MENVSC / Department of Geography and Geo-Information Sciences / Out of all rain-producing weather systems, cut-off lows (COLs) are linked with the occurrence of high impact rainfall and in some cases short-lived floods which can last for 24 hours over South Africa. This study examined the characteristics associated with the present occurrence of the severe COL systems over South Africa from 2011 to 2017. The accuracy of the 4.4 km Unified Model (UM) which is currently in use for simulating areas of deep moist convection in South Africa was evaluated. The UM simulated geopotential height at 500 hPa as well as the associated 24 hours precipitation which were compared against the daily fields of geopotential height and 6-hourly precipitation from the European Centre for Medium-Range Weather Forecasts (ECMWF). COL events were categorized and analyzed according to the associated surface circulation patterns at 850 hPa. The seasonal distribution and duration of the systems over northern (10°E-33°E //22°-32°S) and southern (10°E-33°E //32°-35°S) regions of the study area were also analyzed. Results show COL systems shifting with season towards the north eastern parts of the country, with an increased number of events during the austral winter season during the study period. Systems which lasted for long time were observed during the austral winter and spring seasons. The UM tends to simulate areas of heavy precipitation accurately with poor simulation during the initial stages of the systems. The UM provided a more realistic-looking closed geopotential height and rainfall fields for systems which are coupled with a cold front at the surface. Application of the knowledge about the evolution in the characteristics of COL events from this study can improve the operational forecasting of these weather systems over the country. / NRF

Thesis (MSc)--Stellenbosch University, 2012. / Severe floods in the Western Cape have caused significant damage to hydraulic structures, roads and other infrastructure over the past decade. The current design criteria for these structures and flood return level calculations are based on the concept of stationarity, which assumes that natural systems vary within an envelope of variability that does not change with time. In the context of regional climate change and projected changes in rainfall intensity, the basis for these calculations may become unrealistic with the passage of time. Hydraulic structures and other infrastructure may become more vulnerable to damaging floods because of changing hydroclimatic conditions. This project assesses the changes in extreme rainfall values over time across the Western Cape, South Africa. Using a Generalised Pareto Distribution, this study examines the changes in return levels across the Western Cape region for the periods 1900-1954 and 1955-2010. Of the 137 rainfall stations used in this research, 85 (62%) showed an increase in 50-year return level, 30 (22%) a decrease in 50-year return level and 22 (16%) stations displayed little change in rainfall intensity over time. While there were no clear spatial patterns to the results, they clearly indicate an increase in frequency of intense rainfalls in the latter half of the 20th and early 21st century. The changes in return level are also accompanied by a change in the frequency of high intensity 2-3 day long storms. 115 (84%) of the 137 rainfall stations showed an increase in the frequency of long duration, high intensity storms over the data record. This change generates a shifting risk profile of extreme rainfalls, which, in turn, creates challenges for the design of hydraulic structures and any infrastructure exposed to the resulting damaging floods. It can therefore be argued that it is inappropriate to design structures or manage water resources assuming stationarity of climate and that these principles should be assessed in order to reduce the risk of flood damage owing to increasing storm intensity. KEY WORDS Flood Risk, Stationarity, Disaster Risk, Hazard, Extreme Rainfall, Generalized Pareto Distribution, Climate

