Browsing by Author "Clark, David John."

Now showing 1 - 8 of 8

Assessment of satellite derived rainfall and its use in the ACRU hydrological model.
(2017) Suleman, Shuaib.; Chetty, Kershani Tinisha.; Clark, David John.
Many parts of southern Africa are considered water scarce regions. Therefore, sound management and decision making is important to achieve maximum usage with sustainability of the precious resource. Hydrological models are often used to inform management decisions; however model performance is directly linked to the quality of data that is input. Rainfall is a key aspect of hydrological systems. Understanding the spatial and temporal variations of rainfall is of paramount importance to make key management decisions within a management area. Rainfall is traditionally measured through the use of in-situ rain gauge measurements. However, rain gauge measurements poorly represent the spatial variations of rainfall and rain gauge networks are diminishing, especially in southern Africa. Due to the sparse distribution of rain gauges and the spatial problems associated with rain gauge measurements, the use of satellite derived rainfall is being increasingly advocated. The overall aim of this research study was to investigate the use of satellite derived rainfall into the ACRU hydrological model to simulate streamflow. Key objectives of the study included (i) the validation of satellite derived rainfall with rain gauge measurements, (ii) generation of time series of satellite derived rainfall to drive the ACRU hydrological model, and (iii) validation of simulated streamflow with measured streamflow. The products were evaluated in the upper uMngeni, upper uThukela (summer rainfall) as well as the upper and central Breede catchments (winter rainfall). The satellite rainfall products chosen for investigation in this study included TRMM 3B42, FEWS ARC2, FEWS RFE2, TAMSAT-3 and GPM. The satellite rainfall products were validated using rain gauges in and around the study sites from 1 January 2010 to 30 April 2017. The rainfall products performed differently at each location with high variation in daily magnitudes of rainfall. Total rainfall volumes over the period of analysis were generally in better agreement with rain gauge volumes with TRMM 3B42 tending to overestimate rainfall volumes whereas the other products underestimated rainfall volumes. The ACRU model was applied using satellite rainfall and rain gauge measurements in the aforementioned study catchments from 1 October 2007 to 30 September 2016. Streamflow results were generally poor and variable amongst products. Daily correlations of streamflow were poor. Total streamflow volumes were in better agreement with total volumes of observed streamflow. TRMM 3B42 and rain gauge driven simulations produced the best results in the summer rainfall region, whereas the FEWS driven simulations produced the best results in the winter rainfall region.
Development and assessment of an improved continuous simulation modelling system for design flood estimation in South Africa using the ACRU model.
(2019) Rowe, Thomas James.; Smithers, Jeffrey Colin.; Clark, David John.
An estimate of the risk associated with flood events is required to adequately design hydraulic structures and limit negative socio-economic impacts as a result of floods. The methods used to estimate design floods in South Africa are outdated and are in need of revision. A National Flood Studies Programme (NFSP) has recently been initiated by Smithers et al. (2016) to overhaul Design Flood Estimation (DFE) procedures in South Africa. One of the recommendations of the NFSP is development and assessment of a Continuous Simulation Modelling (CSM) approach to DFE. Consequently, the aim of this study is to further develop and assess the performance of an improved comprehensive CSM system, to consistently and reliably estimate design flood discharges in small catchments (0 - 100 km2) in South Africa using the ACRU model. In the development of the approach a strong emphasis has been placed on ease of use from a practitioner’s point of view. The aim is achieved through several specific objectives as summarised below. The first objective was to review CSM approaches applied locally and internationally for DFE, in order to identify research gaps and guide the development of an improved national CSM system for DFE in South Africa. The review culminates with a list of recommendations and steps required to develop and adopt a CSM approach for DFE in practice. The first critical step identified and required was the development of a comprehensive CSM system using the ACRU model (Schulze, 1995). This included: the structure of the system and how to implement the system, an enhanced land cover and soils classification to apply with the system and default input information and databases to use with the system. The second objective addresses the