Estimating the critical phosphorus loading of shallow lakes with the ecosystem model PCLake: Sensitivity, calibration and uncertainty

There is a vast body of knowledge that eutrophication of lakes may cause algal blooms. Among lakes, shallow lakes are peculiar systems in that they typically can be in one of two contrasting (equilibrium) states that are self-stabilizing: a ‘clear’ state with submerged macrophytes or a ‘turbid’ state dominated by phytoplankton. Eutrophication may cause a switch from the clear to the turbid state, if the P loading exceeds a critical value. The ecological processes governing this switch are covered by the ecosystem model PCLake, a dynamic model of nutrient cycling and the biota in shallow lakes. Here we present an extensive analysis of the model, using a three-step procedure. (1) A sensitivity analysis revealed the key parameters for the model output. (2) These parameters were calibrated on the combined data on total phosphorus, chlorophyll-a, macrophytes cover and Secchi depth in over 40 lakes. This was done by a Bayesian procedure, giving a weight to each parameter setting based on its likelihood. (3) These weights were used for an uncertainty analysis, applied to the switchpoints (critical phosphorus loading levels) calculated by the model. The model was most sensitive to changes in water depth, P and N loading, retention time and lake size as external input factors, and to zooplankton growth rate, settling rates and maximum growth rates of phytoplankton and macrophytes as process parameters. The results for the ‘best run’ showed an acceptable agreement between model and data and classified nearly all lakes to which the model was applied correctly as either ‘clear’ (macrophyte-dominated) or ‘turbid’ (phytoplankton-dominated). The critical loading levels for a standard lake showed about a factor two uncertainty due to the variation in the posterior parameter distribution. This study calculates in one coherent analysis uncertainties in critical phosphorus loading, a parameter that is of great importance to water quality managers.     

Linking species- and ecosystem-level impacts of climate change in lakes with a complex and a minimal model

To study the interaction between species- and ecosystem-level impacts of climate change, we focus on the question of how climate-induced shifts in key species affect the positive feedback loops that lock shallow lakes either in a transparent, macrophyte-dominated state or, alternatively, in a turbid, phytoplankton-dominated state. We hypothesize that climate warming will weaken the resilience of the macrophyte-dominated clear state. For the turbid state, we hypothesize that climate warming and climate-induced eutrophication will increase the dominance of cyanobacteria. Climate change will also affect shallow lakes through a changing hydrology and through climate change-induced eutrophication. We study these phenomena using two models, the full ecosystem model PCLake and a minimal dynamic model of lake phosphorus dynamics. Quantitative predictions with the complex model show that changes in nutrient loading, hydraulic loading and climate warming can all lead to shifts in ecosystem state. The minimal model helped in interpreting the non-linear behaviour of the complex model. The main output parameters of interest for water quality managers are the critical nutrient loading at which the system will switch from clear to turbid and the much lower critical nutrient loading – due to hysteresis – at which the system switches back. Another important output parameter is the chlorophyll-a level in the turbid state. For each of these three output parameters we performed a sensitivity analysis to further understand the dynamics of the complex model PCLake. This analysis showed that our model results are most sensitive to changes in temperature-dependence of cyanobacteria, planktivorous fish and zooplankton. We argue that by combining models at various levels of complexity and looking at multiple aspects of climate changes simultaneously we can develop an integrated view of the potential impact of climate change on freshwater ecosystems.     

