The sub-Antarctic Zone (SAZ) is a zone of vigorous vertical mixing in the Southern Ocean where it is difficult to obtain data for model validation on the turbulence conditions. In this study, a onedimensional configuration from the Nucleus for the European Modelling of the Ocean (NEMO) model was implemented in order to determine the sensitivity and turbulence response of an idealized SAZ water column. Various turbulence scheme parameterizations that are available for ocean models were tested. Furthermore, the number of vertical levels were varied in order to ascertain the sensitivity of the grid. The forcing data were obtained from various reanalyses (ERA-Interim, NASA, NCEP and JRA55) and were likewise tested. Different turbulence diagnostics and univariate indicators were chosen to ascertain the turbulence response and to analyse the energetics of the water column. It was found that using different reanalyses produced different tracer (salinity and temperature) results. Even though the results varied considerably, very high correlations were found for the potential energy anomaly between reanalyses and insignificant correlations were found for the other indicators. This suggested that it was a valuable descriptor which captured the buoyancy fluxes and wind stress information and can be efficiently used to assess the vertical turbulent state with data such as ARGO profiles. It was further found that for a single reanalysis, the turbulence schemes had produced similar results (with small variability and not to the extent as changing the reanalysis) for the turbulence diagnostics and univariate indicators. An important finding of an entrapped warm water parcel beneath cooler waters was found in simulation outputs as well as ARGO validation data. For realistic conditions observed from the ARGO floats, as the season progressed, there were no more instances of a warm water parcel. There was no reason however, to why there should not have been eddies passing by the region. In simulations, the warm water parcel persisted throughout the season for simulated data, likely causing the early stratification that affects ocean models in the SAZ. The stratification was found to have an approximate one month early onset observed from comparing the ARGO data profiles to simulated profiles. The Brunt Väisälä frequency, potential energy anomaly as well as the buoyancy flux were analysed and these diagnostics indicated that an approximate one month early stratification was found during November. It was likely that this false stratification signal may have influenced the summer stratification leading to a poor representation of the Mixed Layer Depth (MLD) and various other indicators. It was found that during the austral winter months, the model simulated comparable MLD's to the ARGO float data as well as theWinter Cruise data (obtained from the SA Agulhas II), capturing the winter dynamics well.

Droughts occurred over South Africa during the summer seasons of 2014/2015 and 2015/2016. At the same time, the Pacific Ocean was warmer than normal starting in 2014 and leading to the very strong 2015/2016 El Niño. The first objective of this study is to document the ocean and climate conditions that occurred during the summer seasons 2014/2015 and 2015/2016 in southern Africa. NCEP Reanalysis data is used to compute the monthly and seasonal scale composite mean and anomalies of large-scale circulations during the summer seasons of 2014/2015 and 2015/2016. Results obtained from the study suggest that some months of 2014/2015 and 2015/2016 were canonical to the effect of El Niño over southern Africa, but not all of them during the summer seasons were dry. The wetter than normal conditions in northeast South Africa, Botswana and Zimbabwe during December 2014 are unfamiliar for a canonical summer El Niño event over southern Africa. Anomalous cooler than normal Sea Surface Temperature (SST) occurred over the west coast and south coast during December 2014 and February 2015, while it is usually warmer during El Niño. Additionally, the colder than normal SST at the south coast during February 2016 and Namibian and West Coast during March 2016 does not mimic the canonical El Niño patterns. However, this did not influence the El Niño-like warmer seasonal SST average during 2015/2016. The lower than normal pressure anomalies over the subcontinent during December 2014 and January 2015 were not portraying a canonical El Niño pattern but the other months were. The seasonal larger than normal pressure at 500 hPa over the subcontinent was more typical of El Niño during summer 2015/2016 and acted to suppress rainfall. Secondly, the study uses the Standard Precipitation Index (SPI) at different time scales (3 months duration, 5 months duration and 17 months duration) to assess the severity of 2015/2016 summer drought compared to the other droughts of the 20th and 21st century (1921 to 2016) and to analyse the relationship between droughts and ENSO. The South African Weather Service (SAWS) rainfall data shows that KwaZulu-Natal was the only region within South Africa, to have the 2015/2016 as the strongest summer drought since 1921 but 2015/2016 was still one of the worst droughts on record in South Africa, especially at the 2 consecutive seasonal scales. In general, droughts are favored by El Niño and wetter conditions by La Niña, but the second strongest El Niño of 1997/1998 led to near normal rainfall over the north-eastern region at all time-scales. The SPI has proven to be very versatile, flexible and very effective to monitor the 2015/2016 summer drought in the complex South African rainfall regime. However, there was little difference between 3 months SPI at the end of February and 5 months SPI at the end of March. For South Africa, the summer rainfall 2015/2016 season had the fifth worst drought after El Niño related drought of 1982/1983 and 1991/1992 and the non-El Niño related droughts of 1967/1968 and 1944/1945. At the 17-month scale, an index that encompasses two summer seasons 2015/2016 was the third worst drought since summer 1921/1922 due to dry conditions in 2014/2015 and 2015/2016.