Born and bred in Nigeria, Oluwaseun Kunle Oyebode earned a Bachelor of Science degree in Agricultural and Environmental Engineering from the University of Ibadan, Nigeria in 2009. He graduated with a second class honours, upper division and he was one of the top three students in his graduating class. Oyebode has through specialized courses such as hydrology, water resources conservation and management as well as computer modelling of agricultural, water and environmental systems, created a platform for himself to acquire basic skills of systematic research, analytical thinking and information evaluation. These skills have in turn provided him with the drive and insight to proffering solutions to complex hydrological modelling problems.
“These skills became evident in my final year research project entitled, ‘Simulating effects of drainage design parameters on optimum crop yield using DRAINMOD,’ which received the approval of the American Society of Agriculture and Biological Engineers (ASABE),” said Oyebode. A paper on this study was presented at the ASABE TMDL 2010 Conference: Watershed Management to Improve Water Quality, held in Baltimore, Maryland, USA. The paper was also published in the conference proceedings. In 2009, he was one of the few beneficiaries of the Oyo State Scholarship Awards of Excellence granted to promising Nigerian students by the Ministry of Education, Oyo State Government in Nigeria.
In 2012, Oyebode undertook his Master of Technology degree programme in Civil Engineering at the Durban University of Technology. His speciality was in water resources engineering. He completed the programme at DUT Cum Laude and his research findings were published in internationally accredited journals and conference proceedings. He is also a member of the International Water Association (IWA), Water Institute of Southern Africa (WISA), Nigerian Society of Engineers (NSE), International Association of Engineers (IAENG) and the International Association of Hydrological Sciences (IAHS).
His current dissertation is on ‘Modelling Streamflow Response to Hydro-climatic Variables in the upper Mkomazi River, South Africa.’ Speaking about his research and his findings, Oyebode said: “The availability of data has always been a major factor that influences the accuracy and reliability of predictions in hydrological studies. Furthermore, to achieve accurate and reliable predictions, large datasets are often required for the purpose of training and validation in all categories of models. However, these huge amounts of datasets are limited to certain regions and often unavailable in some areas, especially in developing countries such as South Africa. Therefore, limitations of datasets are considered as a major obstacle to the implementation of hydrological modelling studies in data sparse regions.”
The aim of Oyebode’s research was to determine a suitable technique for modelling the impacts of hydro-climatic variables on streamflow dynamics in the Upper Mkomazi River under limited availability of datasets. In order to achieve this aim, the predictive ability of two evolutionary computation (EC) data-driven modelling techniques namely genetic programming (GP), and differential evolution (DE)-supervised artificial neural network (DE-ANN) was investigated comparatively. The study is unique in that most studies from literature have employed only one EC technique for streamflow modelling without investigating their performance against other EC techniques. Therefore, comparative studies of this nature (ie the application of GP and a DE-ANN) were conspicuous by their near absence. This served as an inciting cause for employing two EC techniques. His study to the best of his knowledge is the first application of GP to hydrological analysis in South Africa. Results from the study found the GP-derived models better than the DE-ANN models in terms of prediction accuracy, computation speed and adaptation to occurrences of rare events. Furthermore, an evolutionary approach which involved the conjunctive use of GP and DE was introduced to improve the performance of the ANN models. GP was used as a screening tool, to reduce the dimensionality of the ANN models while DE provided more invigorated search for better solutions through sensitivity analysis. These facilitated quicker convergence and also reduced over-parameterization effects in the ANN models. Generally, results obtained from the individual monthly models demonstrate that the GP approach is capable of providing an accurate and reliable estimate of streamflow in the upper Mkomazi River.
The reason he chose to conduct such research was because of his love for innovation and creation, and due to higher demand for water in Pietermaritzburg and Durban cities. This study would help to understand the streamflow dynamics of the river, especially in this era of climate variability. He was supervised by Dr Josiah Adeyemo, a Senior Lecturer at the Department of Civil Engineering and Surveying at the University.
One of the challenges he faced during the course of his MTech studies was harnessing plausible information on the river using the knowledge of evolutionary computing; as his research was inter-disciplinary in nature. He overcame this hurdle by forming synergistic relationships with information technology experts, especially in the area of computer programming. This facilitated the development and interpretability of the working principles of the models. “Mentoring from renowned researchers in this field was also a bottle neck. All in all, challenges are part of life and when you cross to the other side of it, they become experiences for the future,” he said.
Oyebode hopes to be employed in an organisation that will give him room for capacity building and development of skills that will be instrumental to fulfilling his career goals. He also plans to do his PhD as soon as possible. “My long term career goal is to become a renowned researcher and consultant, providing solutions to real-world science and engineering problems, especially as it relates to the field of hydrology, water resources and environmental management.”
Oyebode will be graduating on 11 April 2014 at 2pm during DUT’s graduation ceremony for the Faculty of Engineering and the Built Environment.
Noxolo Memela
