Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / The eutrophication is an environmental disorder that happens due to nutrients excess, mainly in lakes and reservoirs, from human action. This disorder causes the algal bloom, impairing the water quality, increasing the water turbidity and toxicity. This work aimed to assess the trophic state of Vacacai Mirim reservoir, located in the city of Santa Maria, southern Brazil, between 2010 and 2011, using a mathematical model for simulation and verify the water quality based on CONAMA resolution 357/2005 and on the Trophic State Index. The model used in this work is denominated CE-QUAL-W2 and was developed to water quality simulation in rivers and, principally in reservoirs. Data and information for the study area that have been published in previous work were used. These data were entered in the model as input data and then made the calibration of it. The calibration was made manually, using perturbation of parameters and has proven adequate, verifying the accuracy by the average absolute error and root mean square error. After calibration, the validation was made using a data set that was not entered as input data in calibration process. Afterwards, it was proposed extreme scenarios to verify the behavior of reservoir in front of situations of drought (30 days in length), high volume of precipitation (20x and 50x increase in rainfall volume), population growth (increase in nutrients concentration at 33, 66 and 99%) and finally, a possible treatment of effluents in the drainage basin (reduction in nutrients concentration at 33, 66 and 99%). The simulation of scenarios has shown that a drought of 30 days worsens the water quality, making the reservoir hypereutrophic and harming the classification of reservoir in reference values of Class 3 from CONAMA resolution 357/2005. The increase of precipitation dilutes nutrients and pollutants, decreasing their concentration, improving the water quality and keeping the trophic state of the reservoir. The scenario of the increase of pollutant load evidences that total phosphorous and chlorophyll-a exceeds limits reference values of Class 3, being necessary attention, because they raise the trophic state of the reservoir to hypereutrophic. The effluent discharge reduction scenario shows that, by reducing 66% the launch, there are considerable water quality gains, classifying the reservoir in Class 3 and being classified as mesotrophic. / A eutrofização é um distúrbio ambiental que acontece devido ao excesso de nutrientes, principalmente em lagos e reservatórios, proveniente da ação antrópica. Esse distúrbio causa a proliferação de algas, prejudicando a qualidade da água, aumentando os níveis de turbidez e também de toxicidade da água. Assim, este trabalho teve como objetivo avaliar o estado de trofia do reservatório do rio Vacacaí Mirim, na cidade de Santa Maria/RS, entre os anos de 2010 e 2011 utilizando um modelo matemático de simulação e verificar a qualidade da água frente a resolução CONAMA 357/2005 e o Índice de Estado Trófico. O modelo utilizado neste trabalho é denominado CE-QUAL-W2 e foi desenvolvido para simulações de qualidade da água em rios e, principalmente, em reservatórios. Foram utilizados dados e informações referentes a área de estudo, que foram publicados em trabalhos anteriores. Esses dados foram inseridos no modelo como dados de entrada, para então ser feita a calibração do mesmo. A calibração foi realizada de forma manual, utilizando a perturbação de parâmetros e se mostrou adequada, ao se verificar a precisão por meio de erro médio absoluto e da raiz do erro médio quadrático. Após o modelo calibrado, foi feita a validação utilizando uma campanha que havia ficado de fora da calibração. Posteriormente, foram propostos cenários extremos para se verificar o comportamento do reservatório frente situações de estiagem, de volumes intensos precipitados, do aumento populacional e, por último, de um possível tratamento dos efluentes na bacia de drenagem. A simulação de cenários mostrou que uma estiagem de 30 dias piora a qualidade da água, tornando o reservatório hipereutrófico e prejudicando a classificação do reservatório nos valores de referência da Classe 3. O aumento da precipitação dilui nutrientes e poluentes, fazendo com que a sua concentração diminua, melhorando a qualidade da água e mantendo o estado de trofia do reservatório. O cenário de aumento da carga poluidora evidencia que o fósforo total e a clorofila a excedem os valores de referência da Classe 3, sendo necessário uma atenção maior, visto que elevam o estado trófico do reservatório para hipereutrófico. O cenário de redução do lançamento de efluentes mostra que, ao se reduzir em 66% o lançamento, há consideráveis ganhos de qualidade da água, classificando o reservatório nos valores de referência da Classe 3 e sendo classificado como mesotrófico.

