Willy Bauwens | Universiteit Gent and Vrije Universiteit Brussel (original) (raw)

Papers by Willy Bauwens

This paper describes the development of the Flash Flood Manager, abbreviated as FlaFloM. The Flas... more This paper describes the development of the Flash Flood Manager, abbreviated as FlaFloM. The Flash Flood Manager is an early warning system for flash floods which is developed under the EU LIFE project FlaFloM. It is applied to Wadi Watier located in the Sinai peninsula (Egypt) and discharges in the Red Sea at the local economic and tourist hub of Nuweiba city. FlaFloM consists of a chain of four modules: 1) Data gathering module, 2) Forecasting module, 3) Decision support module or DSS and 4) Warning module. Each module processes input data and consequently send the output to the following module. In case of a flash flood emergency, the final outcome of FlaFloM is a flood warning which is sent out to decision-makers. The ‘data gathering module’ collects input data from different sources, validates the input, visualise data and exports it to other modules. Input data is provided ideally as water stage (h), discharge (Q) and rainfall (R) through real-time field measurements and external forecasts. This project, however, as occurs in many arid flash flood prone areas, was confronted with a scarcity of data, and insufficient insight in the characteristics that release a flash flood. Hence, discharge and water stage data were not available. Although rainfall measurements are available through classical off line rain gauges, the sparse rain gauges network couldn’t catch the spatial and temporal characteristics of rainfall events. To overcome this bottleneck, we developed rainfall intensity raster maps (mm/hr) with an hourly time step and raster cell of 1*1km. These maps are derived through downscaling from two sources of global instruments: the weather research and forecasting model (WRF) and satellite estimates from the Tropical Rainfall Measuring Mission (TRMM). The ‘forecast module’ comprises three numerical models that, using data from the gathering module performs simulations on command: a rainfall-runoff model, a river flow model, and a flood model. A rainfall-runoff model transforms the (forecasted) rainfall into a runoff volume (m³) and consequently a time-dependent discharge (m³/s) for each of the subwadis which is then routed through the main channel. The flood model then converts the discharges into water stages and generates a spatially-distributed flood map. The rainfall-runoff model is developed in Matlab-Simulink. The latter two models are implemented in Infoworks and Floodworks (both Wallingford Software), which allows an automatic feed into the warning module. The ‘warning module’ has two tasks: 1) to generate specific flags when modelling results exceed pre-established thresholds for rainfall, discharge, water stage, volumes, etc… 2) to communicate the given flags as warning signals to operators and/or stakeholders. The ‘decision support module’ or DSS finally gives to the user the capability of performing alternative analysis in order to have a better idea of the reliability of the forecasts by means of the comparison of already made forecasts with new data and a sensitivity analysis. Although FlaFloM is now able to send out warnings, the forecasts of this first version are expected to be insufficiently accurate which may lead to false warnings and loss of trust with decision-makers if not communicated well. When new insights and data are available, the model will be updated which improves the forecast accuracy. At this moment, we see two major fields of improvement: 1) better rainfall forecasts and 2) better insights of the response of an arid area to storm events. Firstly, the rainfall maps provided better insights in the spatial and temporal extent of a rainfall event, though absolute rainfall values are not considered accurate. The major reason behind is the fact that both global systems are insufficiently parameterized for arid areas. New data from an improved rain gauge network is expected to add value. Secondly, better insights need to be gained on the response of the Wadi to rainfall. The calibration of the hydrological models is currently based on literature and a geological surface map from which we derived infiltration rates. Modelled discharges or flood volumes can only be assessed qualitatively based on the field knowledge of local Bedouins inhabitants. To reduce uncertainty on forecasts and to guide on new data to be collected, a sensitivity analysis with rainfall scenarios is performed.

