Establishing a water resources assessment system for Eswatini (Swaziland) incorporating data and modelling uncertainty
- Authors: Ndzabandzaba, Coli
- Date: 2021-10-29
- Subjects: Water resources development Eswatini , Water-supply Eswatini Management , Hydrologic models Eswatini , Runoff Mathematical models , Rain and rainfall Mathematical models , Pitman model
- Language: English
- Type: Doctoral theses , text
- Identifier: http://hdl.handle.net/10962/189009 , vital:44806 , 10.21504/10962/189009
- Description: The uneven distribution of water resources availability globally puts pressure on environmental and human or socio-economic systems and has complex implications for the interactions within these systems. The natural environment and water resources are increasingly threatened by development, and water management crises are still occurring. This is exacerbated by the lack of accurate and adequate information on these systems. In Eswatini, for example, the pressure on the available water resources is mounting due to increasing water demand for irrigation while information about natural hydrological conditions and levels of water resources developments are uncertain. In addition, the practical application of hydrological models for water resources assessments that incorporate uncertainty in Eswatini has yet to be realised. The aim of the study, therefore, was to develop a water resource assessment system that is based on both observed and simulated information and that includes uncertainty. This study focusses on a regional water resource assessment using an uncertainty version of the Pitman monthly rainfall-runoff model whose outputs are constrained by six indices of natural hydrological response (i.e., mean monthly runoff, mean monthly groundwater recharge, Q10, Q50 and Q90 percentage points of the flow duration curve and % time of zero flows) for each of the 122 sub-basins of the transboundary catchments of Eswatini. A 2-step uncertainty modelling approach was tested, validated and then applied to all the sub-basins of Eswatini. The first step of the model run establishes behavioural, but uncertain model parameter ranges for natural incremental sub-basin hydrological responses and the model is typically run 100 000 times for each sub-basin. The parameter space that defines the uncertainty in parameter estimation is sampled based on simple Monte Carlo approach. The second step links all the sub-basin outputs and allows for water use parameters to be incorporated, where necessary, in order to generate cumulative sub-basin outflows. The results from the constraint index analysis have proved to be useful in constraining the model outputs. Generally, the behavioural model outputs produced realistic uncertainty estimates as well as acceptable simulations based on the assessment of the flow duration curves. The modelling results indicated that there is some degree of uncertainty that cannot be easily accounted for due to some identified data issues. The results also showed that there is still a possibility to improve the simulations provided such issues are resolved. The issues about the simulation of stream flow that were detected are mainly related to availability of data to estimate water use parameters. Another challenge in setting up the model was associated with establishing constraints that match the parameters for natural hydrological conditions for specific sub-basins and maintaining consistency in the adjustment of the model output constraints for other sub-basins. In an attempt to overcome this problem, the study recommends additional hydrological response constraints to be used with the Pitman model. Another main recommendation relates to the strong cooperation of relevant catchment management authorities and stakeholders including scientists in order to make information more available to users. The new hydrological insight is derived from the analysis of hydrological indices which highlighted the regional variations in hydrological processes and sub-basin response across the transboundary basins of Eswatini. The adopted modelling approach provides further insight into all the uncertainties associated with quantifying the available water resources of the country. The study has provided further understanding of the spatial variability of the hydrological response and existing development impacts than was previously available. It is envisaged that these new insights will provide an improved basis for future water management in Eswatini. , Thesis (PhD) -- Faculty of Science, Institute for Water Research, 2021
- Full Text:
- Date Issued: 2021-10-29
- Authors: Ndzabandzaba, Coli
- Date: 2021-10-29
- Subjects: Water resources development Eswatini , Water-supply Eswatini Management , Hydrologic models Eswatini , Runoff Mathematical models , Rain and rainfall Mathematical models , Pitman model
- Language: English
- Type: Doctoral theses , text
- Identifier: http://hdl.handle.net/10962/189009 , vital:44806 , 10.21504/10962/189009
- Description: The uneven distribution of water resources availability globally puts pressure on environmental and human or socio-economic systems and has complex implications for the interactions within these systems. The natural environment and water resources are increasingly threatened by development, and water management crises are still occurring. This is exacerbated by the lack of accurate and adequate information on these systems. In Eswatini, for example, the pressure on the available water resources is mounting due to increasing water demand for irrigation while information about natural hydrological conditions and levels of water resources developments are uncertain. In addition, the practical application of hydrological models for water resources assessments that incorporate uncertainty in Eswatini has yet to be realised. The aim of the study, therefore, was to develop a water resource assessment system that is based on both observed and simulated information and that includes uncertainty. This study focusses on a regional water resource assessment using an uncertainty version of the Pitman monthly rainfall-runoff model whose outputs are constrained by six indices of natural hydrological response (i.e., mean monthly runoff, mean monthly groundwater