[Portion of abstract]: Sediment dynamics are influenced by transformed landscape connectivity in catchments worldwide. The upper Thina catchment, an important high rainfall resource in the northern Eastern Cape, South Africa, is an example of where ongoing subsistence farming on communal land has led to overgrazing and trampling that has initiated large erosive features (e.g. gullies) and river incision. The formation of gullies led to increased hillslope-channel connectivity and the resultant river incision decreased the channel-valley fill connectivity. These two changes in connectivity led to increased sediment export from the catchment that has various down-stream ecological and socio-economic impacts. This study investigates how the change in hillslope-channel and channel-valley fill connectivity has altered the sediment dynamics in the Vuvu catchment, a headwater tributary of the Thina River. A combination of methods were used to assess the changes in hillslope-channel and channel-valley fill connectivity. High resolution aerial images were used to map source features, such as fields, gullies, sheet erosion, landslides, roads and livestock tracks. Topographic and geological characteristics of the source features were extracted using a Geographic Information System. Furthermore, hillslope-channel pathways, such as the natural drainage network, continuous gullies, discontinuous gullies, roads and livestock tracks were mapped and analysed in terms of topographic and geological characteristics. Historic aerial images were assessed to calculate the date the larger gullies began forming. Recent aerial photos and cross sectional surveys of the valley fill were combined to map the various sediment sinks. Particle size and organic content were analysed for flood bench cores and terrace samples. The chronology of the flood benches was determined using unsupported Pb-210 and Cs-137 dating, and determined for the terraces using Optically Stimulated Luminescence dating. Quantitative and qualitative sediment tracing approaches, using mineral magnetic properties, were used to trace the origin of suspended sediment (collected during flood events), sediment stored in the flood benches and sediment stored in the terraces. Hydrological monitoring was used to assess the potential to store sediment on flood benches along the valley fill through flood bench inundation frequency. Hydrological and hydraulic modelling extended the measured inundation frequencies to a 73 year period and other cross sections along the valley fill. Furthermore, a future scenario of an increased vegetation cover and reduced hillslope-channel connectivity was assessed in terms of channel-valley fill inundation frequency.

PhDH / Department of Hydrology and Water Resources / Demand for water resources has been on the increase and is compounded by population growth and related development demands. Thus, numerous sectors are affected by water scarcity and therefore effective management of drought-induced water deficit is of importance. Luvuvhu River Catchment (LRC), a tributary of the Limpopo River Basin in South Africa has witnessed an increasing frequency of drought events over the recent decades. Drought impacts negatively on communities’ livelihoods, development, economy, water resources, and agricultural yields. Drought assessment in terms of frequency and severity using Drought Indices (DI) in different parts of the world has been reported. However, the forecasting and prediction component which is significant in drought preparedness and setting up early warning systems is still inadequate in several regions of the world. This study aimed at characterising, assessing, and predicting drought conditions using DI as a drought quantifying parameter in the LRC. This was achieved through the application of hybrid statistical and machine learning models including predictions via a combination of hybrid models. Rainfall and temperature data were obtained from South African Weather Service, evapotranspiration, streamflow, and reservoir storage data were obtained from the Department of Water and Sanitation while root zone soil moisture data was derived from the NASA earth data Giovanni repository. The Standardised Precipitation Index (SPI), Standardised Precipitation Evapotranspiration Index (SPEI), Standardised Streamflow Index (SSI), and Nonlinear Aggregated Drought Index (NADI) were selected to assess and characterise drought conditions in the LRC. SPI is precipitation based, SPEI is precipitation and evapotranspiration based, SSI is based on streamflow while NADI is a multivariate index based on rainfall, potential evapotranspiration, streamflow, and storage reservoir volume. All indices detected major historical drought events that have occurred and reported over the study area, although the precipitation based indices were the only indices that categorised the 1991/1992 drought as extreme at 6- and 12- month timescales while the streamflow index and multivariate NADI underestimated the event. The most recent 2014/16 drought was also categorised to be extreme by the standardised indices. The study found that the multivariate index underestimates most historical drought events in the catchment. The indices further showed that the most prevalent drought events in the LRC were mild drought. Extreme drought events were the least found at 6.42%, 1.08%, 1.56%, and 4.4% for SPI, SPEI, SSI, and NADI, respectively. Standardised indices and NADI showed negative trends and positive upward trends, respectively. The positive trend showed by NADI depicts a decreased drought severity over the study period. Drought events were characterised based on duration, intensity, severity, and frequency of drought events for each decade of the 30 years considered in this study i.e. between 1986 – 1996, 1996 – 2006, 2006 – 2016. This was done to get finer details of how drought