recommendations made from the literature review, where a comprehensive CSM system for DFE using the ACRU model is developed and described in detail. Based on similarities identified between the ACRU (Schulze, 1995) and SCS-SA models (Schmidt and Schulze, 1987a), as well as the fact that the SCS-SA model is relatively simple and widely applied in practice, the CSM system was adapted to be consistent with the land cover classification used in the SCS-SA model. This included the incorporation of a methodology and rules, developed by Rowe (2015), to represent land management practices and hydrological conditions within the ACRU model. The development of this comprehensive CSM system with default national scale inputs and land cover classifications contributes to new knowledge on how to package a CSM system for DFE in South Africa. The third objective focuses on the assessment and verification of the CSM system developed, using observed data. Through the verifications and assessments performed an inconsistency between daily simulated stormflow volumes and the volume of stormflow used in the daily stormflow peak discharge equation was identified. Therefore, a revision, which is more conceptually correct than the current assumption that all stormflow generated from an event contributes to the peak discharge on the day, was applied to the fraction of the simulated daily stormflow used in the peak discharge equation. This corrected the inconsistency and significantly improved the results, thereby providing an improved methodology to more accurately estimate peak discharges in the ACRU model than had hitherto been the case. Despite the improvement in the results, a general over-simulation of peak discharges was still evident. Consequently, further investigation of the ACRU stormflow peak discharge computations was performed in order to identify which approach provides the most satisfactory results (Objective 4). This included a performance assessment of both the SCS single Unit Hydrograph (UH) approach and the incremental UH approach. The performance of each approach was assessed using both estimated parameters and parameters derived from observed data. These parameters include stormflow volumes, catchment lag times, and the distribution of daily rainfall, where applicable, to each approach. Comparison of the results from the two approaches indicated that more accurate results are obtained when applying the incremental UH approach, when using both estimated or observed parameter inputs. In terms of the incremental UH approach, it was identified that the approach is more sensitive to the use of synthetic daily rainfall distributions compared to estimated lag times. Based on the results obtained new knowledge and additional research gaps related to: (i) improved estimation of the distribution of daily rainfall within the ACRU model, (ii) links between the distribution of daily rainfall and catchment lag time, and (iii) the need to further verify and possibly recalibrate CNs for South Africa were identified. The fifth objective addressed is an assessment of the impact of model configuration on the performance of the ACRU CSM system developed, in order to propose a final CSM system for DFE in South Africa. Results when using site-specific land cover and soils information are compared to those obtained when different sources of input information are used, such as the national land cover and soils maps developed for the entire country. The results when using these default national datasets were not particularly good, however recommendations are made to improve on the results. In addition, the most appropriate current databases to use with the CSM system are defined, providing users with the most appropriate default information currently available to use in the absence of site-specific information. The last objective addressed was a comparison of the performance of the final ACRU CSM system proposed in this study to that of the widely applied SCS-SA model and associated approaches, when using the same input information. Ultimately, the final ACRU CSM system proposed provides results that are superior to those from the SCS-SA model and associated approaches. In addition, several advantages of the ACRU CSM system over the traditional SCS-SA approaches were identified. Recommendations were, however, made to improve on the CSM system developed in this study and to use the results to update the SCS-SA model. New knowledge on the performance of the SCS-SA model and its associated approaches compared to that of the comprehensive CSM system developed for South Africa is therefore provided in this study.
Development and assessment of an integrated largescale hydrological modelling tool for water resources management in the Cauvery Catchment, India.
(2022) Horan, Robyn.; Smithers, Jeffrey Colin.; Kjeldsen, Thomas Rodding.; Clark, David John.; Toucher, Michele Lynn.