recharge, Q10, Q50 and Q90 percentage points of the flow duration curve and % time of zero flows) for each of the 122 sub-basins of the transboundary catchments of Eswatini. A 2-step uncertainty modelling approach was tested, validated and then applied to all the sub-basins of Eswatini. The first step of the model run establishes behavioural, but uncertain model parameter ranges for natural incremental sub-basin hydrological responses and the model is typically run 100 000 times for each sub-basin. The parameter space that defines the uncertainty in parameter estimation is sampled based on simple Monte Carlo approach. The second step links all the sub-basin outputs and allows for water use parameters to be incorporated, where necessary, in order to generate cumulative sub-basin outflows. The results from the constraint index analysis have proved to be useful in constraining the model outputs. Generally, the behavioural model outputs produced realistic uncertainty estimates as well as acceptable simulations based on the assessment of the flow duration curves. The modelling results indicated that there is some degree of uncertainty that cannot be easily accounted for due to some identified data issues. The results also showed that there is still a possibility to improve the simulations provided such issues are resolved. The issues about the simulation of stream flow that were detected are mainly related to availability of data to estimate water use parameters. Another challenge in setting up the model was associated with establishing constraints that match the parameters for natural hydrological conditions for specific sub-basins and maintaining consistency in the adjustment of the model output constraints for other sub-basins. In an attempt to overcome this problem, the study recommends additional hydrological response constraints to be used with the Pitman model. Another main recommendation relates to the strong cooperation of relevant catchment management authorities and stakeholders including scientists in order to make information more available to users. The new hydrological insight is derived from the analysis of hydrological indices which highlighted the regional variations in hydrological processes and sub-basin response across the transboundary basins of Eswatini. The adopted modelling approach provides further insight into all the uncertainties associated with quantifying the available water resources of the country. The study has provided further understanding of the spatial variability of the hydrological response and existing development impacts than was previously available. It is envisaged that these new insights will provide an improved basis for future water management in Eswatini. , Thesis (PhD) -- Faculty of Science, Institute for Water Research, 2021
- Full Text:
- Date Issued: 2021-10-29
Quantification of water resources uncertainties in two sub-basins of the Limpopo River basin
- Authors: Oosthuizen, Nadia
- Date: 2018
- Subjects: Hydrologic models -- Limpopo River Watershed , Water-supply -- Limpopo River Watershed , Water-supply -- Management , Sustainable development , Rain and rainfall -- Mathematical models , Runoff -- Mathematical models , Reservoirs -- Limpopo River Watershed
- Language: English
- Type: text , Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10962/63267 , vital:28388
- Description: The demand for water is rapidly growing, placing more strain on access to the resources and subsequently its management. For sustainable management, there is a need to accurately quantify the available water resources. Unfortunately, the data required for such assessments are frequently far from sufficient in terms of availability and quality, especially in southern Africa. In the absence of historical observed data, models are generally used to describe the different hydrological processes and generate data and information that will inform management and policy decision making. Ideally, any hydrological model should be based on a sound conceptual understanding of the processes in the basin and be backed by quantitative information for the parameterization of the model. Such data is however, often inadequate in many sub-basins necessitating the incorporation of the uncertainty related to the estimation process. Model parameter estimation and input data are significant sources of uncertainty that should be quantified. Also, in southern Africa water use data are unreliable because available databases consist of licensed information and actual use is generally unknown. In this study, the water resources of two sub-basins of the Limpopo River basin – the Mogalakwena in South Africa and the Shashe shared between Botswana and Zimbabwe – are estimated. The study assessed how uncertainties in the Pitman model parameterisation and input water use data affect the estimation of surface water resources of the selected sub-basins. Farm reservoirs and irrigated areas data from various sources were collected and used to run the Pitman model. Results indicate that the total model output uncertainty is higher for the Shashe sub-basin which is more data scarce than the Mogalakwena sub-basin. The study illustrates the importance of including uncertainty in the water resources assessment process to provide baseline data for decision making in resource management and planning. The study reviews existing information sources associated with the quantification of water balance components and gives an update of water resources of the sub-basin. The flows generated by the model at the outlet of the basin were between 22.6 Mm3 and 24.7 Mm3 per month when incorporating uncertainty to the main physical runoff generating parameters. The total predictive uncertainty of the model increased to between 22.2 Mm3 and 25.0 Mm3 when anthropogenic water use data such as small farm and large reservoirs and irrigation were included. The flows generated for Shashe was between 11.7 Mm3 and 14.5 Mm3 per month when incorporating uncertainty to the main physical runoff generating parameters. The predictive uncertainty of the model changed to 11.7 Mm3 and 17.7 Mm3 after the water use uncertainty was added. However, it is expected that the uncertainty could be reduced by using higher resolution remote sensing imagery.