Using artificial neural network for reservoir eutrophication prediction

Reservoirs provide approximately 70% of water supply for domestic and industrial use in Taiwan. The water quality of reservoirs is now one of the key factors in the operation and water quality management of reservoirs. Transient weather patterns result in highly variable magnitudes of precipitation and thereby sharp fluctuations in the surface elevation of the reservoirs. In addition, excessive watershed development in the past two decades has contributed to continuing increase in nutrient loads to the reservoirs. The difficulty in quantifying watershed nutrient loads and uncentainties in kinetic mechanism in the water column present a technical challenge to the mass balance based modeling of reservoir eutrophication. This study offers an alternative approach to quantifying the cause-and-effect relationship in reservoir eutrophication with a data-driven method, i.e., capturing non-linear relationships among the water quality variables in the reservoir. A commonly used back-propagation neural network was used to relate the key factors that influence a number of water quality indicators such as dissolved oxygen (DO), total phosphorus (TP), chlorophyll-a (Chl-a), and secchi disk depth (SD) in a reservoir in central Taiwan. Study results show that the neural network is able to predict these indicators with reasonable accuracy, suggesting that the neural network is a valuable tool for reservoir management in Taiwan.     

Modeling linkages between sediment resuspension and water quality in a shallow,eutrophic, wind-exposed lake

The interactions between bed sediments and the water column in shallow, eutrophic lakes have tremendous implications for the fate and transport of nutrients in those water bodies. This has resulted in the development of water quality models for lakes incorporating the processes of sediment resuspension. Reliable resuspension models are thus needed to accurately represent this phenomenon. In this paper, three different sediment-resuspension models are combined with a hydrodynamic and water quality model, dynamic lake model-water quality (DLM-WQ), and the resulting models are used to simulate nutrient distributions in the highly eutrophic Salton Sea, California, USA. One of the resuspension formulas is based upon sediment characteristics as well as the bed shear stress exerted by wind-induced waves and currents, while the other two are standard, power-law-type formulas for cohesive sediments with two different exponents. The outputs for water quality variables, such as temperature, chlorophyll a, dissolved oxygen and nutrients, obtained from the three resulting models and from an earlier DLM-WQ run with a simple empirical sediment-resuspension model are compared with measured data. The level of agreement between the simulations and the measured data is assessed by using both statistical and graphical model evaluation methods, including measures of residual errors, sample autocorrelations, t-tests, and box plots. Based on these assessments, DLM-WQ with an extended version of the García and Parker [García, M.H., Parker, G., 1993. Experiments on the entrainment of sediment into suspension by a dense bottom current. J. Geophys. Res.-Oceans 98, 4793–4807] relationship gave the best results for water quality in the Salton Sea, confirming that the use of formulas with more information on the sediment characteristics yields more accurate results. To the best of our knowledge, this is the first effort to combine water quality models for lakes and reservoirs with a sediment-resuspension model which was originally intended for open-channel flows. The simulations confirm that sediment resuspension is the most dominant process in the Salton Sea's nutrient cycling. The effect of proposed physical changes to the Salton Sea on water quality characteristics is also addressed.     

Spatial and temporal hydrodynamic and water quality modeling analysis of a large reservoir on the South Carolina (USA) coastal plain

Two-dimensional, 31-segment, 61-channel hydrodynamic and water quality models of Lake Marion (surface area 330.7 km²; volume 1548.3×10⁶ m³) were developed using the WASP5 modeling system. Field data from 1985 to 1990 were used to parameterize the models. Phytoplankton kinetic rates and constants were obtained from a related in situ study; others from modeling literature. The hydrodynamic model was calibrated to estimates of daily lake volume; the water quality model was calibrated for ammonia, nitrate, ortho-phosphate, dissolved oxygen, chlorophyll-a, biochemical oxygen demand, organic nitrogen, and organic phosphorus. Water quality calibration suggested the model characterized phytoplankton and nutrient dynamics quite well. The model was validated (Kolmogorov–Smirnov two-sample goodness-of-fit test at P<0.05) by reparameterizing the nutrient loading functions using an independent set of field data. The models identified several factors that may contribute to the spatial variability previously reported from other research in the reservoir, despite the superficial absence of complex structure. Sensitivity analysis of the phytoplankton kinetic rates suggest that study site-specific estimates were important for obtaining model fit to field data. Sediment sources of ammonia (10–60 mg m⁻² day⁻¹) and phosphate (1–6 mg m⁻² day⁻¹) were important to achieve model calibration, especially during periods of high temperatures and low dissolved oxygen. This sediment flux accounted for 78% (nitrogen) and 50% (phosphorus) of the annual load. Spatial and temporal variability in the lake, reflected in the calibrated and validated models, suggest that ecological factors that influence phytoplankton productivity and nutrient dynamics are different in various parts of the lake. The WASP5 model as implemented here does not fully accommodate the ecological variability in Lake Marion due to model constraints on the specification of rate constants. This level of spatial detail may not be appropriate for an operational reservoir model, but as a research tool the models are both versatile and useful.     