Economic development and population growth in southern India have resulted in rapid changes to land use, land management and water demand, significantly impacting and degrading water resources. The significant anthropogenic influences across the catchment have contributed to changes in hydrological functioning. Focussing on the highly contentious inter-state Cauvery River Catchment, this study aims to address the key scientific challenges faced within this catchment. The study was designed to develop an integrated large-scale hydrological model to improve water resource assessments in a highly heterogeneous and data-scarce region whilst considering the primary water resource challenges facing the Cauvery Catchment. The Upper Cauvery region, located in the Western Ghats, acts as the water tower of the catchment. The rainfall in the region is monsoonal, the topography is complex, and the rain gauge network is sparse, resulting in the estimation of rainfall being particularly challenging. The scarce rainfall data available in the Western Ghats region is hindering the understanding of the regional weather system, and the accepted rainfall dataset for India, Indian Meteorological Department rainfall grids, are known to have inaccurate estimations within the Western Ghats. The current knowledge of the meteorology and hydrology of the Upper Cauvery is limited. Additionally, the anthropogenic impact on local hydrological processes, such as streamflow, groundwater recharge and evapotranspiration, is poorly constrained. The current understanding of how these diverse local changes cumulatively impact water availability at the broader catchment scale is minimal. Small-scale rural water management and urban heterogeneity may strongly affect water resource availability across southern India. However, how such fine-scale factors propagate to the river catchment is largely unclear. The Global Water Availability Assessment (GWAVA) model was applied initially to the Upper Cauvery region to determine the suitability and compare model results from other modelling tools applied in the region. Two new versions of the GWAVA model were then developed. The first aimed to include small-scale runoff harvesting interventions (SSRHIs) into the model and quantify their impact on catchment water resources to address a renewed scientific interest in assessing their effectiveness in improving local water resources and the effects at a catchment scale. The second aimed to enhance the representation of groundwater and large operational dams whilst maintaining the model’s applicability to regions with low-data availability. The Indian Meteorological Department (IMD) gridded rainfall was compared to available gauges and selected remotely sensed datasets within the Upper Cauvery region. GWAVA will be utilised to assess the applicability of the remotely sensed data for a catchment rainfall estimation. GWAVA was determined to be a suitable tool to represent the Cauvery Catchment; however, the importance of an accurate spatial representation of rainfall for input into hydrological models and that comprehensive dam functionality is paramount to obtaining good results in this region was highlighted. Furthermore, the average GWAVA, VIC and SWAT ensemble provided a better predictive ability in catchments with dams than the individual models. The average ensemble offset uncertainty in input data and poor dam operation functionality within individual models. The inclusion of SSRHIs demonstrated that farm bunds appear to have a negligible effect on the average annual simulated streamflow. In contrast, tanks and check dams have a more significant and time-varying impact. The open water surface of the SSRHIs contributed to an increase in evaporation losses across the sub-catchment. The change in simulated groundwater storage with the inclusion of SSRHIs was not as significant as sub-catchmentscale literature, and field studies suggest. Including groundwater processes into GWAVA improved streamflow simulation in the headwater sub-catchments and the representation of the baseflow component such that low-flow model skill increased approximately 33-67% in the Cauvery and 66-100% in the Narmada. The existing dam routine was extended to account for large, regulated dams with two calibratable parameters. The routine improved streamflow simulation in sub-catchments downstream of major dams, where the streamflow was largely reflective of dam releases. The model performance was improved between 15 and 30% in the Cauvery and 7-30% in the Narmada when the regulated dams were considered. The model provides a more robust representation of the annual outflow volume from major dams, reducing the average bias from -17% to -1% in the Cauvery and from 14% to 3% in the Narmada. The daily dam releases were significantly improved in the Cauvery, approximately 26-164%. The improvement of the groundwater and dam routines in GWAVA proved successful in improving the overall model performance, the low-flow model skill and bias, and the inclusions allowed for improved traceability of simulated water balance components. It was found that the IMD rainfall within the high-altitude regions of the Western Ghats is underestimated, resulting in the under-simulation of streamflow in the Upper Cauvery. CHIRPS 0.25- and 0.05- degree, MSWEP and PERSIANN remotely sensed rainfall datasets were applied within this region. None of the individual rainfall datasets provided a more accurate representation of the rainfall than the commonly utilised IMD grids. However, using an ensemble of remotely sensed rainfall datasets, primarily the average ensemble, improved the accuracy of rainfall estimation in the catchment. The ‘off-the-shelf’ remotely sensed rainfall products provided a high variation in performance against the in-situ rain gauge data. The IMD grids provided the most accurate representation of rainfall compared to the individual remotely sensed rainfall datasets, despite underestimating the rainfall depths at high altitudes. In the case of the Upper Cauvery, the average ensemble provided a more accurate representation of the rainfall. An integrated large-scale hydrological model was developed to improve water resources assessments in a highly heterogeneous and data-scarce region whilst considering the major water resource challenges facing the Cauvery Catchment. The effects of runoff harvesting interventions, accounting for hard-rock aquifer groundwater processes and the impact of major dams were represented. The inclusion of these features improved the model performance throughout the Cauvery Catchment.
