Climate change is regarded as one of the critical factors surrounding the changing distribution and decreased availability of water resources (Kundzewicz et al., 2008). As such, the rapid rise in surface temperatures since 1970, surpassing any 50-year interval over the previous two millennia as recorded by the Intergovernmental Panel on Climate Change (IPCC, 2021), may threaten the future sustainability of global water resources. However, the IPCC noted that present and projected climate changes are attributed to global warming and its related increases in atmospheric greenhouse gas emissions, owing to recent population and economic growth driving changes in human activity (IPCC, 2001; 2007; 2013; 2021). Therefore, understanding the relationship between anthropogenic activities and climate is essential for quantifying and predicting climate change, and thus, its potential impact on water resources.
Human activities contribute to climate change as they play a major role in altering hydrological circulation (Kuchment, 2004) and influence hydro-climatic variables such as precipitation, temperature, and precipitation extremes (Ahn and Merwade, 2014). In turn, Dale (1997) noted that climate change and human activities are interrelated as climate change could alter the politics, social attitudes, and affluence governing the choice over anthropogenic activities. In addition, as water resources are central to the predicted impacts of climate change (IPCC, 2007), streamflow is regarded as one of the most crucial resultants for water resources management (Dey and Mishra, 2017). Given that human activities and climate variability impact streamflow (Zhang et al., 2011), their effects need to be delineated and mitigated to ensure water resources sustainability, especially for countries that are vulnerable to climate change impacts such as South Africa.
South Africa is deemed at risk to the projected effects of climate change (Kusangaya et al., 2014). Climate change predictions suggest that the country will continue to warm at a rate significantly higher than the 0.15°C per decade observed over the 20th Century (Engelbrecht et al., 2015), resulting in temperature rises of 5 to 8°C over the interior parts by the end of the 21st century, with reduced increases occurring over coastal regions (IPCC, 2007). Moreover, South Africa is expected to suffer from reduced rainfall in the future (ASSAf, 2017), which is increasingly worrying given that the 490 mm it already receives is less than half of the global average (WWF-SA, 2016). South Africa's geographic location in the African continent's southern region may further exacerbate its climate change impacts. Africa has experienced surface temperature increases above the global average (IPCC, 2021) and is regarded as one of the most vulnerable continents to climate change due to its low adaptive capacity (IPCC, 2001; Callaway, 2004). Furthermore, southern Africa is considered the most vulnerable region in Africa to climate change impacts (IPCC, 2007) due to large spatial and temporal variability in climate (IPCC, 2007; Gallego-Ayala and Juizo 2011). Subsequently, climate change impacts on southern Africa's water resources are expected to be even more pronounced than previously anticipated (Kusangaya et al., 2014), posing a serious threat to South Africa's water resources, given its climate change outlook (Ziervogel et al., 2014).
The developing trend in South Africa's climate towards lower rainfalls and higher temperatures leads to decreased terrestrial moisture (Graham et al., 2011; Engelbrecht et al., 2015), ultimately affecting streamflow components (Legesse et al., 2010). Precipitation directly affects the quantity of water entering a hydrological system (Trenberth, 1999); therefore, decreases in precipitation has an adverse effect on streamflow. Temperature directly affects the quantity and rate of evaporation (Arnell and Liv, 2001); hence, due to higher temperatures, a rain falling overland may be quickly evaporated back into the atmosphere, subsequently reducing water availability for run-off and infiltration (Gleick, 1989). Interception and water uptake from vegetation further exacerbate water's availability for run-off and infiltration, thereby reducing groundwater recharge (Healy, 2010). Decreases in groundwater recharge translate to lowering of the water table and, hence, reduced subsurface flow into the stream network (Matalas et al., 1998). Consequently, streamflow projections for South Africa indicate a substantial decrease by 2050, compromising access to water for human consumption, socio-economic development, agriculture, and the aquatic environment (Kusangaya et al., 2014).