湖泊营养物基准的候选变量和指标

湖泊营养物基准指标变量是可用于衡量水质、评价或预测水体的营养状态或富营养化程度的变量,是构成建立区域和湖泊营养物基准的基础。详细介绍和分析了氮、磷两个营养物变量和有机碳、叶绿素、透明度、溶解氧、大型植物、生物群落结构等生物学指标,以及营养状态指标、自养指数、温度、pH值、电导率以及土地利用等一系列相关变量（度量或指标）,并总结了当前美国各州在湖泊营养物基准指标选取方面的基本情况。归纳出选择营养物基准指标时应遵循的若干原则,即综合考虑富营养化控制关键因子和区域差异性,所选指标应具有相对稳定、早期预警和易于监测等特点。最后针对我国湖泊营养物基准指标的选取提出了几点建议。     

Analysis of spatial and temporal patterns in a large reservoir using water quality and hydrodynamic modeling

Effective management of reservoir water resources demands a good command of ecological processes in the waterbody. In this work the three-dimensional finite element hydrodynamic model RMA10 was coupled to an eutrophication model. The models were used together with a methodology for loads estimation to foster the understanding of such processes in the largest reservoir in Western Europe—the Alqueva. Nutrient enrichment and eutrophication are water quality concerns in this man-made impoundment. A total phosphorus and nitrogen loads quantification methodology was developed to estimate the inputs in the reservoir, using point and non-point source data.Field data (including water temperature, wind, water elevation, chlorophyll-a, nutrient concentration and dissolved oxygen) and estimated loads were used as forcing for simulations.The analysis of the modeling results shows that spatial and temporal distributions for water temperature, chlorophyll-a, dissolved oxygen and nutrients are consistent with measured in situ data.Modeling results allowed the identification of likely key impact factors on the water quality of the Alqueva reservoir. It is shown that the particular geomorphological and hydrological characteristics of the reservoir together with local climate features are responsible for the existence of distinct ecological regions within the reservoir.     

Population trends of Steller sea lions (Eumetopias jubatus) with respect to remote sensing measures of chlorophyll-a in critical habitat

The recovery plan for Steller sea lions (SSL; Eumetopias jubatus) suggests critical habitat should be enhanced to incorporate the spatio-temporal variation in dynamic oceanographic features that influence the prey and survival of SSL. It is necessary, therefore, to determine which features affect SSL. Demographics for sub-regions of the endangered, western stock of SSL were examined with respect to corresponding average, maximum, and variance of chlorophyll-a data (SeaWIFS), a proxy for primary productivity. Overall, SSL trends (2000–2008) and pup productivity (1999–2009) were related to maximum values of chl-a in critical habitat. Additionally, conditions in critical habitat appeared worse in areas of decline (i.e., dispersed patterns of chl-a hotspots and greater distances from SSL sites to productive areas). Although there may be a low feasibility of mitigating the effects of dynamic features on the recovery of SSL, the interactive effects of primary productivity and other stressors should be investigated for safeguarding their prey.     

