A hybrid neural-genetic algorithm for reservoir water quality management

A combined neural network and genetic algorithm (GA) was developed for water quality management of Feitsui Reservoir in Taiwan. First, an artificial neural network (ANN) model was employed to simulate the behavior of nutrient loads into the reservoir. The data from watershed loads, precipitation in the watershed, and outflow were used in the ANN model to forecast the total phosphorus concentration in the reservoir. A 6-year (1992-97) record of water quality data was used for network training, and additional data collected in 1998-2000 were used for model verification. Further, a GA was used with this ANN model to optimize the control of nutrient loads from the watershed. The GA was used as a search strategy to determine the proper reduction rates of nutrient loads from the watershed so that the objective function could be as close to the optimal value as possible. The study results indicate that the ANN model can effectively simulate the dynamics of reservoir water quality. The GA is able to identify control schemes that reduce the in-reservoir total phosphorus concentration by as much as 60%, and water quality in the reservoir can be expected to achieve an oligotrophic (most of the time) or mesotrophic level if the watershed nutrient loads are reduced by 10-80%.     

Cost assessment and ecological effectiveness of nutrient reduction options for mitigating Phaeocystis colony blooms in the Southern North Sea: an integrated modeling approach

Nutrient reduction measures have been already taken by wealthier countries to decrease nutrient loads to coastal waters, in most cases however, prior to having properly assessed their ecological effectiveness and their economic costs. In this paper we describe an original integrated impact assessment methodology to estimate the direct cost and the ecological performance of realistic nutrient reduction options to be applied in the Southern North Sea watershed to decrease eutrophication, visible as Phaeocystis blooms and foam deposits on the beaches. The mathematical tool couples the idealized biogeochemical GIS-based model of the river system (SENEQUE-RIVERSTRAHLER) implemented in the Eastern Channel/Southern North Sea watershed to the biogeochemical MIRO model describing Phaeocystis blooms in the marine domain. Model simulations explore how nutrient reduction options regarding diffuse and/or point sources in the watershed would affect the Phaeocystis colony spreading in the coastal area. The reference and prospective simulations are performed for the year 2000 characterized by mean meteorological conditions, and nutrient reduction scenarios include and compare upgrading of wastewater treatment plants and changes in agricultural practices including an idealized shift towards organic farming. A direct cost assessment is performed for each realistic nutrient reduction scenario. Further the reduction obtained for Phaeocystis blooms is assessed by comparison with ecological indicators (bloom magnitude and duration) and the cost for reducing foam events on the beaches is estimated. Uncertainty brought by the added effect of meteorological conditions (rainfall) on coastal eutrophication is discussed. It is concluded that the reduction obtained by implementing realistic environmental measures on the short-term is costly and insufficient to restore well-balanced nutrient conditions in the coastal area while the replacement of conventional agriculture by organic farming might be an option to consider in the nearby future.     

An integrated multi-level watershed-reservoir modeling system for examining hydrological and biogeochemical processes in small prairie watersheds

Eutrophication of small prairie reservoirs presents a major challenge in water quality management and has led to a need for predictive water quality modeling. Studies are lacking in effectively integrating watershed models and reservoir models to explore nutrient dynamics and eutrophication pattern. A water quality model specific to small prairie water bodies is also desired in order to highlight key biogeochemical processes with an acceptable degree of parameterization. This study presents a Multi-level Watershed-Reservoir Modeling System (MWRMS) to simulate hydrological and biogeochemical processes in small prairie watersheds. It integrated a watershed model, a hydrodynamic model and an eutrophication model into a flexible modeling framework. It can comprehensively describe hydrological and biogeochemical processes across different spatial scales and effectively deal with the special drainage structure of small prairie watersheds. As a key component of MWRMS, a three-dimensional Willows Reservoir Eutrophication Model (WREM) is developed to addresses essential biogeochemical processes in prairie reservoirs and to generate 3D distributions of various water quality constituents; with a modest degree of parameterization, WREM is able to meet the limit of data availability that often confronts the modeling practices in small watersheds. MWRMS was applied to the Assiniboia Watershed in southern Saskatchewan, Canada. Extensive efforts of field work and lab analysis were undertaken to support model calibration and validation. MWRMS demonstrated its ability to reproduce the observed watershed water yield, reservoir water levels and temperatures, and concentrations of several water constituents. Results showed that the aquatic systems in the Assiniboia Watershed were nitrogen-limited and sediment flux played a crucial role in reservoir nutrient budget and dynamics. MWRMS can provide a broad context of decision support for water resources management and water quality protection in the prairie region.     