The shortage of water in Mexico has made public and private institutions look at reservoirs as an alternative solution for present and future water supply. However, eighty percent of the existing reservoirs in Mexico are contaminated at some level, many severely. Water quality models are water-management tools used to diagnose water quality problems and the impact of various environmental conditions. They can be effective in assessing various measures of remediation leading to improved water quality. In most of the cases such water quality models have been successfully applied in reservoirs located in temperate climates. However, the use of water quality models in subtropical reservoirs, especially those in developing countries, have relatively little application because either basic data are not available or because they are not sufficient. In this study, a preliminary water quality model was developed for a subtropical reservoir to assess both the ability to collect adequate data and the model's underlying applicability in a subtropical region. The Aguamilpa reservoir is located in the western part of Mexico (Nayarit). It was built for power generation, irrigation and as a fishery. CE-QUAL-W2 is a two-dimensional hydrodynamic and water quality model suitable for long and narrow water bodies. Geometrically the Aguamilpa reservoir is long and deep, making it an ideal candidate to be modeled by CE-QUAL-W2. The model was developed for 1995 and 1996 because of a wider availability of historical data during this period. In addition to a preliminary model and assessment of applicability in this subtropical region, a monitoring and data acquisition plan was designed to identify the minimum required data which must be used to update, calibrate and simulate the water quality parameters. Once the model is calibrated, it may be used to simulate the water quality changes occurring with respect to environmental, climatological and anthropogenic effects. Further, the model may be used to prescribe operating procedures upstream as well as at the dam which can serve to improve the overall water quality. The development of the model at Aguamilpa can serve as a guideline for developing similar water quality models in this and other similar subtropical locations.

Verification of applicability of the CE-QUAL-W2 model for simulation of phytoplankton, epiphyton and macrophyte growth in Lipno Reservoir was tested. The model was first calibrated for the year 2004, then was applicated to the whole investigated period from 1.1. 2000 to 31.12. 2005. Sensitivity analysis was executed for mean growth parameters of phytoplankton, periphyton and macrophytes.

A water quality model was created for Elephant Butte Reservoir, New Mexico to understand nutrient dynamics and algal response during a three year period. The model chosen for this study was CE-QUAL-W2 because of its proven ability to represent hydrodynamics and the ability to represent multiple algal groups. Elephant Butte has been subject to large algal blooms. This study examined the phosphorus loading into the reservoir to see if it could sustain the observed algal growth. Data showed that the amount of bioavailable phosphorus was more than enough to support large blooms. This study also was an initial attempt to model multiple groups of algae that will later be used in other water quality models built and maintained by the US Bureau of Reclamation. Four algal species were modeled; diatoms, dinoflagellates, cyanobacteria and greens. Research into literary values of kinetic parameters for multiple algal species was conducted and the results were compiled in this paper to assist future modeling efforts. Because Elephant Butte was often nitrogen-limited the calibration of algal growth was difficult. Algal growth was very sensitive to the adjustment of kinetic parameters for nitrogen half-saturation, light requirements, growth rates and temperature rate multipliers.

To evaluate the potential impacts of future climate change on a temperate reservoir, I used a calibrated water quality and hydrodynamic model validated using three years of data (2007-2009) from Deer Creek Reservoir (Utah). I evaluated the changes due to altered air temperatures, inflow rates, and nutrient loads that might occur under Global Climate Change (GCC). I developed methods to study GCC on reservoirs. I produced Average Water Temperature Plots, Stratification Plots, and Total Concentration Plots. Average Water Temperature Plots show the sensitivity of the water temperature to various parameters. Stratification Plots quantify stratification length and strength as well as ice-cover periods. Total Concentration Plots analyze the reservoir as a whole concerning water quality parameters. Increasing air temperature increased the water temperature, lengthened stratification time, increased stratification strength, decreased the ice-cover period, decreased the total algae concentration, decreased the flows, and caused peak nutrient concentrations to occur earlier. Decreasing flows caused increased water temperature, shorter stratification periods, weaker stratification, and increased nutrient concentrations. Increasing phosphate concentrations caused increases in total algae, dissolved oxygen, and phosphate concentrations. Variations in Nitrate-Nitrite concentrations did not influence the tested parameters. I found that the reservoir is only sensitive to these changes during the spring and summer. The tools which I developed were used to run the model scenarios, organize the data, and plot the results. They can be used on other reservoirs and for other water quality parameters.

Assessment of water quality within the Escatawpa River system was accomplished utilizing the Better Assessment Science Integrating Point and Nonpoint Sources (BASINS 2.0) to develop the watershed model, and the CE-QUAL-W2 software to develop the estuary model. The watershed model was utilized to quantify both hydrodynamic and water quality (fecal coliforms) characteristics of the watershed for a simulation period spanning from 1990 through 1999. Herein, calibration and application results are presented for watershed and estuary simulations made in an uncoupled manner. The models were developed such that loose coupling of watershed and estuary models can be accomplished as a subsequent phase of this ongoing project. CE-QUAL-W2 model calibration was performed utilizing a set of site specific data acquired on the Escatawpa Estuary System during an intensive survey period of September 10-15, 1997. Dissolved oxygen levels in the system were closely examined, with regards to the impacts from point source discharges.

