Dr Leanne Seeliger is a researcher at the Unit for Environmental Ethics and the Water Institute at Stellenbosch University.
Image: Supplied.
By Dr Leanne Seeliger
World Water Day is observed annually on 22 March to raise awareness about the importance of water in our lives and the sustainable management of freshwater resources. In celebration of this year’s event, I would like to reflect on the following question: Can digitilisation and Artificial Intelligence (AI) improve water management in Cape Town and other major cities around the world?
Last year, researchers at Stellenbosch University completed a project in the Philippi Horticultural Area funded by the Water Research Commission, WWF and the Western Cape Department of Agriculture. Situated on the Cape Flats Aquifer and surrounded by racing stables, factories and informal as well as formal housing settlements, this agricultural area is known as the breadbasket of Cape Town because it supplies many residents of surrounding areas with vegetables.
The Cape Flats Aquifer is part of the City of Cape Town’s additional water supply that is earmarked to replenish the City’s water resources during drought. In recent years, there has been a noticeable deterioration of the water quality in the Aquifer with farmers, industry and formal and informal residents contributing to it.
As part of the project, the agricultural canals that ran through farms, informal settlements and formal neighbourhoods were cleaned up using GIS (Geographic Information System) data input by communities to monitor water use and pollution in the area. The rationale behind digitalising the canals and using community data for improved water governance was to provide the City with accurate information on where the canals were located and what pollution points were problematic. In theory, the citizen science data, if captured by local government timeously, could serve as an early warning system to prevent pollution from damaging the water supply.
If followed up with more scientific analysis, this could lead to better prosecution of polluters and identifying of risks in water contamination. While the City of Cape Town already has a digital portal that citizens use, the speed at which these complaints are followed up and reported to the correct department was often not as fast and efficient as the project would have liked. Unfortunately, the project ended before further advocacy or research on this issue of the integration of citizen science data could be done.
Participants acknowledged that while the digitilisation of water data gathered through citizen science was desirable as an early warning system for local government, it was not easy to achieve. Citizen science data was not always accurate or clean. It also needed to be interpreted, and the results sent to the right department for follow up. This required a civil service with the capacity to receive the information in the correct format, process it fast enough to work proactively on its resolution. Artificial intelligence systems that can read a wide variety of data from video material, to photographs and words could speed up the identification of red flags in certain geographical zones, it was thought.
It was, however, emphasised that AI could not replace the need for officials to spend face-to- facetime with communities, nor could it replace jobs. Rather, its value lay in speeding up the processing of data and the sending of it to the right departments. Scientific water quality and water quantity information also presented problems because it was expensive to obtain and it was also often privately owned and not in the public domain. The protection of private data by legislation such as the POPI Act posed another problem, as farmers and landholders were often reluctant to release information on their use of water.
Since AI required a lot of electricity to power up the servers that hosted the data, the environmental costs of using it were also examined. In a country like South Africa where load-shedding is more frequent, this might pose a problem. The impacts of more evidence-based water management in the public sector, however, were expected to be transformative. Civil servants and politicians would be able to prioritise water management concerns in the City with the availability of more real time data.
Stakeholder engagement could improve in places like the Philippi Horticultural Area where densification and the informality of residence structures sometimes made it difficult to locate water issues timeously. However, if digitalisation was to improve the public’s engagement in water monitoring, then citizens would need a reliable source of free public internet. They would also have to be able to follow instructions on webpages and public portals to report their water concerns. This would require an education awareness campaign through schools, eco-clubs and other non-governmental organisations working in the water sector. In this regard, Stellenbosch University Water Institute and the Department of Water and Sanitation have been working to raise citizen awareness about water issues.
During this year’s National Water Week (20 – 26 March), they will be testing a water app on local rivers in Stellenbosch. Schoolchildren in the area will be able to monitor the rivers in their neighbourhoods and record the data they gather through a mini water questionnaire based on work from the Grade 7 school curriculum.I recently attended a conference in Mexico City on digitilisation and sustainable development. In one of the workshops, researchers and public officials from Brazil, Indonesia, South Africa, India and Europe discussed how local government could use digitalisation and AI to improve water management in cities where water was becoming increasingly scarce and more polluted.
They also focused on the difficulties and costs of achieving this. An important takeaway from the workshop was the realisation that major cities in parts of Latin America, Africa, Europe, India and Central Asia share many similar water management issues. These ranged from aging infrastructure, a lack of rain harvesting, dwindling rainfall and the inability to handle extreme weather conditions. Greater cooperation between cities, governments and researchers worldwide is required to fully realise the impact of digitalising the water management sector.
* Seeliger is a researcher at the Unit for Environmental Ethics and the Water Institute at Stellenbosch University
Weekend Argus