Maviza, Auther2022-07-192022-07-192021https://hdl.handle.net/10539/33033A thesis submitted to the School of Geography, Archaeology and Environmental Studies, Faculty of Science, University of the Witwatersrand, Johannesburg, in fulfilment of the requirements for the degree of Doctor of Philosophy, 2021Climate change is one the most topical subjects in today’s world. Numerous studies have linked climate to increased incidence and severity of impacts of natural hazards such as droughts, floods and wildfires. To this end, climate science research has been and still continues to be one of the most active areas of scientific enquiry in a quest to better understand climate systems dynamics and more recently their interlink with other systems. In this study, the climate and land use –landcover change dynamics are explored and then their impacts on surface hydrological conditions in the upper Mzingwane subcatchment (UMS) of Zimbabwe assessed. Initially, an in-depth review of existing climate and hydrology published research in Zimbabwe over the past 29 years is undertaken using a systematic review approach. It emerged that of the 107 studies reviewed, the two predominant themes covered were climate impact (39%) and climate vulnerability, adaptation and mitigation (39%) while climate and hydrological modelling were the least covered themes at4%. Most of the research is outdated in Zimbabwe and has limited use of more recent climate and hydrological modelling tools and techniques. With regards to landscape degradation, historical land use and landcover changes and modelled future land use and landcover scenarios in UMS were explored the using Geographic Information Systems and Remote Sensing. It emerged that extensive deforestation has been taking place in the UMS with losses of over 700km2 between 1089 and 2018 and this trend is projected to continue into the future with over 40% of forest cover lost by 2038. These changes are most likely to be driven by increased human activities in the area especially small-scale and illegal gold mining. To better understand historical precipitations conditions in the UMS, climate station historical precipitation records are used to assess twentieth century climate extreme events over UMS. Though results indicate statistically insignificant trends, indication is that high intensity and short period precipitation events have been increasing despite the overall decrease in total precipitation levels over the UMS. Generally, mean precipitation anomalies show a general negative trend of between -0.06mm and -6.36mm in the northern and western region of the UMS suggesting general drying between 1920 and 2001. For example, results show declining trends in 5-day maximum precipitation (RX5Day), (mean =-0.879mm/annum). Overall, the smoothed 5-year moving average trends for most index anomalies seem to reveal a near 20 to 30 year periodicity over the UMS. Future climate projections in the UMS for the near future (2021 –2040), mid-term future (2021-2060) and long term future (2061-2099) are explored using the Conformal Cubic Atmospheric Model (CCAM) data downscalings of 6 GCMs. Models show an overall anomaly signal ranging from -15mm to +18mm change in annual total precipitation over the UMS. Mid to long-term precipitation anomalies range between +3mm and +9mm compared to +3mm and +15mm in the near-future period suggesting decreases magnitude of change in total precipitation in the future. Four of the ensemble members generally show positive spatial patterns of change while two show the opposite trends in long term average of monthly precipitation though the is consensus on a south to north increasing gradient in precipitation in all future periods. The Max Plank Institute (MPI), the Geophysical Fluid Dynamics Laboratory Climate Model (GFDL CM2.5) and the National Centre for Meteorological Research Climate Model version 5 (CNRM-CM5) have highest competence in simulating monthly average precipitation while the Community Climate System Model Version 4 (CCSM4) has lowest performance. With regards to temperature projections, all ensemble members show a clear consensus on an increasing trend in both maximum and minimum temperature with magnitudes of changes varying between 1.1°Cup to 6.4°C from the near to the long-term future. These changes translate to between +0.7°C to +1.15°C decade-1 in mean temperature changes which shows consistent gradual warming over the UMS. This could have significant impacts on both human health, agriculture and water security. For the first time in the Northern Limpopo basin, a physical, semi-distributed hydrological model is successfully applied to simulate stream run-off with satisfactory levels of accuracy using downscaled CCAM data. Majority of the ensemble members simulate simulated peak stream discharges ranging between 1244.8 m3s-1 and 42.9 m3s-1 though most (4) of the downscalings simulate a mean increase in peak discharge of ~620m3s-1 i.e. ~7.83 m3s-1yr-1 over the entire future period. The NorESMI and the GFDL-CM3 model project declining trends in peak stream discharge though the former simulates the highest peak discharge (1244.8m3s-1) in the baseline period (1996 –2015). The increases could be related to projected increases in precipitation by the CCAM ensemble members as earlier presented. The impact of land use and land cover changes on the stream peak and total volume discharge seems to be inconclusive in that in the 2018 (2038) LULC change forcing of the hydrological model show a decrease (increase) in stream peak discharge and volume in the UMS. The increase in discharge could relate to the projected increase in bareland/ impervious surfaces related to human development/ activities in the UMS. Overall, the study manages to add new knowledge to fill gaps in historical and project future climate change/ variability scenarios in the UMS. The extent of past and future landscape degradation is quantified indicating extensive deforestation to continue into the future. With projected future climate projections of increasing total precipitation and temperatures, the projected impact is increased stream run-off and peak discharge. Novelty in this study is the first time use of downscaled, and/ bias-corrected high-resolution CCAM data together with simulated LULC scenarios to model future multi-temporal surface stream run-off at a subcatchment level. These study findings are consequential in understanding the climate-LULC-hydrology nexus and therefore valuable in planning for climate change impact mitigation, strategy and policy development so as to avert negative water security and livelihood impacts in the UMS and beyondenThesis