When properly designed, hypolimnetic aeration and oxygenation systems can replenish dissolved oxygen in water bodies while preserving stratification. A comprehensive literature review of design methods for the three primary devices was completed. Using fundamental principles, a discrete-bubble model was first developed to predict plume dynamics and gas transfer for a circular bubble-plume diffuser. This approach has subsequently been validated in a large vertical tank and applied successfully at full-scale to an airlift aerator as well as to both circular and linear bubble-plume diffusers. The unified suite of models, all based on simple discrete-bubble dynamics, represents the current state-of-the-art for designing systems to add oxygen to stratified lakes and reservoirs. An existing linear bubble plume model was improved, and data collected from a full-scale diffuser installed in Spring Hollow Reservoir, Virginia (U.S.A.) were used to validate the model. The depth of maximum plume rise was simulated well for two of the three diffuser tests. Temperature predictions deviated from measured profiles near the maximum plume rise height, but predicted dissolved oxygen profiles compared very well to observations. Oxygen transfer within the hypolimnion was independent of all parameters except initial bubble radius. The results of this work suggest that plume dynamics and oxygen transfer can successfully be predicted for linear bubble plumes using the discrete-bubble approach. To model the complex interaction between a bubble plume used for hypolimnetic oxygenation and the ambient water body, a model for a linear bubble plume was coupled to two reservoir models, CE-QUAL-W2 (W2) and Si3D. In simulations with a rectangular basin, predicted oxygen addition was directly proportional to the update frequency of the plume model. W2 calculated less oxygen input to the basin than Si3D and significantly less mixing within the hypolimnion. The coupled models were then applied to a simplified test of a full-scale linear diffuser. Both the W2 and Si3D coupled models predicted bulk hypolimnetic DO concentrations well. Warming within the hypolimnion was overestimated by both models, but more so by W2. The lower vertical resolution of the reservoir grid in W2 caused the plume rise height to be over-predicted, enhancing erosion of the thermocline. / Ph. D.

The Lake Manassas Watershed is a 189 km2 basin located in the Northern Virginia suburbs of Washington, DC. Lake Manassas is a major waterbody in the watershed and serves as a drinking water source for the City of Manassas. Lake Manassas is experiencing eutrophication due to nutrient loads associated with agricultural activities and urban development in its drainage areas. Two watershed model applications using HSPF, and one receiving water quality model application using CE-QUAL-W2, were linked to simulate Lake Manassas as well as its drainage areas: the Upper Broad Run (126.21 km2) and Middle Broad Run (62.79 km2) subbasins. The calibration of the linked model was for the years 2002-05, with a validation period of 2006-07. The aspects of effective modeling of nutrient losses and nutrient management practices in the Lake Manassas watershed were investigated. The study was mainly conducted in the Upper Broad Run subbasin, which was simulated with an HSPF model. For nutrient simulation, HSPF provides two algorithms: PQUAL (simple, empirically based) and AGCHEM (detailed, process-based). This study evaluated and compared the modeling capabilities and performance of PQUAL and AGCHEM, and investigated significant inputs and parameters for their application. Integral to the study was to develop, calibrate and validate HSPF/PQUAL and HSPF/AGCHEM models in the Upper Broad Run subbasin. "One-variable-at-a-time" sensitivity analysis was conducted on the calibrated Upper Broad Run HSPF/PQUAL and HSPF/AGCHEM models to identify significant inputs and parameters for nutrient load generation. The sensitivity analysis results confirmed the importance of accurate meteorological inputs and flow simulation for effective nutrient modeling. OP (orthophosphate phosphorus) and NH4-N (ammonium nitrogen) loads were sensitive to PQUAL parameters describing pollutant buildup and washoff at land surface. The significant PQUAL parameter for Ox-N (oxidized nitrogen) load was groundwater nitrate concentration. For the HSPF/AGCHEM model, fertilizer application rate and time were very important for nutrient load generation. NH4-N and OP loads were sensitive to the AGCHEM parameters describing pollutant adsorption and desorption in the soil. On the other hand, plant uptake of nitrogen played an important role for Ox-N load generation. A side by side comparison was conducted on the Upper Broad Run HSPF/PQUAL and HSPF/AGCHEM models. Both PQUAL and AGCHEM provided good-to-reasonable nutrient simulation. The comparison results showed that AGCHEM performed better than PQUAL for OP simulation, but PQUAL captured temporal variations in the NH4-N and Ox-N loads better than AGCHEM. Compared to PQUAL, AGCHEM is less user-friendly, requires a lot more model input parameters and takes much more time in model development and calibration. On the other hand, use of AGCHEM affords more model capabilities, such as tracking nutrient balances and evaluating alternative nutrient management practices. This study also demonstrated the application of HSPF/AGCHEM within a linked watershed-reservoir model system in the Lake Manassas watershed. By using the outputs generated by the HSPF/AGCHEM models in the Upper Broad Run and Middle Broad Run subbasins, the Lake Manassas CE-QUAL-W2 model adequately captured water budget, temporal and spatial distribution of water quality constituents associated with summer stratification in the lake. The linked model was used to evaluate water quality benefits of implementing nutrient management plan in the watershed. The results confirmed that without the nutrient management plan OP loads would be much higher, which would lead to OP enrichment and enhanced algae growth in Lake Manassas. / Ph. D.