- Full Text:
- Date Issued: 2018
- Authors: Oosthuizen, Nadia
- Date: 2018
- Subjects: Hydrologic models -- Limpopo River Watershed , Water-supply -- Limpopo River Watershed , Water-supply -- Management , Sustainable development , Rain and rainfall -- Mathematical models , Runoff -- Mathematical models , Reservoirs -- Limpopo River Watershed
- Language: English
- Type: text , Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10962/63267 , vital:28388
- Description: The demand for water is rapidly growing, placing more strain on access to the resources and subsequently its management. For sustainable management, there is a need to accurately quantify the available water resources. Unfortunately, the data required for such assessments are frequently far from sufficient in terms of availability and quality, especially in southern Africa. In the absence of historical observed data, models are generally used to describe the different hydrological processes and generate data and information that will inform management and policy decision making. Ideally, any hydrological model should be based on a sound conceptual understanding of the processes in the basin and be backed by quantitative information for the parameterization of the model. Such data is however, often inadequate in many sub-basins necessitating the incorporation of the uncertainty related to the estimation process. Model parameter estimation and input data are significant sources of uncertainty that should be quantified. Also, in southern Africa water use data are unreliable because available databases consist of licensed information and actual use is generally unknown. In this study, the water resources of two sub-basins of the Limpopo River basin – the Mogalakwena in South Africa and the Shashe shared between Botswana and Zimbabwe – are estimated. The study assessed how uncertainties in the Pitman model parameterisation and input water use data affect the estimation of surface water resources of the selected sub-basins. Farm reservoirs and irrigated areas data from various sources were collected and used to run the Pitman model. Results indicate that the total model output uncertainty is higher for the Shashe sub-basin which is more data scarce than the Mogalakwena sub-basin. The study illustrates the importance of including uncertainty in the water resources assessment process to provide baseline data for decision making in resource management and planning. The study reviews existing information sources associated with the quantification of water balance components and gives an update of water resources of the sub-basin. The flows generated by the model at the outlet of the basin were between 22.6 Mm3 and 24.7 Mm3 per month when incorporating uncertainty to the main physical runoff generating parameters. The total predictive uncertainty of the model increased to between 22.2 Mm3 and 25.0 Mm3 when anthropogenic water use data such as small farm and large reservoirs and irrigation were included. The flows generated for Shashe was between 11.7 Mm3 and 14.5 Mm3 per month when incorporating uncertainty to the main physical runoff generating parameters. The predictive uncertainty of the model changed to 11.7 Mm3 and 17.7 Mm3 after the water use uncertainty was added. However, it is expected that the uncertainty could be reduced by using higher resolution remote sensing imagery.