Long-term changes in watershed nutrient inputs and riverine exports in the Neuse River, North Carolina

We compared patterns of historical watershed nutrient inputs with in-river nutrient loads for the Neuse River, NC. Basin-wide sources of both nitrogen and phosphorus have increased substantially during the past century, marked by a sharp increase in the last 10 years resulting from an intensification of animal production. However, this recent increase is not reflected in changes in river loading over the last 20 years. Temporal patterns in river loads more closely parallel short-term changes in point sources and cropland nutrient application despite their overall lower magnitude. Total phosphorus loads have declined at all stations considered, corresponding to a 1988 phosphate detergent ban. Nitrogen load temporal patterns vary by location and the nitrogen fraction considered. The furthest upstream station exhibited nitrogen decreases after the completion of a dam in 1983. At a station just downstream of a rapidly growing urban area, the total nitrogen load has increased since the mid-1980s, primarily as a nitrate concentration increase. This is consistent with concurrent increases in chemical fertilizer use and point source discharges, as well as increased nitrification at treatment plants. This increase in nitrate loading is not reflected at the most downstream station, where no clear nitrogen trends are discernable. The lack of clear downstream nutrient increases suggests that current water quality impairment in the lower river and estuary may result from chronic nutrient overload rather than recent changes in the watershed. If this is true, then the impact of a planned 30% nitrogen loading reduction may not be immediately apparent. We calculate that, given annual variability, detecting a load reduction of this magnitude will take at least four years, and, should nutrients accumulated in the watershed become a significant source, detecting the resulting ecological improvements is likely to take substantially longer.     

Dynamics of diffuse pollution from US southern watersheds

To understand the effects of diffuse pollution information on the source of pollutants, quantities in transport, mode of transport, transient nature of the pollution event, and most importantly, a consideration of remediation efforts need to be known. For example, water quality research in the Yazoo Basin uplands in Mississippi has shown sediment loads from a conventional-till upland soybean watershed to be about 19,000kg/ha/yr, and responsible for 77-96% of P and N in transport. In contrast, sediment loads from a comparable no-till soybean watershed were only 500 kg/ha/yr. transporting about 31% of P and N in transport. Sediment loads from a nearby forested area were low, about 200 kg/ha/yr, but responsible for about 47-76% of P and N in transport. Transient pollution events are responsible for the transport of large quantities of sediment, nutrients, and pesticides; in some storm events nearly the annual load. Best management practices (BMPs) must be designed to remediate diffuse pollution and the transient nature of pollution events which can have a profound effect on the ecological health of steams and reservoirs.     

Effects of critical source area on sediment yield and streamflow

The spatial discretization scale determines the degree to which the spatial heterogeneity of the surface condition and the stream network can be represented. Discretization scale can have a significant impact on watershed model results, with respect to both hydrology and water quality. The impact of critical source area (CSA) on simulated water yield for each cell can be assessed for streamflow and run‐off using the Annualized Agricultural Non‐Point Source (AnnAGNPS) model. The total number of cells simulated for the six watersheds ranged from 8 to 352 for various CSA combinations. Increasing the number of subwatersheds increased water and sediment yield in many watersheds. Increasing the number of cells influences the depiction of land use and soil type accordingly and therefore influences run‐off. The applied CSAs were also compared with topographic parameters such as average cell slope, average channel slope and length to width ratio of each watershed.     

Estimating nutrient releases from agriculture in China: An extended substance flow analysis framework and a modeling tool

Agriculture related pollution has attracted the attention of policy makers as well as scientists in China as its contribution to water impairment has increased, and quantitative information at the national and regional levels is being sought to support decision making. However, traditional approaches are either time-consuming, expensive (e.g. national surveys) or oversimplified and crude (e.g. coefficient methods). Therefore, this study proposed an extended substance flow analysis (SFA) framework to estimate nutrient releases from agricultural and rural activities in China by depicting the nutrient flows in Chinese agro-ecosystems. The six-step process proposed herein includes: (a) system definition; (b) model development; (c) database development; (d) model validation; (e) results interpretation; and (f) uncertainty analysis. The developed Eubolism (Elementary Unit based nutrient Balance mOdeLIng in agro-ecoSysteM) model combined a nutrient balance module with an emission inventory module to quantify the nutrient flows in the agro-ecosystem. The model was validated and then applied to estimate the total agricultural nutrient loads, identify the contribution of different agricultural and rural activities and different land use types to the total loads, and analyze the spatial pattern of agricultural nutrient emissions in China. These results could provide an entire picture of agricultural pollution at the national level and be used to support policy making. Furthermore, uncertainties associated with the structure of the elementary units, spatial resolution, and inputs/parameters were also analyzed to evaluate the robustness of the model results.     