Lake Victoria is situated within an elevated plateau in the western part of Africa's Great Rift V... more Lake Victoria is situated within an elevated plateau in the western part of Africa's Great Rift Valley and lies within the territory of three countries: Tanzania, Uganda and Kenya. It is Africa's largest lake and the second widest fresh water lake in the world in terms of surface area. It is also the source of the longest branch of the River Nile, the White Nile. The lake's shallowness, limited river inflow, and large surface area relative to its volume make it vulnerable to climate changes and fluctuations of the water level. This affects the surrounding countries and their people a lot, especially in terms of their food supply and economy. The aim of this study was to get more information on the causes of these fluctuations by analysing the water balance of the lake for the period 1970-1974. It was based both on historical data and measurements and new calculations, and compared with previous studies (e.g. Suttcliffe and Parks, 1999). Precipitation and evaporation over the lake surface were calculated with the Thiessen Polygons method, using measurements from stations around the lake and on the islands. The total inflow of the lake is the sum of the contributions of twelve subbasins. One of these subcatchments, the Nzoia-catchment, was modeled with SWAT (Soil and Water Assessment Tool), a physically based, semi-distributed river basin simulator, as a contribution to the development of a water balance model for Lake Victoria. To calculate the outflow at the Owen Falls Dam in Jinja (Uganda), gauge heights of the lake were used in combination with the "Agreed Curve" (the relationship between water level and flow that was set by the policy makers). As the lake is assumed to be a system with a closed mass balance, the combination of the variations in the above mentioned components resulted in changes of the lake's storage, leading to fluctuations of the water level. For the period 1970-1974 the calculated mean monthly evaporation is 133 mm, with a standard deviation of 11% and variations similar with previous studies. The mean monthly precipitation was calculated as 150 mm, with a standard deviation of 55%, which is very high, but previous studies also show large fluctuations and similar results. The monthly inflow has an average value of 26 mm (standard deviation 39%), with a large peak from April till May and a smaller from August till November due to the rainy seasons. The study of Suttcliffe and Parks showed even larger peaks for the April-May period. The results for the outflow (monthly 49 mm, standard deviation 9%) are almost equal to the results of Suttcliffe and Parks. Remarkable is the decreasing trend of the outflow value, indicating the lowering of the water level. As a result of the analysis of the components of the water balance in this study, the water level of Lake Victoria was calculated and the results were compared with historical measurements. The level shows (both in measurements and calculations) a large peak in April and May and a smaller one in November. For the period 1970-1974 a decrease of 360 mm of the water level was measured resulting in a yearly decrease of 72 mm. The calculation results in an average yearly decrease of 76 mm. Although this is very similar to the measurements, there is a big standard deviation (113 mm) on the difference between the monthly measured and calculated fluctuations. This is caused by the uncertainty on some of the components. If however the difference between calculated and measured water level is too large (e.g. three times the standard deviation), this can be seen as human interference. Suttcliffe J. and Y. Parks, 1999, The Hydrology of the Nile. Institute of Hydrology, Wallingford, Oxfordshire, UK: International Association of Hydrological Sciences Press.

Environmental Modelling and Software

Complex environmental models are controlled by a high number of parameters. Accurately estimating... more Complex environmental models are controlled by a high number of parameters. Accurately estimating the values of all these parameters is almost impossible. Sensitivity analysis (SA) results enable the selection of the parameters to include in a calibration procedure, but can also assist in the identification of the model processes. Additionally, a sensitivity analysis can yield crucial information on the use and meaning of the model parameters.

Water Science and Technology, 1997

ABSTRACT

Water Science and Technology, 1997

ABSTRACT

Water Science and Technology, 2006

n this research the applicability of two different water quality concepts, a QUAL2E-based and a R... more n this research the applicability of two different water quality concepts, a QUAL2E-based and a RWQM1-based water quality model is evaluated in terms of management decisions. The Dender river in Belgium serves as a case study for the application of the methodology. By using sensitivity analysis on both model concepts the important processes are revealed. Further, the differences between the predictions for a future scenario are analysed. The scenario chosen here is a reduction in fertiliser use of 90%, which reduces the diffuse pollution. This way, the advantages or disadvantages of using one concept against the other for this scenario are formulated. It was found that the QUAL-based models are more focussing on algae processes while the RWQM1 also takes into account processes in the sediment. Further the QUAL-based models are easier to calibrate, especially when only a small amount of data is available. Both concepts lead to more or less the same conclusions. However for some periods the differences become important and to reduce the uncertainty in those periods, more efforts should be spent in calibration and in better detection of parameters concerning sediment processes and diffusion.

