. The simulated results showed a good agreement withtheobserved data. The developed model is a new tool to study the aquatic environmental problems in TGP.
Introduction
The Three Gorges Reservoir(TGR), with a storage capacity of 39.3 billionm3, is locatedin the upstream of the Yangtze River, China. The Yangtze River and its tributaries in the reservoir area are characterized by both narrow and deep boundary shape and steep slope. Meanwhile, obvious exchanges ofwater body and various solutes (including nutrients and suspended sediments)occur between the mainstream and the tributaries under frequent reservoir operations(for the purpose of flood control and hydroelectricity generation). Many pollutants flow into the reservoir especially the tributary ofXiangxi River (XR) from many chemical factories and sanitarysewage; this has caused the algal bloom phenomenon (resulting from massive phytoplankton reproduction) in the tributaries after water impoundment in 2003.
Over the past years, many researchers have conducted field observations and numerical simulations for studying problems of water quality deterioration and ecosystem evolution in TGR and its tributaries. Particularly, many hydrodynamic and water quality models have been applied for examining the TGR algal blooms, which include1-Dimensional [1], vertical 2-Dimensional [2] and 3-Dimensional[3] models.From the previous works, it is found that the flowing field, solute transport and phytoplankton biomass evolution in TGR presentsignificant spatial and temporal variations[4];hence,a 3-dimensional model is better used.Generally, the 3-dimensioal models can be divided into structured and unstructured meshes considering the topographical structure. For example, Dai et al. (2013) used a structured-mesh model, Delft3D, for modeling the TGR water quality [3]; however,the model cannot well follow the complex boundary and subaqueous relief in TGR. Therefore, in this study, an unstructuredmesh water quality model wasintroduced based on the hydrodynamic model SELFE[5] and the Water Quality Analysis Simulation Program (WASP) model[6].The WASP model used here has been extended considering the updated research literatures[7] and the local characteristics about TGR water quality problem[4].In this study, the algal bloom occurred in June 2007 was preliminarily simulated by the coupled model aiming at understanding the changing process of the horizontal and vertical distribution of phytoplankton biomass (chlorophyll-a). The research could not onlyhelp planners establish effective water quality management policies but also improve the ecosystem sustainabilityfor the TGP reservoir area.
Basic Theory of Water Quality Model
The model was developed based on an unstructured-mesh hydrodynamic model SELFE[5]. The extend WASP model was based on the theory of original WASP model[6] and the most updated water quality research works [7].The coupled hydrodynamic and extended WASP model covers the basic physical, chemical, and biological processes of an aquatic ecosystem in TGR as well as the effect of sediments including suspended and bottom layer sediment on the water quality processes. The model also considers the interactionsamong hydrodynamic conditions, temperature, underwater light intensity, nutrients, sediment, dissolved oxygen (DO), biochemical oxygen demand (BOD) and phytoplankton biomass (represented by chlorophyll-a concentration).The author has added more modules into the original WASP model including nonlinear adsorption-desorption interaction between suspended sediment and phosphate, solutes exchange between bottom sediment layer and overlying water considering bottom stress [8,9]. Meanwhile, the Nitrogen and Phosphorus nutrients as Total Nitrogen(TN) and Total Phosphorus (TP) used in the model have been divided into organic and inorganic nutrients including nitrate(NO3), ammonia(NH4), organic nitrogen(ON), phosphate(PO4) and organic phosphorous(OP) [8,9]. For technical details, readersare referred toliteratures [6, 7, 8, 9].
Modeling Area
The XR islocated in the Hubei Province of China andflows across Xingshan and Zigui Counties.The XR has a length of 94km length, with a catchment area of 2,939km2. Many other tributaries exist in TGR along the Yangtze River including Duanfangxi, Tongzhuanghe, and Qingshuihe,Yuanshuihe(as shown in Fig. 1). The Xingshan and Jianyangpin Stations are controlled hydrological stations withan average annual discharge of 47.4and 10.0 m3/s, respectively.Since the discharge and pollutant load into XR arepredominantly larger [10], other tributaries were considered as closed bay in the simulation.Serious water quality problems had occurred in these tributaries due to insufficient water mixing or flowing based on the in-situ observation [4]. Therefore,a high initial concentration of nutrients and chlorophyll-ais set in the calculation.The hydrodynamic condition at local tributaries is closely linked with the main streamflow of the Yangtze River.For example,both backflow mixing and the vertical exchange occurduring the water impoundmentand water release processes of TGR. Thus, the nutrients, sediment, and phytoplankton near the XR mouth interact with the main streamflow.Therefore, the water body exchange between Yangtze River and the tributaries influencing the algal bloom process was preliminarily simulated in this study.
