Flooding in KwaZulu-Natal. File photo Rajesh Jantilala/AFP/Getty
The University of KwaZulu-Natal is developing a groundbreaking early-warning system for extreme weather after natural disasters in the province, such as the deadly April 2022 floods, caught the attention of international researchers and funders.
The project is a multi-partner collaboration with the University of the West of Scotland, the Royal College of Surgeons in Ireland’s faculty of nursing and midwifery, and the University of Portsmouth, said acting deputy vice-chancellor of research and innovation at UKZN, Anil Chuturgoon.
He said the UK’s National Institute for Health and Care Research had awarded the project a grant of £2.2 million to develop the system, which represents “a transformative step” in safeguarding lives and strengthening communities against the growing threats of “climate-induced disasters”.
“Together, we are transforming innovative ideas into actionable solutions. This initiative is more than just a response to climate change. It is a commitment to a safer, healthier and more resilient future,” Chuturgoon said.
“The weather project is a beacon of hope, a demonstration of how science and technology, when combined with community engagement, can mitigate the impact of a changing climate.”
University of the West of Scotland and Royal College of Surgeons in Ireland representative Mary Lynch said that when she became aware of the devastating effects of the April 2022 floods — in which more than 400 people died and the economy lost an estimated R17 billion in damages — she contacted Professor Saloshni Naidoo, the head of public health medicine at the UKZN, to advise her about the grant call for projects.
Lynch said the research partners collaborated and applied for the grant, which was approved, to support the development and implementation of the project.
The University of Portsmouth’s David Ndzi said researchers would work closely with communities and the South African Weather Service to develop an early-warning system that would be able to precisely predict rainfall and flooding.
“The other key aspect is to predict disease outbreaks. So, as we’ve heard, climate change is no longer a prediction. It is something that, within the area you are experiencing unprecedented rainfall within short spaces of time, there has been an unprecedented loss of life,” Ndzi said.
The system, based on a high-tech system already in use in Malaysia, will include a phone app that residents can download to access weather predictions for their specific location. It will use machine learning to adapt current rainfall patterns to make predictions, rather than relying on 10- to 15-year-old data, enhancing its accuracy.
“Part of the problem that we have with the conventional systems of measuring rain with satellite and radar is a lack of high resolution to tell you what is going to happen on your street,” Ndzi said.
“This system is going to work on devices, using devices that will cost less than $5, just measuring temperature, humidity and atmospheric pressure, to predict rain. And hence you can scatter them around, and it gives you its prediction at each location and, based on that, you can then know how much water is going to be getting to a specific river or residential area.
“The aim of the project is to have a system that helps the population to know well in advance when it is going to rain and whether that rain is going to cause flooding. And in this world, it’s not about us as the researchers, it’s a co-creation with the communities that are affected.”
He said the researchers were taking a different approach to traditional rain predictions that focus on analysing past data, because the climate is changing. They would use artificial intelligence (AI) to predict when it would rain, and how heavily, as well as the danger of flooding.
“It is not very valuable to solve a problem after it has occurred. So, the predictive system is to give that advance warning for people to really take precautionary measures and minimise harm to property or the loss of life,” Ndzi said.
The eThekwini municipality will collaborate closely with researchers to ensure a system is developed to assist the most vulnerable communities across KwaZulu-Natal, said Geoff Tooley, a former engineer at the municipality’s coastal stormwater and catchment management department.
He said the municipality had previously collaborated with the University of KwaZulu-Natal to develop a successful community-based early-warning system with the Quarry Road informal settlement.
“During the 2022 floods, we saw 400 dwellings out of that community washed away. But, as a result of the collaboration between the city’s forecast early-warning system and the community-based early-warning system, we didn’t lose a single life to drowning,” Tooley said.
“The challenge that we faced was how [to] expand this to the other 187-odd informal settlements that are vulnerable, and within floodplains, in our city. When we saw this project, and the planned use of AI, and how we can use that to predict, inform and disseminate information to our vulnerable communities, we saw this as a massive opportunity.”