- Full Text:
- Date Issued: 2018
Understanding and modelling of surface and groundwater interactions
- Authors: Tanner, Jane Louise
- Date: 2014
- Subjects: Groundwater -- South Africa , Water-supply -- Management , Integrated water development , Hydrogeology , Water resources development -- South Africa , Water -- Analysis , Groundwater -- Management , Watersheds -- South Africa , Hydrologic models
- Language: English
- Type: Thesis , Doctoral , PhD
- Identifier: vital:6043 , http://hdl.handle.net/10962/d1012994
- Description: The connections between surface water and groundwater systems remain poorly understood in many catchments throughout the world and yet they are fundamental to effectively managing water resources. Managing water resources in an integrated manner is not straightforward, particularly if both resources are being utilised, and especially in those regions that suffer problems of data scarcity. This study explores some of the principle issues associated with understanding and practically modelling surface and groundwater interactions. In South Africa, there remains much controversy over the most appropriate type of integrated model to be used and the way forward in terms of the development of the discipline; part of the disagreement stems from the fact that we cannot validate models adequately. This is largely due to traditional forms of model testing having limited power as it is difficult to differentiate between the uncertainties within different model structures, different sets of alternative parameter values and in the input data used to run the model. While model structural uncertainties are important to consider, the uncertainty from input data error together with parameter estimation error are often more significant to the overall residual error, and essential to consider if we want to achieve reliable predictions for water resource decisions. While new philosophies and theories on modelling and results validation have been developed (Beven, 2002; Gupta et al., 2008), in many cases models are not only still being validated and compared using sparse and uncertain datasets, but also expected to produce reliable predictions based on the flawed data. The approach in this study is focused on fundamental understanding of hydrological systems rather than calibration based modelling and promotes the use of all the available 'hard' and 'soft' data together with thoughtful conceptual examination of the processes occurring in an environment to ensure as far as possible that a model is generating sensible results by simulating the correct processes. The first part of the thesis focuses on characterising the 'typical' interaction environments found in South Africa. It was found that many traditional perceptual models are not necessarily applicable to South African conditions, largely due to the relative importance of unsaturated zone processes and the complexity of the dominantly fractured rock environments. The interaction environments were categorised into four main 'types' of environment. These include karst, primary, fractured rock (secondary), and alluvial environments. Processes critical to Integrated Water Resource Management (IWRM) were defined within each interaction type as a guideline to setting a model up to realistically represent the dominant processes in the respective settings. The second part of the thesis addressed the application and evaluation of the modified Pitman model (Hughes, 2004), which allows for surface and groundwater interaction behaviour at the catchment scale to be simulated. The issue is whether, given the different sources of uncertainty in the modelling process, we can differentiate one conceptual flow path from another in trying to refine the understanding and consequently have more faith in model predictions. Seven example catchments were selected from around South Africa to assess whether reliable integrated assessments can be carried out given the existing data. Specific catchment perceptual models were used to identify the critical processes occurring in each setting and the Pitman model was assessed on whether it could represent them (structural uncertainty). The available knowledge of specific environments or catchments was then examined in an attempt to resolve the parameter uncertainty present within each catchment and ensure the subsequent model setup was correctly representing the process understanding as far as possible. The confidence in the quantitative results inevitably varied with the amount and quality of the data available. While the model was deemed to be robust based on the behavioural results obtained in the majority of the case studies, in many cases a quantitative validation of the outputs was just not possible based on the available data. In these cases, the model was judged on its ability to represent the conceptualisation of the processes occurring in the catchments. While the lack of appropriate data means there will always be considerable uncertainty surrounding model validation, it can be argued that improved process understanding in an environment can be used to validate model outcomes to a degree, by assessing whether a model is getting the right results for the right reasons. Many water resource decisions are still made without adequate account being taken of the uncertainties inherent in assessing the response of hydrological systems. Certainly, with all the possible sources of uncertainty in a data scarce country such as South Africa, pure calibration based modelling is unlikely to produce reliable information for water resource managers as it can produce the right results for the wrong reasons. Thus it becomes essential to incorporate conceptual thinking into the modelling process, so that at the very least we are able to conclude that a model generates estimates that are consistent with, and reflect, our understanding (however limited) of the catchment processes. It is fairly clear that achieving the optimum model of a hydrological system may be fraught with difficulty, if not impossible. This makes it very difficult from a practitioner's point of view to decide which model and uncertainty estimation method to use. According to Beven (2009), this may be a transitional problem and in the future it may become clearer as we learn more about how to estimate the uncertainties associated with hydrological systems. Until then, a better understanding of the fundamental and most critical hydrogeological processes should be used to critically test and improve model predictions as far as possible. A major focus of the study was to identify whether the modified Pitman model could provide a practical tool for water resource managers by reliably determining the available water resource. The incorporation of surface and groundwater interaction routines seems to have resulted in a more robust and realistic model of basin hydrology. The overall conclusion is that the model, although simplified, is capable of representing the catchment scale processes that occur under most South African conditions.
