History of Water Resource Assessment in South Africa

This blog post will focus on the history of water resource assessment.

For the peers in Geog 3057 and anyone out there who happened to stumble upon my blog, you might be wondering, ‘What has the history of water resource assessment got to do with this module? I am here to take a geography module, not a history module!’ Well, I’ll tell you why. The short answer is: For all the analyses that we are able to carry out today, they are all built on the ones in the past. As Pitman rightly wrote in his article, ‘Where we stand todayis a consequence of what has been done before – a prime example of the ‘weightof history’’.

We improve our assessment by learning from our mistakes and insufficiencies in the past and hopefully progress forwards. There are many techniques that would not have been available had the power of computer and methods of calculations not been advanced. They allow the development of hydrological models from the simple black box models to sophisticated and data intensive spatially-distributed models. Of course, some basic techniques are still well and alive today, e.g. mean annual discharge and rainfall and flow regime. They have all been instrumental in the understanding of hydrologic system and the paradigm shift of water resource management (shift from large scale dam to more integrated water resource management).

The methods used to assess the water resources in South Africa from the 1950s onwards have seen significant progress, nonetheless faced with new challenges as well. In the 1952 study done by Midgley, there were no such things as computers! For a millennial like me who has been blessed with the invention of laptops and their applications, it seems insane to carry out all the calculations manually. Later, in 1969, Midgley and Pitman were able to advantage of the main-frame computer at the time as an aid for calculations, although its power is nowhere near to what we have today. Luckily, in the 1994 study by Midgley et al, PC and (Geographic Information System) GIS were available. They allowed the researchers to add in components of land use change and better redefine catchment boundaries. This was fundamental as in the past, the hydrological records in catchment area that had undergone significant land use change would simply be rejected for use in models. This means that the models are able to simulate the actual historical changes on land, allowing a much better calibration in models. Processes such as urbanisation, afforestation, irrigation and abstractions as well as the presence of reservoirs can finally be considered. Additionally, the resolution of maps has improved from 1:50 000 in 1981 to 1:1 000 000 in 1994. We therefore see a major improvement in the ability of hydrological models in less than 50-year time.

It is not just the power of the model that has improved, hydrological theories have also advanced during this period. In the 1969 study, Midgley and Pitman introduced the concept of risk when calculating the relationships between reservoir storage and yield. Mean annual runoff and precipitation were compared for all gauged catchments so zones with similar rainfall-runoff relationships were recognized. In 1981, Midgley et al through using a deterministic rainfall-runoff model were able to extend the flow records given there were available and suitable rainfall data. This increases the volume of data up to 334 flow records, considerably greater than previous studies done in South Africa. In the 1994 study, Midgley et al recognised the importance of land use change and were able to investigate their effects on the hydrological systems.  In Middleton and Bailey (2008), the experience of many practising hydrologists was compiled and they suggested greater attention needs to be paid to the interaction between surface water and groundwater in models. Also, they recognised issues such as deteriorating water quality and runoff reduction caused by afforestation and non-native vegetation. Many more aspects of hydrology have therefore been taken into account in contemporary water resource assessments compared to those in the 1950s.

Despite the advances made in both technology and theories, the quest for better understanding of hydrology has been fought with several challenges in South Africa.

There has been considerable decline in the number of both rain and river gauging stations (see Figure 1 and 2). The number of flow gauge peaked in 1980s and dropped by 100 in less than 20-years-time whereas the number of rainfall stations have returned to the level in 1920s. This is largely due to a reduction in funding for water resource agency such as South African Weather Services (SAWs) and Department of Water Affairs (DWA) and lack of training in education system (Herold 2010). This presents a huge problem for the researchers if the trends are to continue. As Pitmann (2011) has stated, ‘all of the computing power in the world is useless without the appropriate data to be processed.’ This is especially true when the scientists are trying to discern the impact of climate change on hydrological systems. Without good quality data, it becomes even harder to make predictions and prepare for extreme natural events in the future.

Figure1. Number of rain gauges.

Figure2. Number of river gauges.

Processes such as land use changes, population growth, urbanisation, infrastructure leakage and pollution also present challenges to water resource assessment. All put greater pressure on the available water resource. Hydro-ecologist also recognised the role of natural flow regime for the maintenance of ecosystem services. This calls for a much more comprehensive assessment of water resources, not just simply looking at the volume of water. To able to do so, we absolutely must encourage the government to invest more in the hydrological monitoring network and personnel!