我国淡水藻华长期变动特征综合分析

水体富营养化与淡水藻华（水华）是全球普遍现象,我国是一个高氮磷投入的国家,在人类活动和全球气候变化背景条件下,河流湖库富营养化问题日益突出。文章收集了近30年来我国水华事件的相关文献资料,基于综合研究和统计学分析,回顾性评估了我国水华发生的基本特征、变动规律和成因。数据表明,自上世纪80年代以来,水华发生频率从上世纪80年代每年1～2次上升到2000年以后的每年近10次（不完全统计）,总体呈现上升趋势;水华从湖泊向河流库区扩展,从点到面蔓延,且主要集中在人口密集、污染严重的长江中下游地区和东部沿海地区;在我国化肥施用、畜禽养殖、电站大坝建设等人类活动以及气候变暖的叠加影响下,河流湖库的氮磷浓度上升、氮磷比值下降,水体水化学和生态环境的长期变动加大了水华发生的风险。文章认为,在流域海域综合管理框架下进行氮磷污染的联合控制,以及多学科交叉开展特定水体的水华过程和机理研究是今后藻华防控的根本要求。     

Evaluation of biological water purification functions of inland lakes using an aquatic ecosystem model

An effective measure to cope with eutrophication of lakes is to remove nutrients that can cause algal blooming by taking advantage of natural water purification processes. Here the term “purification” is defined, in a wide sense, as the potential role of a water body to contribute to the reduction of pollutants and thus controlling eutrophication. Also regarded as a kind of ecological regulating services, biological purification involves various processes concerning seasonal nutrient fixation, such as uptake by aquatic macrophytes, biofouling onto foliage substrates, feeding by organisms in higher trophic level, and eternal loss or removal of substance from the water. In order to evaluate the water purification ability, a numerical lake ecosystem model highlighting the role of macrophyte colonies in the shore zone was developed and applied to Lakes Suwa, Kasumi and Biwa, as well as five small lakes attached to Lake Biwa.     

Assessing the effects of disease and bleaching on Florida Keys corals by fitting population models to data

Coral diseases have increased in frequency over the past few decades and have important influences on the structure and composition of coral reef communities. However, there is limited information on the etiologies of many coral diseases, and pathways through which coral diseases are acquired and transmitted are still in question. Furthermore, it is difficult to assess the impacts of disease on coral populations because outbreaks often co-occur with temperature-induced bleaching and anthropogenic stressors. We developed spatially explicit population models of coral disease and bleaching dynamics to quantify the impact of six common diseases on Florida Keys corals, including aspergillosis, dark spots, white band, white plague, white patch, and Caribbean yellow band. Models were fit to an 8-year data set of coral abundance, disease prevalence, and bleaching prevalence. Model selection was used to assess alternative pathways for disease transmission, and the influence of environmental stressors, including sea temperature and human population density, on disease prevalence and coral mortality. Classic disease transmission from contagious to susceptible colonies provided the best-fit model only for aspergillosis. For other diseases, external disease forcing, such as through a vector or directly from pathogens in the environment, provided the best fit to observed data. Estimates of disease reproductive ratio values (R₀) were less than one for each disease, indicating coral colonies were below densities required for diseases to become established through contagious spread alone. Incidences of white band and white patch disease were associated with greater susceptibility or slower recovery of bleached colonies, and no disease outbreaks were associated with periods of elevated sea temperatures alone. Projections of best-fit models indicated that, atleast during the period of this study, disease and bleaching did not have substantial impacts on populations and impaired rates of population growth appeared to be attributable to other stressors. By applying epidemiological models to field data, our study gives qualitative insights into the dynamics of coral diseases, relative stressor impacts, and directions for future research.     

Development and application of “U.S.E.D.”: A hydroclimate lake stratification model

Two model types are currently in use to model the thermal stratification cycle in lakes and reservoirs: the eddy diffusion and the mixed layer (or integral energy) approaches. Here the former is analysed and developments are proposed to remove the empiricisms previously implicit in these models. These discussions permit the reformulation of K_H0 independently of current shear, together with an expression for R_i. The deduced formulae are in good agreement with observations. The newly formulated model (the University of Salford eddy diffusion model, U.S.E.D.) is subsequently used in simulations of lakes and reservoirs at different latitudes which are found to be in good agreement with observations without requiring inter-site calibration.     