Water Science and Technology, 2006

n this research the applicability of two different water quality concepts, a QUAL2E-based and a R... more n this research the applicability of two different water quality concepts, a QUAL2E-based and a RWQM1-based water quality model is evaluated in terms of management decisions. The Dender river in Belgium serves as a case study for the application of the methodology. By using sensitivity analysis on both model concepts the important processes are revealed. Further, the differences between the predictions for a future scenario are analysed. The scenario chosen here is a reduction in fertiliser use of 90%, which reduces the diffuse pollution. This way, the advantages or disadvantages of using one concept against the other for this scenario are formulated. It was found that the QUAL-based models are more focussing on algae processes while the RWQM1 also takes into account processes in the sediment. Further the QUAL-based models are easier to calibrate, especially when only a small amount of data is available. Both concepts lead to more or less the same conclusions. However for some periods the differences become important and to reduce the uncertainty in those periods, more efforts should be spent in calibration and in better detection of parameters concerning sediment processes and diffusion.

Environmental Modelling and Software, 2011

A modelling approach is presented that determines the most cost-effective set of reduction measur... more A modelling approach is presented that determines the most cost-effective set of reduction measures to reach an in-stream concentration target. The framework is based on the coupling of two models: the hydrological water quality model SWAT and an economic optimization model (Environmental Costing Model, ECM). SWAT is used to determine the relationship between the modelled in-stream concentration at the river basin outlet and the associated emission reduction. The ECM is used to set up marginal abatement cost curves for nutrients and oxygen demanding substances. Results for nitrogen are presented for the Grote Nete river basin in Belgium for the year 2006.

Environmental Modelling and Software, 2011

A modelling approach is presented that determines the most cost-effective set of reduction measur... more A modelling approach is presented that determines the most cost-effective set of reduction measures to reach an in-stream concentration target. The framework is based on the coupling of two models: the hydrological water quality model SWAT and an economic optimization model (Environmental Costing Model, ECM). SWAT is used to determine the relationship between the modelled in-stream concentration at the river basin outlet and the associated emission reduction. The ECM is used to set up marginal abatement cost curves for nutrients and oxygen demanding substances. Results for nitrogen are presented for the Grote Nete river basin in Belgium for the year 2006.

INTRODUCTIONIn the past, the growth of urban agglomerations in Europe caused major disruptions in... more INTRODUCTIONIn the past, the growth of urban agglomerations in Europe caused major disruptions inbasic needs such as public health, hygiene, and safety against flooding. Starting in thelast century, large scale urban drainage systems were implemented to tackle the worstwater related problems in urban areas. Subsequently, flood protection and public healthimproved dramatically. At the same time, the quality of receiving waters

INTRODUCTIONIn the past, the growth of urban agglomerations in Europe caused major disruptions in... more INTRODUCTIONIn the past, the growth of urban agglomerations in Europe caused major disruptions inbasic needs such as public health, hygiene, and safety against flooding. Starting in thelast century, large scale urban drainage systems were implemented to tackle the worstwater related problems in urban areas. Subsequently, flood protection and public healthimproved dramatically. At the same time, the quality of receiving waters

All the parts of an urban drainage system, i.e. the sewer system, the waste water treatment plant... more All the parts of an urban drainage system, i.e. the sewer system, the waste water treatment plant (WWTP) and the river, should be integrated into one single model to assess the performance of the overall system and for the development of design and control strategies assisting in its sustainable and cost effective management. Existing models for the individual components of the system have to be merged in order to develop the integrated tool. One of the problems arising from this methodology is the incompatibility of state variables, processes and parameters used in the different modelling approaches. Optimisation of an urban drainage system and of the waste water treatment process in particular requires a good knowledge of the waste water composition. As important transformations take place between the emission from the household and the arrival at the treatment facility, sewer models should include these transformations in the sewer system. At present, however, research is still needed in order to increase our knowledge concerning these in-sewer processes. A comparison of the state variables, processes and parameters has been carried out in both sewer models (SMs) and activated sludge models (ASMs). An ASM approach is used for the description of reactions in sewer models. However, a difference is found in the expression for organic material (expressed in terms of BOD) and heterotrophic biomass is absent as a state variable, resulting in differences in processes and parameters. Reconciliation of both the models seems worthwhile and a preliminary solution is suggested in this paper.