Figure1.Schematic map of Three Gorges Reservoir and the field observation
The computational domain including Yangtze River, Xiangxi River,Gaolan River (the second-class tributary of Yangtze River) and some other first-class tributaries in TGR including Yuanshuihe, Qingshuihe, Tongzhuanghe and Duanfangxiis discretized with 15,282 horizontal triangle elements and 21 vertical “pure S-layers”. The time steps for hydrodynamic calculation and solute mass transport calculation are both set to 30s. A total of 11 observation stations in XR and 6 observation stations in Yangtze River were set up along the river course to measure the nutrients,suspended sediment, DO, BOD and chlorophyll-a concentrations in the water surface as well as biochemical and meteorological data for the algal blooms eventsoccurring in June 2007. The inlet boundary conditions of the Yangtze River, XR and Gaolan River were set following the hydrologic stations recorded discharge during the algal bloom, and the outlet boundary at the TGR dam was set to the actual operational water level. Thenutrients, suspended sediment, DO, BOD, water temperature, and chlorophyll-a concentrations on the first day of the algal bloom were defined as the initial calculation condition.
Modeling the Yangtze River mainstream and tributaries
The coupledSELFE-WASP model was calibrated using daily field measured data in June 2007 when the algal blooms, hydrodynamics, and water quality were observed synchronously and suitable for the calibration of 3D coupled modeling. For calibration runs, the velocity field was calculated from SELFE runs. The WASP model parameters were repeatedly adjusted to obtain a reasonable reproduction of field observed algal bloom process.Some of the model parameters were referred tothe previous literatures [1-3].Some parametersweredirectly obtained from experiments and field measurement.All adopted parameter values were in the range reported in Justićand Wang [7]. And the calibrated parameters used in modeling(not listed here) can reproduce the nutrients (TP) and algal bloom process (Chlorophyll-a) well as shown in Fig. 2.
(a) (b)
Figure 2. Comparison of the observed and simulated concentrations for calibration of (a)TP and (b)Chlorophyll-a
The flowing velocity in Yangtze River was obviously higherthan that in the Xiangxi tributary as shown in Fig.3(a) which wouldlead todifferent algal reproduction phenomenon. The chlorophyll-a concentration characterized as phytoplankton biomass was preliminarily simulated considering many factors influencing the growth, death and settling down of phytoplankton cells. The horizontal distribution of chlorophyll-a concentration at both water surface and bottom can providethe information about the algal bloomlocations in the TGR area in Fig.3(b). The algal bloom mainly occursin the upper reach of XR from Pingyikou to the confluence of XR and Gaolan River (Xiakou) as shown in Fig. 1. Meanwhile, the bottom Chlorophyll-a concentration wasalso obviously high at the shallow water zone like Pingyikou and Tongzhuanghe. The phytoplankton biomass in the Yangtze River was relatively lower than that in the tributaries. Generally, the algal bloomoccurringlocation could be preliminarily analyzed through the 3-dimensional modelingstudy.