Development of a geographic data model for hydrological modelling.
(2006) Bollaert, M. J.; Clark, David John.
Hydrology is a multi-disciplinary science, and therefore derives data from diverse sources, with the data often of a spatio-temporal nature. A recent trend has been to combine these data with GIS, due to the data’s geographic origin, and inherently requires an abstraction of reality in order to deal with the multitude of data that would otherwise result. Consequently, data models have been developed for this purpose, and these require a generalisation of processes and variables, while offering a simplified structure for their storage. The purpose of this study was to develop a data model for the storage and dissemination of hydrological variables and associated data used in hydrological modelling. Data would be of a spatial and temporal nature, and thus the design of the new data model needed to provide for this. A number of existing geographic data models were therefore reviewed, including the geodatabase model. This data model and the object-relational database model upon which it was built, were ascertained as being the most suitable for the study, and were therefore included in the design of the new data model. The related Arc Hydro data model was subsequently reviewed, since it offered an established means by which to model geographic features associated with surface hydrology. Following this, an investigation into time series storage methods was carried out, as it was important that the new data model be able to store large time series datasets in an efficient manner. Thus a number of methods were identified and evaluated as to their advantages and disadvantages. A new data model was thereby conceived, using the geodatabase as its foundation, and was developed in order to offer efficient storage of hydrological data. The data model developed was subsequently tested by populating it with data from the Quaternary Catchments database which supports the ACRU model. Finally, additional functionality was added to the data model, in the form of export options.
The effect of spatial resolution in remote sensing estimates of total evaporation in the uMgeni catchment.
(2014) Shoko, Cletah.; Clark, David John.; Bulcock, Hartley Hugh.; Mengistu, Michael Ghebrekidan.
The estimation of total evaporation plays a vital role in water resources monitoring and management, especially in water-limited environments. In South Africa, the increasing water demand, due to population growth and economic development, threatens the long-term water supply. This, therefore, underscores the need to account for water by different consumers, for well-informed management, allocation and future planning. Currently, there are different methods (i.e. ground-based and remote sensing-based methods), which have been developed and implemented to quantify total evaporation at different spatial and temporal scales. However, previous studies have shown that ground-based methods are inadequate for understanding the spatial variations of total evaporation, within a heterogeneous landscape; they only represent a small area, when compared to remotely sensed methods. The advent of remote sensing therefore provides an invaluable opportunity for the spatial characterization of total evaporation at different spatial scales. This study is primarily aimed at estimating variations of total evaporation across a heterogeneous catchment in KwaZulu-Natal, South Africa, using remote sensing data. The first part provides an overview of total evaporation, its importance within the water balance and consequently in the management of water resources. It also covers various methods developed