All the parts of an urban drainage system, i.e. the sewer system, the waste water treatment plant... more All the parts of an urban drainage system, i.e. the sewer system, the waste water treatment plant (WWTP) and the river, should be integrated into one single model to assess the performance of the overall system and for the development of design and control strategies assisting in its sustainable and cost effective management. Existing models for the individual components of the system have to be merged in order to develop the integrated tool. One of the problems arising from this methodology is the incompatibility of state variables, processes and parameters used in the different modelling approaches. Optimisation of an urban drainage system and of the waste water treatment process in particular requires a good knowledge of the waste water composition. As important transformations take place between the emission from the household and the arrival at the treatment facility, sewer models should include these transformations in the sewer system. At present, however, research is still needed in order to increase our knowledge concerning these in-sewer processes. A comparison of the state variables, processes and parameters has been carried out in both sewer models (SMs) and activated sludge models (ASMs). An ASM approach is used for the description of reactions in sewer models. However, a difference is found in the expression for organic material (expressed in terms of BOD) and heterotrophic biomass is absent as a state variable, resulting in differences in processes and parameters. Reconciliation of both the models seems worthwhile and a preliminary solution is suggested in this paper.

Environmental Modelling and Software, 2007

To evaluate the future state of river water in view of actual pollution loading or different mana... more To evaluate the future state of river water in view of actual pollution loading or different management options, water quality models are a useful tool. However, the uncertainty on the model predictions is sometimes too high to draw proper conclusions. Because of the complexity of process based river water quality models, it is best to investigate this problem according to the origin of the uncertainty. If the uncertainty stems from input data or parameter uncertainty, more reliable results are obtained by performing specific measurement campaigns. The aim of the research reported in this paper is to guide these measurement campaigns based on an uncertainty analysis. The practical case study is the river Dender in Flanders, Belgium.

Environmental Modelling and Software, 2007

To evaluate the future state of river water in view of actual pollution loading or different mana... more To evaluate the future state of river water in view of actual pollution loading or different management options, water quality models are a useful tool. However, the uncertainty on the model predictions is sometimes too high to draw proper conclusions. Because of the complexity of process based river water quality models, it is best to investigate this problem according to the origin of the uncertainty. If the uncertainty stems from input data or parameter uncertainty, more reliable results are obtained by performing specific measurement campaigns. The aim of the research reported in this paper is to guide these measurement campaigns based on an uncertainty analysis. The practical case study is the river Dender in Flanders, Belgium.

This paper describes the development of the Flash Flood Manager, abbreviated as FlaFloM. The Flas... more This paper describes the development of the Flash Flood Manager, abbreviated as FlaFloM. The Flash Flood Manager is an early warning system for flash floods which is developed under the EU LIFE project FlaFloM. It is applied to Wadi Watier located in the Sinai peninsula (Egypt) and discharges in the Red Sea at the local economic and tourist hub of Nuweiba city. FlaFloM consists of a chain of four modules: 1) Data gathering module, 2) Forecasting module, 3) Decision support module or DSS and 4) Warning module. Each module processes input data and consequently send the output to the following module. In case of a flash flood emergency, the final outcome of FlaFloM is a flood warning which is sent out to decision-makers. The ‘data gathering module’ collects input data from different sources, validates the input, visualise data and exports it to other modules. Input data is provided ideally as water stage (h), discharge (Q) and rainfall (R) through real-time field measurements and external forecasts. This project, however, as occurs in many arid flash flood prone areas, was confronted with a scarcity of data, and insufficient insight in the characteristics that release a flash flood. Hence, discharge and water stage data were not available. Although rainfall measurements are available through classical off line rain gauges, the sparse rain gauges network couldn’t catch the spatial and temporal characteristics of rainfall events. To overcome this bottleneck, we developed rainfall intensity raster maps (mm/hr) with an hourly time step and raster cell of 1*1km. These maps are derived through downscaling from two sources of global instruments: the weather research and forecasting model (WRF) and satellite estimates from the Tropical Rainfall Measuring Mission (TRMM). The ‘forecast module’ comprises three numerical models that, using data from the gathering module performs simulations on command: a rainfall-runoff model, a river flow model, and a flood model. A rainfall-runoff model transforms the (forecasted) rainfall into a runoff volume (m³) and consequently a time-dependent discharge (m³/s) for each of the subwadis which is then routed through the main channel. The flood model then converts the discharges into water stages and generates a spatially-distributed flood map. The rainfall-runoff model is developed in Matlab-Simulink. The latter two models are implemented in Infoworks and Floodworks (both Wallingford Software), which allows an automatic feed into the warning module. The ‘warning module’ has two tasks: 1) to generate specific flags when modelling results exceed pre-established thresholds for rainfall, discharge, water stage, volumes, etc… 2) to communicate the given flags as warning signals to operators and/or stakeholders. The ‘decision support module’ or DSS finally gives to the user the capability of performing alternative analysis in order to have a better idea of the reliability of the forecasts by means of the comparison of already made forecasts with new data and a sensitivity analysis. Although FlaFloM is now able to send out warnings, the forecasts of this first version are expected to be insufficiently accurate which may lead to false warnings and loss of trust with decision-makers if not communicated well. When new insights and data are available, the model will be updated which improves the forecast accuracy. At this moment, we see two major fields of improvement: 1) better rainfall forecasts and 2) better insights of the response of an arid area to storm events. Firstly, the rainfall maps provided better insights in the spatial and temporal extent of a rainfall event, though absolute rainfall values are not considered accurate. The major reason behind is the fact that both global systems are insufficiently parameterized for arid areas. New data from an improved rain gauge network is expected to add value. Secondly, better insights need to be gained on the response of the Wadi to rainfall. The calibration of the hydrological models is currently based on literature and a geological surface map from which we derived infiltration rates. Modelled discharges or flood volumes can only be assessed qualitatively based on the field knowledge of local Bedouins inhabitants. To reduce uncertainty on forecasts and to guide on new data to be collected, a sensitivity analysis with rainfall scenarios is performed.