(a) (b)
Figure 3.Horizontal distribution of variables(a) Flowing velocity at the confluence of Xiangxi River and Yangtze River and (b) Chlorophyll-a concentration near riverbed
The phytoplankton cells at water surface can get more lighting for photosynthesis than that in the deeper water column, which would cause obvious differences of vertical chlorophyll-a concentration. Because of over-saturated phenomenon caused by solar irradiation that will inhibit the photosynthesis, the chlorophyll-a concentration beneath the water surface around water depth of 5.0~20.0m reaches the maximum value.The chlorophyll-a concentration is not the highest at the water surface as shown in Fig. 3.The vertical concentration profiles of chlorophyll-a show stratifications in both the XR tributary (Fig. 4a) and the Yangtze River mainstream (Fig.4b). The chlorophyll-a concentration in XR tributary was obviously higher than that in Yangtze River mainstream because the relatively high flowing speed could increase the self-cleaning capability and also more physical transport woulddecrease the phytoplankton biomass in Yangtze River mainstream as shown in Fig. 3(a).The relatively stable flow in XR provides a suitable condition for the algal reproduction that has led to locally high chlorophyll-a concentrations; while the chlorophyll-a concentration in the Yangtze River is more related to the inflow boundary condition.
(a) (b)
Figure 4.Vertical concentration profile of Chlorophyll-a at: (a) XR and (b)Yangtze River.
Conclusion
Thisstudy presents a preliminary numerical simulation onthe spatial-temporal variation of water quality constituents (i.e.water temperature, suspended sediment, nutrients, DO,BOD, and chlorophyll-a) by using a 3D unstructured-mesh numerical model coupling SELFE and an extended WASP module. The model was successfully applied to the study ofTGR and XR algal bloom event in June 2007. The simulation domain covered Yangtze River and many tributaries which was different from the only simulation about XR algal bloom in the author's previous works[8,9].Realistic trends and magnitudes of chlorophyll-a concentration was preliminarily obtained from the numerical modeling and generally agreed with the field observations. The algal bloom occurring locationscould be analyzedsuccessfully through the proposed 3-dimensional model. Further studies on other water quality constituents are yet to be reported more thoroughlyin the near future.
Acknowledgement
This material is based on research/work supported by the National Science Foundation of China (51309021) and in part by the Singapore Ministry of National Development and National Research Foundation under L2 NIC Award No. L2NICCFP-2013-3 (WBS no.: M4061545.D63). We are also grateful to the XiangxiRiver Ecological Observation Station ofChina,Three Gorges Universityfor providing the field observation data.
References
[1]L. Wang, Z. Yu, H. Dai. Eutrophication model for river-typereservoir tributaries and its applications. Water Science and Engineering 2(2009)16-24.
[2]J. Lian, Y. Yao, C. Ma. Reservoir Operation Rules for Controlling Algal Blooms ina Tributary to the Impoundment of Three Gorges Dam. Water 6(2014) 3200-3223.
[3]H. Dai, J. Mao, D. Jiang. Longitudinal Hydrodynamic Characteristics in Reservoir Tributary Embayments and Effects on Algal Blooms. PLOSONE 8 (2013)1-14.
[4]T. Zheng, J. Mao, H. Dai. Impacts of water release operationson algal blooms in a tributary bay of Three Gorges Reservoir. Science ChinaTechnological Sciences 54(2011) 1588-1598.
[5]Y.Zhang, M. B. António. SELFE: A semi-implicit Eulerian-Lagrangian finite-element modelfor cross-scale ocean circulation. Ocean Modelling 21 (2008) 71-96.
[6]Wool T.M., et al.Water Quality Analysis SimulationProgram (WASP)Version 6 User’s Manual. US EnvironmentalProtection Agency, Atlanta, GA. 2001.
[7]D. Justić, L. Wang. Assessing temporalandspatialvariabilityofhypoxiaovertheinner Louisiana–upper Texasshelf:Applicationofanunstructured-grid three-dimensionalcoupledhydrodynamic-waterqualitymodel. Continental ShelfResearch72 (2014) 163-179.
[8]J. Li, D. Li, X. Wang. Three-dimensional unstructured mesh eutrophication model and its application to Xiangxi River, China. Journal of Environmental Sciences24 (2012)1569-1578.
[9]J. Li, Z. Jin, W. Yang. Numerical modeling of the Xiangxi River algal bloom and sediment-related process in China. Ecological Informatics 22 (2014) 23-35.
[10]K. Zhu. Studies on photoplankton community dynamics and eutrophication in the Three Gorges Reservoir. Doctoral dissertation of Chinese Academy of Sciences, 2013(In Chinese).