Stimulation of phytoplankton production in coastal waters by natural rainfall inputs: Nutritional and trophic implications

Recent evaluations of estuarine and coastal nutrient budgets implicate atmospheric deposition as a potentially significant (20 to 30%) source of biologically available nitrogen. We examined the potential growth stimulating impact of atmospheric nitrogen loading (ANL), as local rainfall, in representative shallow, nitrogen limited North Carolina mesohaline estuarine and euhaline coastal Atlantic Ocean habitats. From July 1988 to December 1989, using in situ bioassays, we examined natural phytoplankton growth responses, as¹⁴CO₂ assimilation and chlorophylla production, to rain additions over a range of dilutions mimicking actual input levels. Rainfall at naturally occurring dilutions (0.5 to 5%) stimulated both¹⁴CO₂ assimilation and chlorophylla production, in most cases in a highly significant manner. Parallel nutrient enrichments consistently pointed to nitrogen as the growth stimulating nutrient source. Generally, more acidic rainfall led to greater magnitudes of growth stimulation, especially at lower dilutions. Nutrient analyses of local rainfall from May 1988 to January 1990 indicated an inverse relationship between pH and NO ₃ ^- content. There have been growing concerns regarding increasing coastal and estuarine eutrophication, including ecologically and economically devastating phytoplankton blooms bordering urban and industrial regions of North America, Europe, Japan, and Korea. It appears timely, if not essential, to consider atmospheric nutrient loading in the formulation and implementation of nutrient management strategies aimed at mitigating coastal eutrophication.     

Macrozoobenthic assemblages in relation to environments of the Yangtze-isolated lakes

Eutrophication can shift lakes from a clear, macrophyte-dominated state to a turbid, algae-dominated state, and different habitat condition supports different fauna. Macrozoobenthos are good indicators of water environment, and studies on macrozoobenthic assemblage characteristics can help us to know which state a lake is in, thus provide the basis for its eutrophication control. In this study, a systematic investigation on macrozoobenthos was conducted in 17 Yangtze-isolated lakes to explore the macroecological laws of macrozoobenthic assemblages. Detrended correspondence analysis (DCA) revealed that variance of benthic assemblage structure occurred in two types of lakes. In macrophytic lakes, altogether 51 taxa of macrozoobenthos were identified. The average density and biomass of total macrozoobenthos were 2231 individuals·m^?2 and 1.69 g dry weight·m^?2, respectively. Macrozoobenthic assemblage was characterized by dominance of scrapers (i.e. gastropods). In algal lakes, altogether 20 taxa of macrozoobenthos were identified. The average density and biomass of total macrozoobenthos were 2814 individuals·m^?2 and 1.38 g dry weight·m^?2, respectively. Macrozoobenthic assemblage was characterized by dominance of collector-gatherers (i.e. oligochaetes). Wet biomass of submersed macrophytes (B _Mac) and phytoplankton chlorophyll a concentration (Chl a) were demonstrated as the key factor structuring macrozoobenthic assemblages in macrophytic and algal lakes, respectively.     

Control concept and countermeasures for shallow lakes’ eutrophication in China

Research on lake eutrophication in China began in the early 1970s, and many lakes in China are now known to be in meso-eutrophic status. Lake eutrophication has been showing a rapidly increasing trend since 2000. Investigations show that the main reasons for lake eutrophication include a fragile lake background environment, excessive nutrient loading into lakes, excessive human activities, ecological degeneration, weak environmental protection awareness, and lax lake management. Major mechanisms resulting from lake eutrophication include nutrient recycling imbalance, major changes in water chemistry (pH, oxygen, and carbon), lake ecosystem imbalance, and algal prevalence in lakes. Some concepts for controlling eutrophication should be persistently proposed, including lake catchment control, combination of pollutant source control with ecological restoration, protection of three important aspects (terrestrial ecology, lake coast zone, and submerged plant), and combination of lake management with regulation. Measures to control lake eutrophication should include pollution source control (i.e., optimize industrial structural adjustments in the lake catchment, reduce nitrogen and phosphorus emission amounts, and control endogenous pollution) and lake ecological restoration (i.e. establish a zone-lake buffer region and lakeside zone, protect regional vegetation, utilize hydrophytes in renovation technology); countermeasures for lake management should include implementing water quality management, identifying environmental and lake water goals, legislating and formulating laws and regulations to protect lakes, strengthening publicity and the education of people, increasing public awareness through participation in systems and mechanic innovations, establishing lake region management institutions, and ensuring implementation of governance and management measures.     