Lake Victoria is situated within an elevated plateau in the western part of Africa's Great Rift V... more Lake Victoria is situated within an elevated plateau in the western part of Africa's Great Rift Valley and lies within the territory of three countries: Tanzania, Uganda and Kenya. It is Africa's largest lake and the second widest fresh water lake in the world in terms of surface area. It is also the source of the longest branch of the River Nile, the White Nile. The lake's shallowness, limited river inflow, and large surface area relative to its volume make it vulnerable to climate changes and fluctuations of the water level. This affects the surrounding countries and their people a lot, especially in terms of their food supply and economy. The aim of this study was to get more information on the causes of these fluctuations by analysing the water balance of the lake for the period 1970-1974. It was based both on historical data and measurements and new calculations, and compared with previous studies (e.g. Suttcliffe and Parks, 1999). Precipitation and evaporation over the lake surface were calculated with the Thiessen Polygons method, using measurements from stations around the lake and on the islands. The total inflow of the lake is the sum of the contributions of twelve subbasins. One of these subcatchments, the Nzoia-catchment, was modeled with SWAT (Soil and Water Assessment Tool), a physically based, semi-distributed river basin simulator, as a contribution to the development of a water balance model for Lake Victoria. To calculate the outflow at the Owen Falls Dam in Jinja (Uganda), gauge heights of the lake were used in combination with the "Agreed Curve" (the relationship between water level and flow that was set by the policy makers). As the lake is assumed to be a system with a closed mass balance, the combination of the variations in the above mentioned components resulted in changes of the lake's storage, leading to fluctuations of the water level. For the period 1970-1974 the calculated mean monthly evaporation is 133 mm, with a standard deviation of 11% and variations similar with previous studies. The mean monthly precipitation was calculated as 150 mm, with a standard deviation of 55%, which is very high, but previous studies also show large fluctuations and similar results. The monthly inflow has an average value of 26 mm (standard deviation 39%), with a large peak from April till May and a smaller from August till November due to the rainy seasons. The study of Suttcliffe and Parks showed even larger peaks for the April-May period. The results for the outflow (monthly 49 mm, standard deviation 9%) are almost equal to the results of Suttcliffe and Parks. Remarkable is the decreasing trend of the outflow value, indicating the lowering of the water level. As a result of the analysis of the components of the water balance in this study, the water level of Lake Victoria was calculated and the results were compared with historical measurements. The level shows (both in measurements and calculations) a large peak in April and May and a smaller one in November. For the period 1970-1974 a decrease of 360 mm of the water level was measured resulting in a yearly decrease of 72 mm. The calculation results in an average yearly decrease of 76 mm. Although this is very similar to the measurements, there is a big standard deviation (113 mm) on the difference between the monthly measured and calculated fluctuations. This is caused by the uncertainty on some of the components. If however the difference between calculated and measured water level is too large (e.g. three times the standard deviation), this can be seen as human interference. Suttcliffe J. and Y. Parks, 1999, The Hydrology of the Nile. Institute of Hydrology, Wallingford, Oxfordshire, UK: International Association of Hydrological Sciences Press.