The Occoquan Watershed is a 1515 square kilometer basin located in northern Virginia and contains two principal waterbodies: the Occoquan Reservoir and Lake Manassas. Both waterbodies are principal drinking water supplies for local residents and experience eutrophication and summer algae growth. They are continuously threatened by new development from the rapid expansion of the greater Washington D.C. region. The Occoquan model, consisting of six HSPF and two CE-QUAL-W2 submodels linked in a complex way, has been developed and applied to simulate hydrology and water quality activities in the two major reservoirs and the associated drainage areas. The studied water quality constituents include temperature, dissolved oxygen, ammonium nitrogen, oxidized nitrogen, orthophosphate phosphorus, and algae. The calibration of the linked model is for the years 1993-95, with a validation period of 1996-97. The results show that a successful calibration can be achieved using the linked approach, with moderate additional effort. The spatial and temporal distribution of hydrology processes, nutrient detachment and transport, stream temperature and dissolved oxygen were well reproduced by HSPF submodels. By using the outputs generated by HSPF submodels, the CE-QUAL-W2 submodels adequately captured the water budgets, hydrodynamics, temperature, temporal and spatial distribution of dissolved oxygen, ammonium nitrogen, oxidized nitrogen, orthophosphate phosphorus, and algae in Lake Manassas and Occoquan Reservoir. This demonstrates the validity of linking two types of state of the art water quality models: the watershed model HSPF and the reservoir model CE-QUAL-W2. One of the advantages of the linked model approach is to develop a direct cause and effect relationship between upstream activities and downstream water quality. Therefore, scenarios of various land use proposals, BMP implementation, and point source management can be incorporated into HSPF applications, so that the CE-QUAL-W2 submodels can use the boundary conditions corresponding with these scenarios to predict the water quality variations in the receiving waterbodies. In this research, two land use scenarios were developed. One represented the background condition assuming all the land covered by forest and the other represented the environmental stress posed by future commercial and residential expansion. The results confirm the increases of external nutrient loads due to urbanization and other human activities, which eventually lead to nutrient enrichment and enhanced algae growth in the receiving waterbodies. The increases of external nutrient loads depend on land use patterns and are not evenly spread across the watershed. The future development in the non urban areas will greatly increase the external nutrient production and BMPs should be implemented to reduce the potential environmental degradation. For the existing urban areas, the model results suggest a potential threshold of nutrient production despite future land development. The model results also demonstrate the catchment function of Lake Manassas in reducing nutrient transport downstream. / Ph. D.

Integrated water resources management has been used for decades in various formats. The limited resources and the ever growing population keep imposing pressure on decision makers to better-, and reliably, manage the available waters. On the other hand, the continuous development in computing and modeling power has helped modelers and decision makers considerably. To use these models, assumptions have to be made to fill in the gaps of missing data and to approximate the current conditions. The type and amount of information available can also be used to help select the best model from the currently available models. Advances in data collection have not kept up to the pace of advances in model development and the need for more and reliable input parameter values. Hence, uncertainty in model input parameters also needs to be quantified and addressed. This research effort develops a spatially-based modeling framework to model watersheds from both water quantity and quality standpoints. In this research, Gridded Surface Sub-Surface Hydrologic Analysis (GSSHA) and CE-QUAL-W2 models are linked within the Watershed Modeling System (WMS); a GIS interface for hydrologic and hydraulic models, to better handle both models pre and post processing. In addition, stochastic analysis routines are developed and used to examine and address the uncertainty inherent in the modeling process of the interface between land and water in the designated watershed. The linkage routines are developed in WMS using C++. The two models are linked spatially and temporally with the general direction of data flow from GSSHA to CE-QUAL-W2. Pre-processing of the CE-QUAL-W2 model is performed first. Then stochastic parameters and their associated distributions are defined for stochastic analysis in GSSHA before a batch run is performed. GSSHA output is then aggregated by CE-QUAL-W2 segments to generate multiple CE-QUAL-W2 runs. WMS then reads the stochastic CE-QUAL-W2 runs upon successful completion for data analysis. Modelers need to generate a WMS Gage for each location where they want to examine the stochastic output. A Gage is defined by a segment and a layer in the CE-QUAl-W2 model. Once defined, modelers are able to view a computed credible interval with lower, upper bounds in addition to the mean time series of a pre-selected constituent. Decision makers can utilize this output to better manage watersheds by understanding and incorporating the spatio-temporal uncertainty for the land-water interface.