The impact of grazing by microzooplankton in northern Hiroshima Bay,the Seto Inland Seam,Japan

The abundance of microzooplankton and their grazing impact on phytoplankton were studied using the dilution technique from May 1990 to November 1991 in northern Hiroshima Bay, a typical eutrophic area in the Seto Inland Sea. Microzooplankton, dominated in number by tintinnid ciliates, were abundant from June to September when chlorophyll-a concentrations were high. Maximum density of microzooplankton ranged from 3.8×10³ to 25.4×10³ ind l^-1. During the period of investigation, mean microzooplankton density and mean chlorophyll-a concentration of the <20-m fraction increased toward the inner region of the bay. The microzooplankton grazing on phytoplankton increased from summer to early autumn, and decreased from late autumn to winter. At an offshore station, the annual means of the daily grazing loss for total chlorophyll-a and the chlorophyll-a of the <20-m fraction were 12 and 15% of the initial standing stock, respectively. At an estuarine station, the microzooplankton grazed 19 and 29% of the total and <20-m initial standing stock, respectively. The quantity of grazed chlorophyll-a correlated positively and linearly with the potential production of chlorophyll-a at both stations. The quantity of chlorophyll-a grazed by microzooplankton and the potential production of chlorophyll-a were nearly equivalent in the <20-m fraction at the estuarine station. This suggests that the microzooplankton assemblage was able to consume almost all the nanoplankton newly produced in the eutrophic estuary.     

Exploring the influence of lake water chemistry on chlorophyll a: A multivariate statistical model analysis

A multivariate statistical approach integrating the absolute principal components score (APCS) and multivariate linear regression (APCS-MLR), along with structural equation modeling (SEM), was used to model the influence of water chemistry variables on chlorophyll a (Chl a) in Lake Qilu, a severely polluted lake in southwestern China. Water quality was surveyed monthly from 2000 to 2005. APCS-MLR was used to identify key water chemistry variables, mine data for SEM, and predict Chl a. Seven principal components (PCs) were determined as eigenvalues >1, which explained 68.67% of the original variance. Four PCs were selected to predict Chl a using APCS-MLR. The results showed a good fit between the observed data and modeled values, with R² = 0.80. For SEM, Chl a and eight variables were used: NH₄-N (ammonia-nitrogen), total phosphorus (TP), Secchi disc depth (SD), cyanide (CN), arsenic (As), cadmium (Cd), fluoride (F), and temperature (T). A conceptual model was established to describe the relationships among the water chemistry variables and Chl a. Four latent variables were also introduced: physical factors, nutrients, toxic substances, and phytoplankton. In general, the SEM demonstrated good agreement between the sample covariance matrix of observed variables and the model-implied covariance matrix. Among the water chemistry factors, T and TP had the greatest positive influence on Chl a, whereas SD had the largest negative influence. These results will help researchers and decision-makers to better understand the influence of water chemistry on phytoplankton and to manage eutrophication adaptively in Lake Qilu.     

AQUATOX: Modeling environmental fate and ecological effects in aquatic ecosystems