Environmental Modelling and Software

Complex environmental models are controlled by a high number of parameters. Accurately estimating... more Complex environmental models are controlled by a high number of parameters. Accurately estimating the values of all these parameters is almost impossible. Sensitivity analysis (SA) results enable the selection of the parameters to include in a calibration procedure, but can also assist in the identification of the model processes. Additionally, a sensitivity analysis can yield crucial information on the use and meaning of the model parameters.

Water Science and Technology, 1997

ABSTRACT

Water Science and Technology, 1997

ABSTRACT

Water Science and Technology, 2006

n this research the applicability of two different water quality concepts, a QUAL2E-based and a R... more n this research the applicability of two different water quality concepts, a QUAL2E-based and a RWQM1-based water quality model is evaluated in terms of management decisions. The Dender river in Belgium serves as a case study for the application of the methodology. By using sensitivity analysis on both model concepts the important processes are revealed. Further, the differences between the predictions for a future scenario are analysed. The scenario chosen here is a reduction in fertiliser use of 90%, which reduces the diffuse pollution. This way, the advantages or disadvantages of using one concept against the other for this scenario are formulated. It was found that the QUAL-based models are more focussing on algae processes while the RWQM1 also takes into account processes in the sediment. Further the QUAL-based models are easier to calibrate, especially when only a small amount of data is available. Both concepts lead to more or less the same conclusions. However for some periods the differences become important and to reduce the uncertainty in those periods, more efforts should be spent in calibration and in better detection of parameters concerning sediment processes and diffusion.

Water Science and Technology, 2006

n this research the applicability of two different water quality concepts, a QUAL2E-based and a R... more n this research the applicability of two different water quality concepts, a QUAL2E-based and a RWQM1-based water quality model is evaluated in terms of management decisions. The Dender river in Belgium serves as a case study for the application of the methodology. By using sensitivity analysis on both model concepts the important processes are revealed. Further, the differences between the predictions for a future scenario are analysed. The scenario chosen here is a reduction in fertiliser use of 90%, which reduces the diffuse pollution. This way, the advantages or disadvantages of using one concept against the other for this scenario are formulated. It was found that the QUAL-based models are more focussing on algae processes while the RWQM1 also takes into account processes in the sediment. Further the QUAL-based models are easier to calibrate, especially when only a small amount of data is available. Both concepts lead to more or less the same conclusions. However for some periods the differences become important and to reduce the uncertainty in those periods, more efforts should be spent in calibration and in better detection of parameters concerning sediment processes and diffusion.

Environmental Modelling and Software, 2011

A modelling approach is presented that determines the most cost-effective set of reduction measur... more A modelling approach is presented that determines the most cost-effective set of reduction measures to reach an in-stream concentration target. The framework is based on the coupling of two models: the hydrological water quality model SWAT and an economic optimization model (Environmental Costing Model, ECM). SWAT is used to determine the relationship between the modelled in-stream concentration at the river basin outlet and the associated emission reduction. The ECM is used to set up marginal abatement cost curves for nutrients and oxygen demanding substances. Results for nitrogen are presented for the Grote Nete river basin in Belgium for the year 2006.

Environmental Modelling and Software, 2011

A modelling approach is presented that determines the most cost-effective set of reduction measur... more A modelling approach is presented that determines the most cost-effective set of reduction measures to reach an in-stream concentration target. The framework is based on the coupling of two models: the hydrological water quality model SWAT and an economic optimization model (Environmental Costing Model, ECM). SWAT is used to determine the relationship between the modelled in-stream concentration at the river basin outlet and the associated emission reduction. The ECM is used to set up marginal abatement cost curves for nutrients and oxygen demanding substances. Results for nitrogen are presented for the Grote Nete river basin in Belgium for the year 2006.