characteristics behaved at a 10-year interval over the study period. An increased drought duration was observed between 1986 - 1996 while the shortest duration was observed between 1996 - 2006 followed by 2006 - 2016. NADI showed an overall lowest catchment duration at 1- month timescale compared to the standardised indices. The relationship between drought severity and duration revealed a strong linear relationship across all indices at all timescales (i.e. an R2 of between 0.6353 and 0.9714, 0.6353 and 0.973, 0.2725 and 0.976 at 1-, 6- and 12- month timescales, respectively). In assessing the overall utilisation of an index, the five decision criteria (robustness, tractability, transparency, sophistication, and extendibility) were assigned a raw score of between one and five. The sum of the weighted scores (i.e. raw scores multiplied by the relative importance factor) was the total for each index. SPEI ranked the highest with a total weight score of 129 followed by the SSI with a score of 125 and then the SPI with a score of 106 while NADI scored the lowest with a weight of 84. Since SPEI ranked the highest of all the four indices evaluated, it is regarded as an index that best describes drought conditions in the LRC and was therefore used in drought prediction. Statistical (GAM-Generalised Additive Models) and machine learning (LSTM-Long Short Term Memory) based techniques were used for drought prediction. The dependent variables were decomposed using Ensemble Empirical Mode Decomposition (EEMD). Model inputs were determined using the gradient boosting, and all variables showing some relative off importance were considered to influence the target values. Rain, temperature, non-linear trend, SPEI at lag1, and 2 were found to be important in predicting SPEI and the IMFs (Intrinsic Mode Functions) at 1, 6- and 12- month timescales. Seven models were applied based on the different learning techniques using the SPEI time series at all timescales. Prediction combinations of GAM performed better at 1- and 6- month timescales while at 12- month, an undecomposed GAM outperformed the decomposition and the combination of predictions with a correlation coefficient of 0.9591. The study also found that the correlation between target values, LSTM, and LSTM-fQRA was the same at 0.9997 at 1- month and 0.9996 at 6- and 12- month timescales. Further statistical evaluations showed that LSTM-fQRA was the better model compared to an undecomposed LSTM (i.e. RMSE of 0.0199 for LSTM-fQRA over 0.0241 for LSTM). The best performing GAM and LSTM based models were used to conduct uncertainty analysis, which was based on the prediction interval. The PICP and PINAW results indicated that LSTM-fQRA was the best model to predict SPEI timeseries at all timescales. The conclusions drawn from drought predictions conducted in this study are that machine learning neural networks are better suited to predict drought conditions in the LRC, while for improved model accuracy, time series decomposition and prediction combinations are also implementable. The applied hybrid machine learning models can be used for operational drought forecasting and further be incorporated into existing early warning systems for drought risk assessment and management in the LRC for better water resources management. Keywords: Decomposition, drought, drought indices, early warning system, frequency, machine learning, prediction intervals, severity, water resources. / NRF

Water quality deterioration is reaching crisis proportions in South Africa. Many South African catchments are over-allocated, and decreasing volumes of source water mean increasing concentrations of pollutants. The Crocodile River Catchment in the Mpumalanga province in South Africa was identified through previous research, as a catchment faced with deteriorating source water quality for water users in the catchment. Poor source water quality has become a sufficiently acute concern for the stakeholders in this catchment to co-operate in developing a process that assists with compliance control of their water use and waste disposal to reduce costs, decrease industrial risks as water quality compliance increases, and improve source water quality. The sugar industry is downstream within the Crocodile River Catchment, and is affected by the activities of all upstream water users; the industry is thus dependent on the stakeholders upstream participating in the effective management of the resource. However, the sugar industry is also located just before the confluence of the Crocodile River and Komati River upstream of the Mozambique border, and thus the water quality of the sugar industry effluent will affect the quality of the water that flows into Mozambique. The sugar industry is on the opposite river bank to the Kruger National Park, which has high water resource protection goals. Therefore, the sugar industry has a national role to play in the management of water resources in the Crocodile River Catchment. This study provides a focused view of the role of the sugar industry in the development of a co-operative, integrated water quality management process (IWQMP) in the Crocodile River Catchment. In order to address the objectives of this study, this research drew from an understanding of the social processes that influence water management practices within the sugar industry as well as social processes that influence the role of the Inkomati-Usuthu Catchment Management Agency as the main governing institution in water resource management in the Inkomati Water Management Area. The study also drew from an understanding of scientific knowledge in terms of a water chemistry which describes the upstream and downstream water quality impacts related to the sugar industry. The water quality analysis for the Lower Crocodile River Catchment shows a decline in water quality in terms of Total Dissolved Solids (TDS) loads when moving from below Mbombela to the Mozambique border. The major sources of TDS in the Lower Crocodile River are point source dominated, which may be attributed to the extensive mining, industrial and municipal activities that occur across the catchment. When observing Total Alkalinity (TAL) and pH values from below Mbombela to the furthest monitoring point, there is deterioration in the quality of the water in the Lower Crocodile River, with the Kaap River contributing a negative effect that is diluted by the Crocodile main stem. The Hectorspruit Waste Water Treatment Works (WWTWs) (located in the Lower Crocodile River Catchment) contributes high concentrations of TDS and TAL into the Crocodile River. Total Inorganic Nitrogen and Soluble Reactive Phosphorus concentrations decrease in the lower reaches of the Crocodile River compared with the river below Mbombela, which can be attributed to the extensive sugar cane plantations located in the Lower Crocodile River Catchment acting as an “agricultural wetland” that serves a function of bioremediation resulting in large scale absorption of nutrients. This is an interesting result as earlier assumptions were that fertiliser application would result in an overall increase in nutrient loads and concentrations. Biomonitoring data show no substantial change in aquatic health in the LowerCrocodile River Catchment. For a catchment that has an extensive agricultural land use in terms of sugarcane and citrus production, the Crocodile River is unexpectedly not in a toxic state in terms of aquatic health. This is a positive result and it suggests that pesticide use is strictly controlled in the sugar and citrus industry in the Crocodile River Catchment. For long term sustainability, it is essential for the sugar industry to maintain (and possibly improve) this pesticide management. The social component of this study aimed to provide an analysis of the management practices of the sugar mill as well as examining agricultural practices in the sugar cane fields in relation to water quality management through the use of Cultural Historical Activity System Theory (CHAT). This component showed that there are contradictions within the sugar industry activity system that are considered to be areas of “tension” that can be loosened or focused on to improve the contribution the sugar industry can make to the IWQMP. Surfacing contradictions within the sugar industry activity system and the Inkomati-Usuthu Catchment Management Agency activity systems highlighted areas of potential for learning and change. While an understanding of biophysical processes through scientific knowledge is critical in water management decision making, it is evident that an understanding of other actors, institutions and networks that inform water quality management decision-making also plays a significant role. The notion of improving the role of scientific or biophysical knowledge in contributing to socio-ecologically robust knowledge co-creation, decisions and actions towards resolving water quality problems is emphasised. Specifically, moving towards improving interactions between scientists and other actors (water users in the Crocodile Catchment in this case), so that scientific practices become more orientated towards societal platforms where water quality management is tackled to enable improved water quality management practices. Therefore, linking the social and biophysical components in this study provides a holistic understanding of how the sugar industry can contribute to the development of an IWQMP for the Crocodile River catchment.

MESHWR / Department of Hydrology and Water Resources / See the attached abstract below

MESHWR / Department of Hydrology and Water Resources / This study was aimed at investigating the long term seasonal and annual changes in rainfall duration and magnitude at Luvuvhu River Catchment (LRC). Rainfall in this catchment is highly variable and is characterised of extreme events which shift runoff process, affect the timing and magnitude of floods and drought, and alter groundwater recharge. This study was motivated by the year to year changes of rainfall which have some effects on the availability of water resources. Computed long term total seasonal, annual rainfall and total number of seasonal rainy days were used to identify trends for the period of 51 years (1965- 2015), using Mann Kendal (MK), linear regression (LR) and quantile regression methods. The MK, LR and quantile regression methods have indicated dominance of decreasing trends of the annual, seasonal rainfall and duration of seasonal rainfall although they were not statistically significant. However, statistical significant decreasing trends in duration of seasonal rainfall were identified by MK and LR at Matiwa, Palmaryville, Levubu, and Entabeni Bos stations only. Quantile regression identified the statistically significant decreasing trends on 0.2, 0.5 and 0.7 quantiles only in the Palmaryville, Levubu and Entabeni Bos, respectively. Stations with non-statistically significant decreasing trends of annual and seasonal rainfall had magnitude of change ranging from 0.12 to 12.31 and 0.54 to 6.72 mm, respectively. Stations with non-statistically increasing trends of annual and seasonal rainfall magnitude had positive magnitude of change ranging from 1.51 to 6.78 and 2.05 to 6.51 mm, respectively. The Study recommended further studies using other approaches to determine the duration of rainfall to improve, update and compare the results obtained in the current study. Continuous monitoring and installation of rain gauges are recommended on the lower reaches of the catchment for the findings to be of complete picture for the whole catchment and to also minimize the rainfall gaps in the stations. Water resources should be used in a sustainable way to avoid water crisis risk in the next generations. / NRF