Please use this identifier to cite or link to this item:
Title: Integrated hydrological modelling for sustainable water allocation planning : Mkomazi Basin, South Africa case study
Authors: Amoo, Taiwo Oseni 
Issue Date: 2018
Allocation of freshwater resources between societal needs and natural ecological systems is of great concern for water managers. This development has challenged decision-makers regarding how to reasonably allocate available water resources to meet the competing demands. Thus, turning these concerns into opportunities requires the need for both water technology innovation and water behavioural change, in order to manage fresh water in a sustainable manner.
This study aimed at investigating the applicability of an integrated hydrological model in a Geographical Information Systems (GIS) environment for sustainable water allocation planning and management, using the Mkomazi Basin in KwaZulu-Natal Province, South Africa, as a case study. Specifically, the study identified ecosystems that depend on Mkomazi River for preservation of their environmental and public benefit values; developed a water allocation mechanism to achieve equitable water distribution and large benefits from water uses across the basin’s users; synthesised rules for sustainable development in sharing the limited water resources and maintaining environmental quality; and finally, established a framework for water trading in order to encourage water use efficiency and allow movement of water to new users.
Historical 15-year (1990-2015) observed streamflows and daily meteorological variables (precipitation sums (mm), relative humidity (%), wind speed (m s_1), mean, minimum and maximum air temperature (oC), solar radiation (MJ/m2), sunshine duration (h) and evapotranspiration (mm)) were used for this study. The hydro-meteorological data collected from various sources were subjected to frequency trend analysis, correlation, regression and the double mass curve to test their accuracy, reliability, homogeneity, consistency and localisation gaps. The ombro-thermic diagram was used to classify the months into wet and dry periods.
The identification of prominent ecosystems that depend on the Mkomazi River was achieved through a comprehensive desktop survey and documentation acquired from the Department of Water and Sanitation (DWS). Multivariate statistical methods; cluster, factor and principal component analysis, were applied to analyse the surface water quality data sets extracted from the repository of South Africa’s water resources website (WR2012), in other assess their impact on the aquatic net benefit values and environmental preservation.
A semi-distributed event process and an integrated Soil and Water Assessment Tool (SWAT) model in a GIS environment, with descriptive statistical of mean, median, mode, standard deviation, skewness, and kurtosis were employed to simulate the basin’s hydrological process in evaluating the basin’s water balance. The SWAT model was parameterised, calibrated and validated from corrected hydro-meteorological data from 2004 to 2013. Sequential Uncertainty Fitting Algorithm (SUFI2) was used for the model sensitivity analysis, calibration and validation of the model.
Artificial Neural Networks (ANNs), Probability Distribution Functions (PDF), and a Flow Duration Curve (FDC) were used to project future available water. Based on the estimated available water, an estimation of allocable water was made based on percentage dependability of the river yield to the different users. The weilbul ranking was used for choosing the dependable flow; this was subsequently used for the different water riparian’s demand distribution. Large benefits derivation from water uses across the basin’s users was based on priority-driven sustainability.
Extensive literature review work was used to synthesise rules for sharing limited water resources and maintaining environmental quality for sustainable development. These rules were all drawn from similar world experiences for efficient and gainful utilisation of water and other natural resources. The synthesised rules and principles were modified to suit
KwaZulu-Natal Province (KZN) water allocation reform regulations. The established water rules were subsequently adapted to the present (Mkomazi) case study area.
The proposed developed water trading framework leans on an inclusive simulation of ‘Hydrology, Environment, Life (aquatics), Policy and Sensitivity’ (HELPS) collective response of the basin in exploring the socio-economic and environmental consequences of water regulation. It uses a System Dynamic (SD) simulation technique to form a composite supply-side augmentation with demand-side improvement system to allow movement of water to new users and encourage water use efficiency.
The results of the agglomerative hierarchical cluster analysis grouped the 10 sub-basin sites into three clusters of highly polluted (HP), medium polluted (MP) and relatively less polluted (LP) group basins with latent factors of 81.9, 3.14 and 0.858 (%) in the total water quality variance data sets. The water quality index analysis shows a mild effect on irrigation farming and aquatic species.
The results of water balance simulation show that mean monthly values were 28.6 m3/s over the years with Nash-Sutcliffe Efficiency (NSE) values of 0.83 and a coefficient of determination (R²) of 0.77 at validation stage. The Curve Number (CN) is the most sensitive parameter for the estimation of both streamflow and water yield within the catchment. Other water balance simulation ratios include: Streamflow/precipitation (0.4 mm); Baseflow/Total flow (0.67 mm); Surface Runoff/Total flow (0.33 mm); Percolation/precipitation (0.20 mm); Deep recharge/precipitation (0.01 mm) with an Evapotranspiration/precipitation ratio of 0.58 mm respectively. The water allocation results in the different dependable flow rates of 60%, 70%, and 85% reliability revealed it to be 17465.56, 8068.04 and 6373.35 (m3/s) at U1H009 discharge station, respectively.
The synthesised literature rules suggest water allocation reform acts should be catalysed
through the institutionalisation of capacity developmental platforms where climate change transformation experts and other stakeholders have input in legislating water reform acts, which should be supported by a strong political will.
The invented SD framework confirms agricultural water use as the highest demand when compared with other users. Its sustainability index was evaluated as the ratio of aggregated possible water demand relative to the corresponding supply in the same period. The result shows an integrated scenario which combines rainfall variation with improved irrigation water use efficiency and gives the optimal sustainability performance index (0.25) of the system at 70% dependable flow.
The simulated water balance results also reveal the use of scientific visualisation techniques in QSWAT to model spatially distributed and time-varying hydrologic-meteorological data sets in evaluating the water balance, while its calibration and validation in SWAT Calibration Uncertainty Procedure (SWAT-CUP) algorithm connotes a strong model efficiency performance.
The developed SD framework provides comprehensive assessment methodology for the decision-maker in assessing water trading. The applied integrated model can be used in similar river basins sharing related attributes to the study area in resolving the current water – stressed challenges in South Africa as well as other regions of the globe.
Considering the extent of the drought and the paucity of the uneven allocation of water resources at the study area, the needfulness of integrated hydrological models such as SWAT and ANNs cannot be overemphasised in ensuring the sustainability of Mkomazi Basin, while unlocking the untapped potential of water resources for the development of the agricultural and industrial sectors, and still meeting the requirements of the ecosystem.
Submitted in fulfilment of the requirements for the degree of Doctor of Engineering: Civil Engineering,
Durban University of Technology, Durban, South Africa, 2018.
Appears in Collections:Theses and dissertations (Engineering and Built Environment)

Files in This Item:
File Description SizeFormat
Amoo_2018.pdf5.07 MBAdobe PDFThumbnail
Show full item record

Page view(s)

checked on Jul 16, 2024


checked on Jul 16, 2024

Google ScholarTM




Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.