INTRODUCTIONIn the past, the growth of urban agglomerations in Europe caused major disruptions in... more INTRODUCTIONIn the past, the growth of urban agglomerations in Europe caused major disruptions inbasic needs such as public health, hygiene, and safety against flooding. Starting in thelast century, large scale urban drainage systems were implemented to tackle the worstwater related problems in urban areas. Subsequently, flood protection and public healthimproved dramatically. At the same time, the quality of receiving waters

INTRODUCTIONIn the past, the growth of urban agglomerations in Europe caused major disruptions in... more INTRODUCTIONIn the past, the growth of urban agglomerations in Europe caused major disruptions inbasic needs such as public health, hygiene, and safety against flooding. Starting in thelast century, large scale urban drainage systems were implemented to tackle the worstwater related problems in urban areas. Subsequently, flood protection and public healthimproved dramatically. At the same time, the quality of receiving waters

All the parts of an urban drainage system, i.e. the sewer system, the waste water treatment plant... more All the parts of an urban drainage system, i.e. the sewer system, the waste water treatment plant (WWTP) and the river, should be integrated into one single model to assess the performance of the overall system and for the development of design and control strategies assisting in its sustainable and cost effective management. Existing models for the individual components of the system have to be merged in order to develop the integrated tool. One of the problems arising from this methodology is the incompatibility of state variables, processes and parameters used in the different modelling approaches. Optimisation of an urban drainage system and of the waste water treatment process in particular requires a good knowledge of the waste water composition. As important transformations take place between the emission from the household and the arrival at the treatment facility, sewer models should include these transformations in the sewer system. At present, however, research is still needed in order to increase our knowledge concerning these in-sewer processes. A comparison of the state variables, processes and parameters has been carried out in both sewer models (SMs) and activated sludge models (ASMs). An ASM approach is used for the description of reactions in sewer models. However, a difference is found in the expression for organic material (expressed in terms of BOD) and heterotrophic biomass is absent as a state variable, resulting in differences in processes and parameters. Reconciliation of both the models seems worthwhile and a preliminary solution is suggested in this paper.

All the parts of an urban drainage system, i.e. the sewer system, the waste water treatment plant... more All the parts of an urban drainage system, i.e. the sewer system, the waste water treatment plant (WWTP) and the river, should be integrated into one single model to assess the performance of the overall system and for the development of design and control strategies assisting in its sustainable and cost effective management. Existing models for the individual components of the system have to be merged in order to develop the integrated tool. One of the problems arising from this methodology is the incompatibility of state variables, processes and parameters used in the different modelling approaches. Optimisation of an urban drainage system and of the waste water treatment process in particular requires a good knowledge of the waste water composition. As important transformations take place between the emission from the household and the arrival at the treatment facility, sewer models should include these transformations in the sewer system. At present, however, research is still needed in order to increase our knowledge concerning these in-sewer processes. A comparison of the state variables, processes and parameters has been carried out in both sewer models (SMs) and activated sludge models (ASMs). An ASM approach is used for the description of reactions in sewer models. However, a difference is found in the expression for organic material (expressed in terms of BOD) and heterotrophic biomass is absent as a state variable, resulting in differences in processes and parameters. Reconciliation of both the models seems worthwhile and a preliminary solution is suggested in this paper.

Environmental Modelling and Software, 2007

To evaluate the future state of river water in view of actual pollution loading or different mana... more To evaluate the future state of river water in view of actual pollution loading or different management options, water quality models are a useful tool. However, the uncertainty on the model predictions is sometimes too high to draw proper conclusions. Because of the complexity of process based river water quality models, it is best to investigate this problem according to the origin of the uncertainty. If the uncertainty stems from input data or parameter uncertainty, more reliable results are obtained by performing specific measurement campaigns. The aim of the research reported in this paper is to guide these measurement campaigns based on an uncertainty analysis. The practical case study is the river Dender in Flanders, Belgium.

Environmental Modelling and Software, 2007

To evaluate the future state of river water in view of actual pollution loading or different mana... more To evaluate the future state of river water in view of actual pollution loading or different management options, water quality models are a useful tool. However, the uncertainty on the model predictions is sometimes too high to draw proper conclusions. Because of the complexity of process based river water quality models, it is best to investigate this problem according to the origin of the uncertainty. If the uncertainty stems from input data or parameter uncertainty, more reliable results are obtained by performing specific measurement campaigns. The aim of the research reported in this paper is to guide these measurement campaigns based on an uncertainty analysis. The practical case study is the river Dender in Flanders, Belgium.