Human practices may compound South Africa's predicted water availability and accessibility shortages by aggravating the quality of its water resources. Mining, recreational, and agricultural activities, and urban and industrial developments significantly alter the quality of natural flows (Edokpayi et al., 2017). In addition, poor wastewater disposal may amplify this problem as many developing countries such as South Africa do not treat wastewater apart from urban wastewater treatment plants. Wastewater treatment plants often discharge their effluents directly into or nearby streams, which is concerning for South Africa's water resources, given that more than half of its wastewater treatment plants do not treat wastewater to acceptable standards (Edokpayi et al., 2015). Resultantly, the streams draining South Africa's urban settlements are characterized by poor water quality. South Africa's population and industrial growth may intensify this issue as it will likely lead to increased wastewater generation (Edokpayi et al., 2017).
The Upper Crocodile River Basin (Fig. 1a) exemplifies the climatic and anthropogenic factors affecting the current and future sustainability of South Africa's water resources. The basin is characterized as having the greatest human impact in South Africa due to the urban sprawls of northern Johannesburg and southern Pretoria (Tshwane) (Department of Water Affairs and Forestry [DWAF], 2008). As a result, there is a considerable water demand from various economic sectors in the basin (Abiye et al.,2015), totalling 556 million m3/year (DWAF, 2008). However, water resources within the UCRB could not meet this demand and it has required enhancement from the neighbouring Vaal River Basin to the south (Abiye et al., 2015; Leketa and Abiye, 2019). The transfer of 550 million m3/year of water from the Vaal Basin subsequently increased wastewater disposal from treatment plants to the rivers draining the UCRB, which is part of the Limpopo River Basin (Leketa and Abiye, 2019), thereby exacerbating the quality of the UCRB's surface water resources (DWAF, 2008). These water quality issues likely most afflict the Hartbeespoort Dam, one of the most severely eutrophicated water bodies in South Africa (Cukic and Venter, 2012), as it is the first restrictive flow body downstream from Johannesburg and Pretoria.
The UCRB is, therefore, expected to be adversely affected by South Africa's envisaged climate and human activity changes. The predicted decrease in rainfall and increase in temperature, coupled with the expected exponential population growth, and associated effluent discharge, pose significant threats to groundwater and surface water availability (Kusangaya et al., 2014). Therefore, managing the UCRB's water resources in response to climate and human activity changes is critical for ensuring water quality and sustainability in the basin.
The Soil and Water Assessment Tool (SWAT) hydrological model has been applied successfully in different regions with variable climatic setting to assess the impacts of changing climates and land-use practices on basin hydrology (Arnold et al., 2012). In South Africa, SWAT has been used extensively to simulate basin hydrology in watersheds with various characteristics and climates (Govender and Everson, 2005; Gyamfi et al., 2016; Thavhana et al., 2018; Mengistu et al., 2019). However, no previous SWAT South African studies have examined SWAT's applicability in simulating streamflow in a large catchment with a pronounced human impact under climate and anthropogenic activity changes. Therefore, this study used the SWAT model to simulate basin hydrology in the UCRB to assess the impact of changing climate and land-use practices on streamflow as a means to justify present and future water resources decision-making in the basin.
The UCRB, covering an area of 6336 km2 between the Gauteng and North-West Provinces, is a catchment that constitutes part of the Crocodile West and Marico Water Management Area as per the Department of Water and Sanitation (DWS) classification. Due to higher elevations in the south of the basin (Fig. 1a), the UCRB is drained in a northerly direction. The confluence of the Jukskei and Hennops Rivers in the eastern part of the basin forms the Crocodile River, which contributes to 90% of the Hartbeespoort Dam's inflow (Leketa et al., 2018), while the Magalies River provides a limited contribution to the Hartbeespoort Dam from the west. The Crocodile River flows northward from the Hartbeespoort Dam into the Roodekoppies Dam before eventually joining the Limpopo River via the Lower Crocodile River Basin (Leketa et al., 2018). Apart from the Hartbeespoort and Roodekoppies Dams, the UCRB contains two other significant water bodies: Rietvlei Dam and Buffelspoort Dam (Fig. 1a) that increased the storage of surface water.