Modeling the contribution of point sources and non-point sources to Thachin River water pollution

Major rivers in developing and emerging countries suffer increasingly of severe degradation of water quality. The current study uses a mathematical Material Flow Analysis (MMFA) as a complementary approach to address the degradation of river water quality due to nutrient pollution in the Thachin River Basin in Central Thailand. This paper gives an overview of the origins and flow paths of the various point- and non-point pollution sources in the Thachin River Basin (in terms of nitrogen and phosphorus) and quantifies their relative importance within the system. The key parameters influencing the main nutrient flows are determined and possible mitigation measures discussed.The results show that aquaculture (as a point source) and rice farming (as a non-point source) are the key nutrient sources in the Thachin River Basin. Other point sources such as pig farms, households and industries, which were previously cited as the most relevant pollution sources in terms of organic pollution, play less significant roles in comparison. This order of importance shifts when considering the model results for the provincial level. Crosschecks with secondary data and field studies confirm the plausibility of our simulations. Specific nutrient loads for the pollution sources are derived; these can be used for a first broad quantification of nutrient pollution in comparable river basins. Based on an identification of the sensitive model parameters, possible mitigation scenarios are determined and their potential to reduce the nutrient load evaluated.A comparison of simulated nutrient loads with measured nutrient concentrations shows that nutrient retention in the river system may be significant. Sedimentation in the slow flowing surface water network as well as nitrogen emission to the air from the warm oxygen deficient waters are certainly partly responsible, but also wetlands along the river banks could play an important role as nutrient sinks.     

Thoughts on Synergetic Development of Watershed Management and Regional Economy

Zhu Qingping, the interviewee, is a prominent expert in China’s watershed management. Starting with the modes of China’s watershed management and the changes of water management philosophy, Zhu emphasizes that watershed management often involves various factors, including public resources, infrastructure construction, ecosystems, historical and cultural traditions, and population, all of which shall be taken into consideration as a whole. The interview then goes to the management of the Yellow River Basin, where Zhu explains the impacts between river flows and urban development, suggests a great opportunity for development the cities in the lower reaches of Yellow River Basin have, and proposes an idea of building a national ecological and cultural belt along the river. He further argues that watershed management requires collaborations across industries, disciplines, and administrative regions and divisions. He also believes that public engagement and maintenance plays an important role in watershed management and an intelligent water / watershed management system needs to be established by networking integrated big-data platforms to facilitate a more intelligent and coordinated water resource management while better ensuring water security at varied scales.     

Using satellite imagery for stormwater pollution management with Bayesian networks

Urban stormwater runoff is the primary source of many pollutants to Santa Monica Bay, but its monitoring and modeling is inherently difficult and often requires land use information as an intermediate process. Many approaches have been developed to estimate stormwater pollutant loading from land use. This research investigates an alternative approach, which estimates stormwater pollutant loadings directly from satellite imagery. We proposed a Bayesian network approach to classify a Landsat ETM(+) image of the Marina del Rey area in the Santa Monica Bay watershed. Eight water quality parameters were examined, including: total suspended solids, chemical oxygen demand, nutrients, heavy metals, and oil and grease. The pollutant loads for each parameter were classified into six levels: very low, low, medium low, medium high, high, and very high. The results provided spatial estimates of each pollutant load as thematic maps from which the greatest pollutant loading areas were identified. These results may be useful in developing best management strategies for stormwater pollution at regional and global scales and in establishing total maximum daily loads in the watershed. The approach can also be used for areas without ground-survey land use data.     

Quantification of stormwater runoff using a combined GIS and curve number approach: a case study for an urban watershed in the Ozark Highlands,USA

Managing stormwater runoff is crucial to preserving water quality in rapidly developing urban watersheds. The objective of this study was to develop a methodology to test existing stormwater drainage infrastructure, identify potential areas of improvement, and estimate potentially contaminated runoff by combining two widely used stormwater runoff prediction models. A watershed containing much of the University of Arkansas-Fayetteville campus was targeted for this study because stormwater from this watershed drains into a local river designated as an impaired water body due to siltation. The curve number method was used to estimate runoff for various flood-return periods and antecedent moisture conditions, while a flow-direction model integrated topography, land use, and stormwater drainage infrastructure in a GIS. The methodology developed and results generated will help stormwater planners visualise localised runoff and potentially adapt existing drainage networks to accommodate runoff, prevent flooding and erosion, and improve the stormwater quality entering nearby surfacewater bodies.     