- Full Text:
- Date Issued: 2014
- Authors: Tanner, Jane Louise
- Date: 2014
- Subjects: Groundwater -- South Africa , Water-supply -- Management , Integrated water development , Hydrogeology , Water resources development -- South Africa , Water -- Analysis , Groundwater -- Management , Watersheds -- South Africa , Hydrologic models
- Language: English
- Type: Thesis , Doctoral , PhD
- Identifier: vital:6043 , http://hdl.handle.net/10962/d1012994
- Description: The connections between surface water and groundwater systems remain poorly understood in many catchments throughout the world and yet they are fundamental to effectively managing water resources. Managing water resources in an integrated manner is not straightforward, particularly if both resources are being utilised, and especially in those regions that suffer problems of data scarcity. This study explores some of the principle issues associated with understanding and practically modelling surface and groundwater interactions. In South Africa, there remains much controversy over the most appropriate type of integrated model to be used and the way forward in terms of the development of the discipline; part of the disagreement stems from the fact that we cannot validate models adequately. This is largely due to traditional forms of model testing having limited power as it is difficult to differentiate between the uncertainties within different model structures, different sets of alternative parameter values and in the input data used to run the model. While model structural uncertainties are important to consider, the uncertainty from input data error together with parameter estimation error are often more significant to the overall residual error, and essential to consider if we want to achieve reliable predictions for water resource decisions. While new philosophies and theories on modelling and results validation have been developed (Beven, 2002; Gupta et al., 2008), in many cases models are not only still being validated and compared using sparse and uncertain datasets, but also expected to produce reliable predictions based on the flawed data. The approach in this study is focused on fundamental understanding of hydrological systems rather than calibration based modelling and promotes the use of all the available 'hard' and 'soft' data together with thoughtful conceptual examination of the processes occurring in an environment to ensure as far as possible that a model is generating sensible results by simulating the correct processes. The first part of the thesis focuses on characterising the 'typical' interaction environments found in South Africa. It was found that many traditional perceptual models are not necessarily applicable to South African conditions, largely due to the relative importance of unsaturated zone processes and the complexity of the dominantly fractured rock environments. The interaction environments were categorised into four main 'types' of environment. These include karst, primary, fractured rock (secondary), and alluvial environments. Processes critical to Integrated Water Resource Management (IWRM) were defined within each interaction type as a guideline to setting a model up to realistically represent the dominant processes in the respective settings. The second part of the thesis addressed the application and evaluation of the modified Pitman model (Hughes, 2004), which allows for surface and groundwater interaction behaviour at the catchment scale to be simulated. The issue is whether, given the different sources of uncertainty in the modelling process, we can differentiate one conceptual flow path from another in trying to refine the understanding and consequently have more faith in model predictions. Seven example catchments were selected from around South Africa to assess whether reliable integrated assessments can be carried out given the existing data. Specific catchment perceptual models were used to identify the critical processes occurring in each setting and the Pitman model was assessed on whether it could represent them (structural uncertainty). The available knowledge of specific environments or catchments was then examined in an attempt to resolve the parameter uncertainty present within each catchment and ensure the subsequent model setup was correctly representing the process understanding as far as possible. The confidence in the quantitative results inevitably varied with the amount and quality of the data available. While the model was deemed to be robust based on the behavioural results obtained in the majority of the case studies, in many cases a quantitative validation of the outputs was just not possible based on the available data. In these cases, the model was judged on its ability to represent the conceptualisation of the processes occurring in the catchments. While the lack of appropriate data means there will always be considerable uncertainty surrounding model validation, it can be argued that improved process understanding in an environment can be used to validate model outcomes to a degree, by assessing whether a model is getting the right results for the right reasons. Many water resource decisions are still made without adequate account being taken of the uncertainties inherent in assessing the response of hydrological systems. Certainly, with all the possible sources of uncertainty in a data scarce country such as South Africa, pure calibration based modelling is unlikely to produce reliable information for water resource managers as it can produce the right results for the wrong reasons. Thus it becomes essential to incorporate conceptual thinking into the modelling process, so that at the very least we are able to conclude that a model generates estimates that are consistent with, and reflect, our understanding (however limited) of the catchment processes. It is fairly clear that achieving the optimum model of a hydrological system may be fraught with difficulty, if not impossible. This makes it very difficult from a practitioner's point of view to decide which model and uncertainty estimation method to use. According to Beven (2009), this may be a transitional problem and in the future it may become clearer as we learn more about how to estimate the uncertainties associated with hydrological systems. Until then, a better understanding of the fundamental and most critical hydrogeological processes should be used to critically test and improve model predictions as far as possible. A major focus of the study was to identify whether the modified Pitman model could provide a practical tool for water resource managers by reliably determining the available water resource. The incorporation of surface and groundwater interaction routines seems to have resulted in a more robust and realistic model of basin hydrology. The overall conclusion is that the model, although simplified, is capable of representing the catchment scale processes that occur under most South African conditions.
- Full Text:
- Date Issued: 2014
- «
- ‹
- 1
- ›
- »