The UCRB forms part of South Africa's interior and is characterized by a subtropical highland climate (Leketa et al., 2018) with a low yet highly variable mean annual rainfall of 700 mm (Abiye, 2011) and mean annual evapotranspiration of approximately 1700 mm (DWAF, 2008). Plots concerning average monthly precipitation and maximum and minimum surface temperatures for the basin's weather stations (Fig. 1b) indicate a directly proportional relationship between rainfall and temperature. Resultantly, the UCRB experiences cold, dry winters between May and July (Leketa, 2019), with winter temperatures ranging from a minimum of 1°C to a maximum of 20°C between April and September (DWAF, 2004). In contrast, the UCRB generally receives rainfall between the summer months of October and March, which occurs as convective rainfall in the form of afternoon thundershowers and occasional hailstorms (Barnard, 2000; DWAF, 2004; Leketa, 2019). This rainfall is instigated by the hot temperatures experienced during summer, commonly ranging from a minimum of 10°C to a maximum of 30°C (DWAF, 2004).
Geological and hydrogeological setting
Underlying the UCRB are rocks of the Archean Basement Complex, the Witwatersrand, Ventersdorp, and Transvaal Supergroups, the Bushveld Igneous Complex, and the Karoo Supergroup, which have a general northerly younging direction relative to Johannesburg (Barnard, 2000; Leketa, 2019) (Fig. 2). The Archean Basement Complex is defined by granodiorites, gabbros, granitic gneiss, and serpentinites (McCarthy and Rubidge, 2005) of the Kaapvaal Craton. Uncomformably overlying the Kaapvaal Craton, the Witwatersrand Supergroup is sequentially divided into the West Rand Group, comprising quartzites and ferruginous magnetic shales, and the Central Rand Group, composed of quartzites, shales, and conglomerates (Pretorius, 1976). The Ventersdorp Supergroup overlies the Witwatersrand Supergroup and outcrops as tuffs and andesites which belong to the Klipriviersberg Group (Barnard, 2000) within the UCRB. The Transvaal Supergroup consists of quartzites of the Black Reef Formation and dolomites of the Malmani Subgroup, which comprise the Chuniespoort Group (Eriksson et al., 2006; Leketa, 2019). The Pretoria Group overlies the Chuniespoort Group and is represented by several formations, namely: the Timeball Hill shale and quartzite, Hekpoort andesite, Strubenkop shale, Daspoort sandstone, Silverton lava and shale, Magaliesberg quartzite, and Rayton shale, sandstone, and volcanic rocks (Eriksson et al., 2006). The Bushveld Igneous Complex outcrops in the northern part of the UCRB and consists of gabbros, norites, and anorthosites of the Rustenburg Layered Suite; Nebo-granites of the Lebowa Granite Suite; and granodiorites of the Rashoop Granophyre Suite (Cawthorn, 2006). The Karoo Supergroup that covers the geological sequence is composed of mudstones, sandstones, and tillites of the Dwyka Group and sandstones, shales, and coal of the Ecca Group (Leketa, 2019), which have small outcrops in the eastern part of the basin.
From a structural perspective, the UCRB has undergone deformation resulting in numerous shear zones and strike-slip faults which penetrate different rock units (Abiye, 2011). Barnard (2000) classified the hydrogeological properties of the UCRB into four aquifer types: fractured aquifers, karstic aquifers, intergranular aquifers, and intergranular and fractured aquifers. Fractured aquifers are commonly found in the granitic gneisses of the Archean basement and in the Witwatersrand Supergroup quartzites (Leketa, 2019). In contrast, karstic aquifers are found within the Malmani Subgroup of the UCRB, which Abiye (2011) noted to occur at greater depths and have higher productivity in comparison to fractured aquifer systems within the UCRB. Intergranular aquifers dominate along riverbanks and in the weathered zones of granites, whilst intergranular and fractured aquifers occur in the basement granites and Bushveld Igneous Complex in the UCRB (Barnard, 2000) (Fig. 2).