Advancing assessment and design of stormwater monitoring programs using a self-organizing map: characterization of trace metal concentration profiles in stormwater runoff

Stormwater runoff poses a great challenge to the scientific assessment of the effects of diffuse pollution sources on receiving waters. In this study, a self-organizing map (SOM), a research tool for analyzing specific patterns in a large array of data, was applied to the monitoring data obtained from a stormwater monitoring survey to acquire new insights into stream water quality profiles under different rainfall conditions. The components of the input data vectors used by the SOM included concentrations of 10 metal elements, river discharge, and rainfall amount which were collected at the inlet and endpoint of an urban segment of the Yeongsan River, Korea. From the study, it was found that the SOM displayed significant variability in trace metal concentrations for different monitoring sites and rainfall events, with a greater impact of stormwater runoff on stream water quality at the upstream site than at the downstream site, except under low rainfall conditions (≤4 mm). In addition, the SOM clearly determined the water quality characteristics for “non-storm” and “storm” data, where the parameters nickel and arsenic and the parameters chromium, cadmium, and lead played an important role in reflecting the spatial and temporal water quality, respectively. When the SOM was used to examine the efficacy of stormwater quality monitoring programs, between 34 and 64% of the sample size in the current data set was shown to be sufficient for estimating the stormwater pollutant loads. The observed errors were small, generally being below 10, 6, and 20% for load estimation, map resolution, and clustering accuracy, respectively. Thus, the method recommended may be used to minimize monitoring costs if both the efficiency and accuracy are further determined by examining a large existing data set.     

Characterisation of diffuse pollutions from forested watersheds in Japan during storm events - its association with rainfall and watershed features

Forest areas have been identified as important sources of nonpoint pollution in Japan. The managers must estimate stormwater quality and quantities from forested watersheds to develop effective management strategies. Therefore, stormwater runoff loads and concentrations of 10 constituents (total suspended solids, dissolved organic carbon, PO(4)-P, dissolved total phosphorus, total phosphorus, NH(4)-N, NO(2)-N, NO(3)-N, dissolved total nitrogen, and total nitrogen) for 72 events across five regions (Aichi, Kochi, Mie, Nagano, and Tokyo) were characterised. Most loads were significantly and positively correlated with stormwater variables (total event rainfall, event duration, and rainfall intensity), but most discharge-weighted event concentrations (DWECs) showed negative correlations with rainfall intensity. Mean water quality concentration during baseflow was correlated significantly with storm concentrations (r=0.41-0.77). Although all pollutant load equations showed high coefficients of determination (R(2)=0.55-0.80), no models predicted well pollutant concentrations, except those for the three N constituents (R(2)=0.59-0.67). Linear regressions to estimate stormwater concentrations and loads were greatly improved by regional grouping. The lower prediction capability of the concentration models for Mie, compared with the other four regions, indicated that other watershed or storm characteristics should be included in the prediction models. Significant differences among regions were found more frequently in concentrations than in loads for all constituents. Since baseflow conditions implied available pollutant sources for stormwater, the similar spatial characteristics of pollutant concentrations between baseflow and stormflow conditions were an important control for stormwater quality.     

Nonpoint source pollution: a distributed water quality modeling approach

A distributed water quality model for nonpoint source pollution modeling in agricultural watersheds is described in this paper. A water quality component was developed for WATFLOOD (a flood forecast hydrological model) to deal with sediment and nutrient transport. The model uses a distributed group response unit approach for water quantity and quality modeling. Runoff, sediment yield and soluble nutrient concentrations are calculated separately for each land cover class, weighted by area and then routed downstream. With data extracted using Geographical Information Systems (GIS) technology for a local watershed, the model is calibrated for the hydrologic response and validated for the water quality component. The transferability of model parameters to other watersheds, especially those in remote areas without enough data for calibration, is a major problem in diffuse modeling. With the connection to GIS and the group response unit approach used in this paper, model portability increases substantially, which will improve nonpoint source modeling at the watershed-scale level.     