Thesis (MSc)--Stellenbosch University, 2015. / ENGLISH ABSTRACT: Eucalyptus camaldulensis Dehnh., or River Red Gum, is a commercially valuable yet recognised invasive alien plant (IAP) of riparian zones throughout South Africa. The invasive potential of E. camaldulensis is widely recognised, with specific regulations aimed at the management of E. camaldulensis. E. camaldulensis is known to use large amounts of water, reduce biodiversity, change river morphology and impact hydrological regimes of rivers. In the native range throughout Australia, E. camaldulensis displays a distinct relationship between rainfall, and flood events, for seed dispersal, germination and establishment, and consequently spatial extent, yet little is known about the relationships in the South African context. The aim of this project was to assess the impact of high rainfall and flood events on the establishment and distribution of E. camaldulensis along the Middle Breede River, between Worcester and Swellendam in the Western Cape, by establishing the current spatial extent of E. camaldulensis along the river, identifying flood events since 1950 and evaluating the impact rainfall and flood events had on the spatial extent thereof. Aerial imagery, rainfall, discharge and river level data was obtained dating back to 1980, as well as field data comprising of GPS-bounding of E. camaldulensis stands. Additionally, density measurements were obtained and interviews conducted with land users. Spatial analysis of aerial imagery, coupled with perimeter (GPS) data and density data were used to conduct spatio-temporal analysis, employing GIS and conventional statistical approaches to address the various objectives. Results indicated E. camaldulensis stands had a small overall increase in spatial extent since 1980. Flooding and rainfall events coincided with an increase in occurrence of E. camaldulensis with elevated river levels and frequent flooding, while spatial variation of this relationship was observed. The hydrological regime of the Breede River coincides with a slow increase in spatial extent of E. camaldulensis stands, but no affirmation of a positive real-world relationship was possible using the available data. Results further suggested, based on the current age class composition, that existing stands originated roughly during 1980, possibly due to commercial forestry related seeding into the river. Reduced fragmentation between stakeholders, educational programmes and improved reporting systems were recommended for improved IAP management within the area. / AFRIKAANSE OPSOMMING: Eucalyptus camaldulensis, of Rooibloekom (RB), is ‘n waardevolle kommersiële, maar erkende indringer plantspesie (IP) wat veral oewersones in Suid-Afrika indring. Die indringerpotensiaal van E. camaldulensis is welbekend, en spesifieke regulasies, gemik op die bestuur van RB en ander spesies is reeds aangeneem. E. camaldulensis is veral bekend vir sy hoë watergebruik, sy vermindering van biodiversiteit, sy vermoë om riviervorme te verander en sy algehele impak op die hidrologiese patroon van riviere waarmee dit in aanraking kom. In sy oorspronklike verspreidingsgebied in Australië toon E. camaldulensis ‘n bepaalde verhouding tussen reënval en vloedgebeurtenisse vir saadverspreiding, ontkieming en vestiging en derhalwe die ruimtelike verspreiding van die spesie; alhoewel hierdie verhouding in die Suid-Afrikaanse konteks steeds redelik onverduidelik bly. Die doelwit van hierdie studie was dus om die impak van hoë reënval en vloedgebeurtenisse op die ruimtelike verspreiding en vestiging van E. camaldulensis teenaan die Middel Breëde Rivier, spesifiek tussen Worcester en Swellendam, te evalueer. Hierdie doelwit was bereik deur die historiese ruimtelike verspreiding teenaan die rivier te meet, hoë reënval en vloedgebeurtenisse vanaf 1980 te identifiseer, en die huidige verspreiding en omtrek met GPS te meet. Digtheidafmetings, sowel as onderhoude met belanghebbendes teenaan die rivier was ook opgeneem. Visuele interpretatasie van lugfotos, sowel as omtrek (GPS) en digtheid-data was gebruik om ruimtelike analise uit te voer, deur die gebruik van GIS en konvensionele statistiese metodes, ten einde die doelwitte te evalueer. Resultate dui aan dat E. camaldulensis areas ‘n klein algemene groei getoon het sedert 1980. Hoë-reënval en gereëlde vloedgebeurtenisse het ook gepaard gegaan met ‘n groei van E. camaldulensis oppervlak, alhoewel hierdie verhouding ruimtelike variasie getoon het, met ‘n algemene groei patroon gemerk oor die volledige studietydperk. Ook geen stimulerende verhouding kon vanuit die beskikbare data bevestig word nie. Addisionele resultate het aangedui dat die verspreiding van E. camaldulensis ongeveer 1980 onstaan het, moontlik as gevolg van kommersiële bosbou-aanplanting en verwante saadverspreiding in die rivier vanaf daardie tyd. Aanbevelings ten opsigte van verbeterde indringerbestuur sluit in die beperking van huidige fragmentasie tussen belanghebbendes en betrokke verwyderingsorganisasies, addisionele onderrigprogramme sowel as die verbetering van terugvoersisteme.