to estimate total evaporation, highlighting their applications, limitations, and finally, the need for further research. Secondly, the study determines the effect of sensor spatial resolution in estimating variations of total evaporation within a heterogeneous uMngeni Catchment. Total evaporation estimates were derived, using multispectral 30 m Landsat 8 and 1000 m MODIS, based on the Surface Energy Balance (SEBS) model. The results have shown that different sensors, with varying spatial resolutions, have different abilities in representing variations of total evaporation at catchment scale. It was found that Landsat-based estimates were significantly different (p < 0.05) from MODIS. The study finally estimates spatial variations of total evaporation from Landsat 8 and MODIS datasets for the uMngeni Catchment. It was found that the Landsat 8 dataset has greater potential for the detection of spatial variations of total evaporation, when compared to the MODIS dataset. For instance, MODIS-based daily total evaporation estimates did not show any significant difference across different land cover types (One way ANOVA; F1.924 = 1.412, p= 0.186), when compared to the 30 m Landsat 8, which yielded significantly different estimates between different land cover types (One way ANOVA; F1.993= 5.185, p < 0.001). The validation results further indicate that Landsat-based estimates were more comparable to ground-based eddy covariance measurements (R2 = 0.72, with a RMSE of 32.34 mm per month (30.30% of the mean)). In contrast, MODIS performed poorly (R2 = 0.44), with a RMSE of 93.63 mm per month (87.74% of the mean). In addition, land cover-based estimates have shown that, not only does the land cover type have an effect on total evaporation, but also the land cover characteristics, such as areal extent and patchiness. Overall, findings from this study underscore the importance of the sensor type, especially spatial resolution, and land cover type characteristics, such as areal extent and patchiness, in accurately and reliably estimating total evaporation at a catchment scale. It is also evident from the study that the spatial and temporal variations in SEBS inputs (e.g., LAI, NDVI and FVC) and energy fluxes (e.g., Rn) calculated by SEBS for the two sensors can affect the spatial and temporal variations in total evaporation estimates. For instance, spatial variations in total evaporation reflected similar spatial variations in Rn. Areas with high NDVI, FVC and LAI (which denotes dense vegetation cover) tend to have higher total evaporation estimates, compared to areas with lower vegetation cover. In addition, the MODIS sensor at 1000 m spatial resolution showed lower estimates of SEBS inputs with less variability across the catchment. This resulted in lower total evaporation estimates, with less variability, compared to the 30 m Landsat 8. In addition, with regard to inputs derived from remote sensing, it was found that the spatial variations in total evaporation are not determined by individual variables (e.g., LST), but are influenced by a combination of many biophysical variables, such as LAI, FVC and NDVI. These findings lay a foundation for a better approach to estimate total evaporation using remote sensing for use in the management and allocation of water.
The estimation and evaluation of a satellite-based drought index using rainfall and evapotranspiration.
(2017) Mahomed, Maqsooda.; Chetty, Kershani Tinisha.; Clark, David John.
Abstract available in PDF file.
An evaluation of priority and fractional methods of water allocation in the Sand River catchment, South Africa.
(2014) Winckworth, Ross.; Smithers, Jeffrey Colin.; Clark, David John.