Using SWAT, Bacteroidales microbial source tracking markers,and fecal indicator bacteria to predict waterborne pathogen occurrence in an agricultural watershed

Developing the capability to predict pathogens in surface water is important for reducing the risk that such organisms pose to human health. In this study, three primary data source scenarios (measured stream flow and water quality, modelled stream flow and water quality, and host-associated Bacteroidales) are investigated within a Classification and Regression Tree Analysis (CART) framework for classifying pathogen (Escherichia coli 0157:H7, Salmonella, Campylobacter, Cryptosporidium, and Giardia) presence and absence (P/A) for a 178 km² agricultural watershed. To provide modelled data, a Soil Water Assessment Tool (SWAT) model was developed to predict stream flow, total suspended solids (TSS), total N and total P, and fecal indicator bacteria loads; however, the model was only successful for flow and total N and total P simulations, and did not accurately simulate TSS and indicator bacteria transport. Also, the SWAT model was not sensitive to an observed reduction in the cattle population within the watershed that may have resulted in significant reduction in E. coli concentrations and Salmonella detections. Results show that when combined with air temperature and precipitation, SWAT modelled stream flow and total P concentrations were useful for classifying pathogen P/A using CART methodology. From a suite of host-associated Bacteroidales markers used as independent variables in CART analysis, the ruminant marker was found to be the best initial classifier of pathogen P/A. Of the measured sources of independent variables, air temperature, precipitation, stream flow, and total P were found to be the most important variables for classifying pathogen P/A. Results indicate a close relationship between cattle pollution and pathogen occurrence in this watershed, and an especially strong link between the cattle population and Salmonella detections.     

Evaluating nitrogen removal by vegetation uptake using satellite image time series in riparian catchments

Nitrogen (N) removal by vegetation uptake is one of the most important functions of riparian buffer zones in preventing non-point source pollution (NSP), and many studies about N uptake at the river reach scale have proven the effectiveness of plants in controlling nutrient pollution. However, at the watershed level, the riparian zones form dendritic networks and, as such, may be the predominant spatially structured feature in catchments and landscapes. Thus, assessing the functions of riparian system at the basin scale is important. In this study, a new method coupling remote sensing and ecological models was used to assess the N removal by riparian vegetation on a large spatial scale. The study site is located around the Guanting reservoir in Beijing, China, which was abandoned as the source water system for Beijing due to serious NSP in 1997. SPOT 5 data was used to map the land cover, and Landsat-5 TM time series images were used to retrieve land surface parameters. A modified forest nutrient cycling and biomass model (ForNBM) was used to simulate N removal, and the modified net primary productivity (NPP) module was driven by remote sensing image time series. Besides the remote sensing data, the necessary database included meteorological data, soil chemical and physical data and plant nutrient data. Pot and plot experiments were used to calibrate and validate the simulations. Our study has proven that, by coupling remote sensing data and parameters retrieval techniques to plant growth process models, catchment scale estimations of nitrogen uptake rates can be improved by spatial pixel-based modelling.     

Growing season surface water loading of fecal indicator organisms within a rural watershed

The loading of microbial contaminants was examined within the Thomas Brook watershed, a 784 ha mixed land-use catchment located in the headwaters of the Cornwallis River drainage basin (Nova Scotia, Canada). The objectives were to: (i) examine spatial and temporal characteristics of fecal bacteria loading during the growing season from five subwatersheds, and (ii) develop areal fecal indicator organism export coefficients for rural landscapes. Fecal coliform, Escherichia coli, total suspended solids (TSS) concentrations and stream flow were monitored at five locations in the watershed over six consecutive growing seasons (May-Oct, 2001-2006). A nested watershed monitoring approach was used to determine bacterial loading from distinct source types (residential vs. agricultural) during both baseflow and stormflow periods. Areal bacterial loading rates increased in each nested watershed moving downstream through the watershed and were highest in the three subcatchments dominated by agricultural activities. Upper watershed bacterial loading throughout the growing season from an agricultural subcatchment (Growing Season Avg 8.92 × 10¹⁰ CFU ha⁻¹) was consistently higher than a residential subcatchment (Growing Season Avg 8.43 × 10⁹ CFU ha⁻¹). As expected, annual average stormflow bacterial loads were higher than baseflow loads, however baseflow loads still comprised between 14 and 35% of the growing season bacterial loads in the five subwatersheds. Fecal bacteria loads were greater during years with higher annual precipitation. A positive linear relationship was observed between E. coli and TSS loading during the 2005 and 2006 growing seasons when both parameters were monitored, indicating that the processes of sediment transport and bacterial transport are linked. It is anticipated that computed areal microbial loading coefficients will be useful in developing watershed management plans. More intensive sampling during stormflow events is recommended for improving these coefficients.