The overall objective of the whole study was to assess whether conservation agriculture (CA) systems can work in the Eastern Cape Province (EC). The CA systems were engaged through cover cropping to address land degradation problems by emphasizing high biomass production in order to realize short term benefits such as moisture conservation, weed suppression and soil fertility benefits under rainfed conditions in the central region of the Eastern Cape province. Since rainfall is the most limiting factor to crop production in the EC, a within season rainfall distribution analysis was conducted to expose the quality of the season (onset, end and duration) and hence the feasibility of CA systems to guide agronomic decisions by farmers in EC. To assess season parameters, thirty four years of daily rainfall was collected from the University of Fort Hare Research station and used to conduct the rainy pentad (5 day rainfall totals) analysis and the daily rainfall analysis using INSTAT software programme. Based on the pentad analysis, results showed that Alice does not have a rainy season in 1 out of 2 years (50% probability) but has one in 1 out of 4 years (25% probability level). This criterion proved to be harsher and conservative when compared to the daily rainfall approach which is more precise in measuring trends on season parameters. The daily rainfall analysis indicated a 65% feasibility for the dry land cropping systems in the EC. The pentad analysis however was effective in illustrating seasonality and it showed that the wet season begins on the 1st of November, ending on the 22nd of March lasting for 140 days. Though the season duration appeared too long, the existence of dry spells during critical growth stages adversely affects the quality of the season. The daily rainfall analysis also managed to derive a signal which can guide planting decisions. For planting to be successful, this analysis determined that 20 mm of rain should be received in two consecutive days after the 1st of November. A screening trial for cover crop biomass production and weed suppression was conducted on-station Fort Hare Research Farm (32°46' S and 26° 50' E), and Msobombvu village (MSBV) (32°44' S, and 26° 55' E) over two seasons (2007/08 and 2008/09). Six summer cover crops i.e. cowpea (Vigna unguiculata), dolichos lablab (Dolichos argenteus), sunnhemp (Crotalaria juncea), buckwheat (Fagopyrum sagittatum), forage sorghum (Sorghum bicolor) and sunflower (Helianthus annus) were evaluated for biomass yield, and weed suppression. Decomposition rates, moisture conservation and residual effects of these cover crops on the succeeding main crop were also evaluated under dryland conditions. The screening trial was laid in randomized complete block design replicated three times. Forage sorghum (Sorghum bicolor) and sunflower (Helianthus annus) were identified as high biomass producers and their dry matter yields ranged from 8 -12 t ha-1. These cover crops can be useful in generating high biomass in rainfed cropping systems in the EC. Other cover crops produced 3 - 4 t ha-1 of biomass which fell short of the 6 t ha-1 expected benchmark. However, these biomass yields were important in weed management since all cover crop species showed a similar degree of weed suppression which surpassed the weed fallow treatment. As dead mulches, the cover crops failed to show residual moisture conservation and weed control benefits for the succeeding maize crop mainly because of poor residue persistence, and low harvestable fallow rainfall. Buckwheat (Fagopyrum esculentum), was selected for further investigations in a follow up trial on station in 2008/09 season because of its weed smothering qualities, suitability to short cycle rotations, and possible allelopathic properties. The trial aimed at finding weed and cost effective management options of buckwheat that are none detrimental to the succeeding maize crop. Results showed that cropping systems where buckwheat is followed by a main crop may not work as they are unprofitable with respect to R100 rand invested. Though perceived to have allelopathic properties, buckwheat failed to demonstrate the possibilities of allelopathic action against weeds. Intercropping trial was conducted on-station in 2007/8-2008/09 seasons to try and find better ways of fitting legume cover crops into maize based cropping systems without compromising production of staple cereals on limited landholdings. The trials evaluated three factors in factorial combination, cover crop planting date, intercropping strategy, and cover crop species. The trial was laid as 2 x 2 x 3 factorial arranged in a split-split plot design. The main plot factor was cover crop planting date, cover crops simultaneously planted with maize and cover crop planted two weeks after planting maize (DKC 61-25). The sub-plot factor was intercropping strategy, strip intercropping and betweenrow intercropping. The sub-sub-plot factor was cover crop species, Dolichos lablab (Dolichos argenteus (Highworth), and Cowpea Vigna ungiculata (Agrinawa) plus control plots of sole maize. Results showed that same time planting of leguminous cover crops with maize using the in-between row intercropping patterns can derive appreciable system biomass (maize/cover crop) yields, utilize land efficiently whilst getting favourable maize grain yield. Based on the rainfall analysis, results showed that the probability of success when relay seeding cover crops after two weeks into standing maize is low (15% chances of success). This suggests that relay intercropping strategies would not work due to the unavailability of a good quality season.