The development and apportionment of water resources is a critical issue, both globally and locally in South Africa. This is particularly true in the development and allocation among states sharing watercourse systems. The competition inherent in access to water resources is increasing. In particular, pressure is being placed on water resources from several activities including irrigation, domestic consumption and industrial requirements. Water allocation mechanisms are therefore critical to sustain the existing allocatable water resources while attempting to combine both efficiency and equity principles. The National Water Act of South Africa (Act 36 of 1998) (NWA (36, 1998)) incorporates both institutional and legal policy which promotes the efficient, equitable and sustainable management of water resources. The aims of the NWA (36, 1998) are achieved by a movement away from a Riparian Rights system (a property adjacent to a water course is allowed reasonable use) to an Administrative System (Hallowes et al., 2008). The inception of an Administrative System for the allocation of water in South Africa is vital given that a number of catchments in South Africa have reached a state of being fully developed and more than 50% of the 19 water management areas in South Africa are water stressed, i.e. the demand exceeds the supply (DWAF, 2004). The NWA (36, 1998) makes allowance for only one right to water; that being the Reserve, which consists of two components, the ecological requirement and basic human needs. The management of the resource is important because the NWA (36, 1998) states that the water resources within South Africa are to be protected, used, developed, conserved, managed and controlled in accordance with the National Water Resources Strategy (DWAF, 2004). The water allocation method currently applied in South Africa is referred to as a Prioritybased River and Reservoir Operating Rule (PRROR) institutional arrangement. Under PRROR, when there is a risk of a reservoir or river failing to meet the supply demanded, restrictions are applied to abstractions. The priority extends not only to those who have the priority of use but which users will relinquish water to the higher priority users and by what quantity. Disadvantages of PRROR include the inability of the Water User to manage their water to meet their needs and are then forced into using it when the water is available. Possible alternate allocation methods include Fractional Water Allocation and Capacity Sharing (FWACS), public water allocation and prior rights systems. The PRROR as currently implemented leads to high priority sectors having dominance over access to water which may lead to those sectors not using water efficiently. The introduction of FWACS creates an atmosphere of water awareness and being responsible for managing water use. In this study, the MIKE BASIN model was used in the simulation of the processes of the PRROR and the FWACS allocation methods. The model routes water based on rules specified for the allocation method under review. The efficiency of each allocation method was evaluated in terms of the reliability of supply to Water Users. In the catchment used as a case study (Sand River Catchment), limited information on Environmental Water Requirement (EWR) was available and the EWRs were set as minimum flows at each reservoir and then set as a minimum flow requirement at a downstream node to prevent Water Users downstream of the dam from immediately abstracting the EWR release. Based on data used in the case study and the rules applied to each scenario, the results from the initial study indicated that PRROR provides a 4% higher reliability of supply in comparison to FWACS in the catchment under investigation. This is true when the supply to a Water User is similar between scenarios. However, if the fractions allocated in FWACS are varied away from this baseline, results indicate that a 50% increase on the original FWACS fractions provides for better reliability of supply. Thus the results show that although PRROR is an alternative method for determining water allocation to water users, FWACS+50 is able to improve on the water reliability of supply within the Sand River Catchment.
Object-orientation and integration for modelling water resource systems using the ACRU model.
(2018) Clark, David John.; Smithers, Jeffrey Colin.; Jewitt, Graham Paul Wyndham.
Water is a limiting resource in South Africa, with demand in many catchments exceeding supply, necessitating transfers of water between catchments. This situation requires detailed and integrated management of the country’s water resources, considering environmental, social and economic aspects as outlined in the National Water Act (Act 36 of 1998). Integrated water resources management (IWRM) will require better data and information through monitoring and integrated water resources modelling. The ACRU hydrological model is an important repository of South African water research and knowledge. In recent years there have been technological advances in computer programming techniques and model integration. The thesis for this study was that the valuable knowledge already existing in the ACRU model could be leveraged to provide a better hydrological model to support IWRM in South Africa by: (i) restructuring the model using object-oriented design and programming techniques, and (ii) implementing a model interface standard. Object-oriented restructuring of the ACRU model resulted in a more flexible model enabling better representation of complex water resource systems. The restructuring also resulted in a more extensible model to facilitate the inclusion of new modules and improved data handling. A new model input structure was developed using Extensible Markup Language (XML) to complement the object-oriented structure of the ACRU model. It was recognised that different models have different purposes and strengths. The OpenMI 2.0 model interface standard was implemented for ACRU, enabling ntegration with other OpenMI 2.0 compliant specialised models representing different domains to provide a more holistic IWRM view of water resource systems. Model integration using OpenMI was demonstrated by linking ACRU to the eWater Source river network model. A case study in the upper uMngeni Catchment in South Africa demonstrated: (i) the benefits of the object-oriented design of the restructured ACRU model, in the context of using ACRU to create modelled catchment-scale water resource accounts, and (ii) the integration of ACRU with another model using OpenMI. The case study also demonstrated that despite the improvements to the ACRU model, the simulations are only as good as the model input data.