AQUATOX combines aquatic ecosystem, chemical fate, and ecotoxicological constructs to obtain a truly integrative fate and effects model. It is a general, mechanistic ecological risk assessment model intended to be used to evaluate past, present, and future direct and indirect effects from various stressors including nutrients, organic wastes, sediments, toxic organic chemicals, flow, and temperature in aquatic ecosystems. The model has a very flexible structure and provides multiple analytical tools useful for evaluating ecological effects, including uncertainty analysis, nominal range sensitivity analysis, comparison of perturbed and control simulations, and graphing and tabulation of predicted concentrations, rates, and photosynthetic limitations. It can represent a full aquatic food web, including multiple genera and guilds of periphyton, phytoplankton, submersed aquatic vegetation, invertebrates, and fish and associated organic toxicants. It can model up to 20 organic chemicals simultaneously. (It does not model metals.) Modeled processes for organic toxicants include chemodynamics of neutral and ionized organic chemicals, bioaccumulation as a function of sorption and bioenergetics, biotransformation to daughter products, and sublethal and lethal toxicity. It has an extensive library of default biotic, chemical, and toxicological parameters and incorporates the ICE regression equations for estimating toxicity in numerous organisms. The model has been implemented for streams, small rivers, ponds, lakes, reservoirs, and estuaries. It is an integral part of the BASINS system with linkage to the watershed models HSPF and SWAT.     

Source identification and prediction of nitrogen and phosphorus pollution of Lake Taihu by an ensemble machine learning technique

● A machine learning model was used to identify lake nutrient pollution sources. ● XGBoost model showed the best performance for lake water quality prediction. ● Model feature size was reduced by screening the key features with the MIC method. ● TN and TP concentrations of Lake Taihu are mainly affected by endogenous sources. ● Next-month lake TN and TP concentrations were predicted accurately. Effective control of lake eutrophication necessitates a full understanding of the complicated nitrogen and phosphorus pollution sources, for which mathematical modeling is commonly adopted. In contrast to the conventional knowledge-based models that usually perform poorly due to insufficient knowledge of pollutant geochemical cycling, we employed an ensemble machine learning (ML) model to identify the key nitrogen and phosphorus sources of lakes. Six ML models were developed based on 13 years of historical data of Lake Taihu’s water quality, environmental input, and meteorological conditions, among which the XGBoost model stood out as the best model for total nitrogen (TN) and total phosphorus (TP) prediction. The results suggest that the lake TN is mainly affected by the endogenous load and inflow river water quality, while the lake TP is predominantly from endogenous sources. The prediction of the lake TN and TP concentration changes in response to these key feature variations suggests that endogenous source control is a highly desirable option for lake eutrophication control. Finally, one-month-ahead prediction of lake TN and TP concentrations (R² of 0.85 and 0.95, respectively) was achieved based on this model with sliding time window lengths of 9 and 6 months, respectively. Our work demonstrates the great potential of using ensemble ML models for lake pollution source tracking and prediction, which may provide valuable references for early warning and rational control of lake eutrophication.     

Neural network modelling of coastal algal blooms

An artificial neural network (ANN), a data driven modelling approach, is proposed to predict the algal bloom dynamics of the coastal waters of Hong Kong. The commonly used back-propagation learning algorithm is employed for training the ANN. The modeling is based on (a) comprehensive biweekly water quality data at Tolo Harbour (1982–2000); and (b) 4-year set of weekly phytoplankton abundance data at Lamma Island (1996–2000). Algal biomass is represented as chlorophyll-a and cell concentration of Skeletonema at the two locations, respectively. Analysis of a large number of scenarios shows that the best agreement with observations is obtained by using merely the time-lagged algal dynamics as the network input. In contrast to previous findings with more complicated neural networks of algal blooms in freshwater systems, the present work suggests the algal concentration in the eutrophic sub-tropical coastal water is mainly dependent on the antecedent algal concentrations in the previous 1–2 weeks. This finding is also supported by an interpretation of the neural networks’ weights. Through a systematic analysis of network performance, it is shown that previous reports of predictability of algal dynamics by ANN are erroneous in that ‘future data’ have been used to drive the network prediction. In addition, a novel real time forecast of coastal algal blooms based on weekly data at Lamma is presented. Our study shows that an ANN model with a small number of input variables is able to capture trends of algal dynamics, but data with a minimum sampling interval of 1 week is necessary. However, the sufficiency of the weekly sampling for real time predictions using ANN models needs to be further evaluated against longer weekly data sets as they become available.