South Africa is a semi arid country in which the average rainfall of 450mm/year is well below the world average of about 860mm/year. As a result, South Africa’s water resources are scarce in global terms and limited in extent. Current predictions are that demand will outstrip water availability in the next 15 years. A coordinated approach to improve both water quality and quantity is needed and in order to achieve that, it is crucial to strengthen capacities of local community involvement in identifying the problems that affect them and strategies to solve them. This research was undertaken to develop a deeper understanding of community learning processes in integrated water resources management (IWRM) practices. The study drew on situated and social learning theory which explains that knowledge and skills are learned and embedded in the contexts in which knowledge is obtained and applied in everyday situations. Multiple data collection techniques were used within a case study design and included document analysis, interviews, focus group discussions and field observations. Data analysis was done in three phases and involved uncovering patterns and trends in the data sets. In this context I discovered, through careful observation and interviews with members of the different communities of practice, that people are learning through social learning interactions with other community members as they engage in their daily water management and food production practices. Learning interactions take place through both informal and formal processes such as meetings, training workshops, conversations and interactions with outsiders. I also discovered that people learn from ‘external groups’ or training programmes which bring new knowledge and expertise, but this needs to be contextualised in the local communities of practice. The research has also shown that there are a number of challenges that appear to exist in these learning contexts. For instance it was found that participation and social learning processes and interactions are influenced by a range of causal mechanisms that are contextual. These insights into how communities learn, as well as the tensions and difficulties that are experienced in the learning processes are important for furthering learning and participation in community-based IWRM practices, projects and programmes.

MESHWR / See the attached abstract below

MESHWR / Department of Hydrology and Water Resources / This study dealt with the influence of climate variability on flood and drought cycles and implications on rainy season characteristics in Luvuvhu River Catchment (LRC) in Limpopo of South Africa. Extreme weather events resulting in hazards such as floods and droughts are becoming more frequent due to climate change. Extreme events affect rainy season characteristics and hence have an influence on water availability and agricultural production. Annual temperature was obtained from Water Research Commission for stations 0723485W, 0766628W and 0766898W from 1950-2013 were used to show/or confirm if there is climate variability in LRC. Daily rainfall data was obtained from SAWS for stations 0766596 9, 0766563 1, 0723485 6 and 0766715 5 were used to detect climate variability and determine the onset, duration and cessation of the rainy season. Streamflow data obtained from the Department of Water and Sanitation for stations A9H004, A9H012, and A9H001 for at least a period of 30 years for each station were used for climate variability detection and determination of flood and drought cycles. Influence of climate variability on floods and droughts and rainy season characteristic were determined in the area of study. Trends were evaluated for temperature, rainfall and streamflow data in the area of study using Mann Kendall (MK) and linear regression (LR) methods. MK and LR detected positive trends for temperature (maximum and minimum) and streamflow stations. MK and LR results of rainfall stations showed increasing trends for stations 0766596 9, and 0766563 1 whereas stations 0723485 6 and 0766715 5 showed decreasing trends. Standardized precipitation index (SPI) was used to determine floods and droughts cycles. SPI results have been classified either as moderately, severely and extremely dry or, moderately, very and extremely wet. This SPI analysis provides more details of dominance of distinctive dry or wet conditions for a rainy season at a particular station. Mean onset of rainfall varied from day 255 to 297, with 0766715 5 showing the earliest onset compared to the rest of the stations. Cessation of rainfall for most of the hydrological years was higher than the mean days of 88, 83 and 86 days in 0766596 9, 0766563 1 and 0723485 6 stations. Mean duration of rainfall varied from 102 to 128, with station 0766715 5 showing shortest duration of rainfall. The results of the study showed that the mean onset, duration and cessation were comparable for all stations except 0766715 5 which had lower values. The study also found that climate variability greatly affects onset, duration and cessation of rainfall during dry years. This led to late onset, early cessation and relatively short duration of the rainfall season. Communities within the catchment must be educated to practice activities such as conservation of indigenous plants, reduce carbon dioxide emissions.

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