Swaziland's First National Communication to the United Nations Framework Convention on Climate Change

United Nations Framework Convention on Climate Change National Report on Climate Change National Report on Climate Change

Chapter 4 Vulnerability and Adaptation

4.3 HYDROLOGY AND WATER RESOURCES

4.3.1 Introduction

Water Resources constitute a major sector for climate change impact assessment for Swaziland due to the sector's importance in supporting the country’s socio-economic development. The objective of this study is to assess the extent and magnitude of impacts due to climate change on the country's water resources with a view to identifying viable measures for adaptation in order to minimize the anticipated impacts. The study examines the performance of surface water, ground water and water quality under climate change conditions.

4.3.2 Baseline Scenario

4.3.2.1 Surface Water

Komati and Mbuluzi are the major basins that contribute to sugar cane irrigation The third basin of economic significance is the Usutu, which also supports sugar cane irrigation and hydro power generation. The two southern basins (Pongola and Lubombo) are smaller and relatively underutilised. All the rivers, with the exception of the Komati, Usutu and Lomati basins originate within Swaziland.

Figure 4.5: drainage basins in Swaziland

Table 4.10 illustrates the country’s major drainage basins and their capacities. All the rivers, save the Lubombo flow from the west to the east and hence traverse the four physiological regions of the country. The catchment areas for the Usutu, Lomati and Komati basins include portions of the basins in the Republic of South Africa.

Table 4.10: Major drainage basins of Swaziland and their corresponding hydrologic variables

The inflows and outflows show that the runoff generated within Swaziland is about 2706 million m3 of water per annum while the mean annual precipitation (MAP) is 850 mm which is equivalent to 14800 million m3 of water per annum. This means only 18% of the rain water is transformed to runoff whilst the remainder could be lost through evaporation and to aquifer recharge.

4.3.2.2 Ground-Water

The country's ground water is an important resource that is of use especially in communities with very low water availability. There is an increase in demand and use of the ground water resource by communities in the rural and peri-urban areas particularly in the dry periods. Groundwater is also a source of stream flow especially during the dry winter months.

It is estimated that groundwater recharge ranges between 5 to 20% of the average rainfall in Swaziland. It is also estimated that the ground water resource potential is equivalent to a sustained flow of about 21 m3/s.

The potential for groundwater resource is highest in the Highveld and Middleveld regions. Of this potential, only 6% of the ground water has been tapped through existing boreholes which number about 1500. Average yield from the boreholes is about 1.4 m/s, with others giving rates as high as 20 ℓ/s. Most of the springs are concentrated in the Highveld and Middleveld regions.

4.3.3 Water Resources Development

Water resources development in the country has been enhanced by the construction of dams. The seven main dams in the country have a total capacity in excess of 230 million cubic metres (MCM). The Mnjoli dam is the largest man made lake in Swaziland with a capacity of 135 MCM. The ministry of Agriculture and Co-operatives is involved in small dams construction country wide to curb water shortages during dry periods.

To-date, about 500 small dams have been constructed impounding a total of about 130 MCM (Murdoch, 1997). The Maguga dam is currently under construction. Water from the dams is mainly to be used for power generation and irrigation. The major dams in the country are listed on table 4.11.

4.3.4 Current Water Demand

The demands for water in Swaziland are mainly for domestic, industrial and agricultural activities. Domestic and industrial water demand does not vary much with season. Since most of irrigated agriculture is practiced in the Lowveld where rainfall is generally low it follows that demands are high in this area particularly during the winter months. Table 4.12 illustrates summarised sectoral water demand.

Table 4.11: Major dams and their storage capacities in the country

4.3.5 Irrigation Water Demand

The crops that are grown under irrigation are sugar cane, citrus fruits, pineapples, vegetables and corn. It has been established that about 70 000 ha of land is under irrigation in the country. Most of the irrigation activities are located in the Lomati catchment (Ngonini Estates), the Lowveld part of the Komati river (sugar cane), the Lowveld part of the Mbuluzi river (Simunye, Mhlume, Tambankulu sugar estates), Middleveld and Lowveld part of the Usutu river (Ubombo Sugar, Big Bend, Malkerns) and the lower part of the Ngwavuma river.

It is estimated that the irrigation water demand is about 1734x106 m3/year compared to amount of 4270x106 m3 of water leaving Swaziland per annum. After some allowance for use by South Africa and Mozambique it is estimated that some 2670x106 m3/year could still be retained in storage facilities. This could mean that there is enough water to meet current and future irrigation water demand. However, it should be noted that the flow in most rivers is low during the winter months and it is during these months that water is critically required to sustain crop growth.

The need to expand the irrigation acreage is also there as crop production expands. However, due to water scarcity especially in the Mbuluzi catchment, the demand cannot be met. As a result some companies like Simunye sugar company are changing their irrigation water application method from sprinkler to sub-surface irrigation. It is anticipated that this will increase water use efficiency and the saved water could then be used for irrigating new areas, or be made available for downstream users as well as for the sustainability of the river environment.

The current Maguga dam construction came about as a result of water scarcity in the Komati river. It is anticipated that the dam will stabilize the flow regime down stream the reservoir. Various other water development activities are also proposed for the Usutu river.

4.3.6 Industrial and Domestic

Water Demand Urbanization and industrial growth is rapidly taking place in Swaziland with major industrial activities in Matsapha and Mbabane. Population growth is creating an increase in domestic and industrial water demand. Migration of rural population to urban centers is also putting pressure on the water demand. It should however be noted that most of the urban centres are served with clean portable water in contrast to only 40% of the rural population.

Rainfall, and hence the water resource, is naturally unevenly distributed in both time and space and this is the case with Swaziland. While the Highveld and Middleveld including Lubombo regions enjoy enough water especially during the rainy season the Lowveld faces water scarcity related problems. This situation calls for proper water resources planning in future in order to adequately meet demands of all the regions.

Table 4.12: Current Sectoral Water Demand in Swaziland

4.3.7 Water Quality

Industrial activities in major cities and in sugar cane estates are a major concern as far as water pollution in the country is concerned. Leachate from improperly managed solid waste disposal sites in urban and industrial sites do find its way into natural water courses.

Accidental spillages of toxic substances like phenol liquors at times do occur. This pollution threatens biodiversity and is also a health hazard to human beings downstream. Due to intensive agricultural activities the sediment yield is affected in the catchments resulting in poor water quality.

Water pollution monitoring is being conducted by the water resources branch at all major rivers and at all industrial and municipal effluents. It has been established that the agro-industrial and industrial effluents from some industries in Matsapha area have shown an increase in COD and phenol. Crossborder effluent pollution from the Republic of South Africa has on occasions been witnessed, especially on the Ndlotane river.

The lowest concentration of dissolved solids is found in the rivers of the Highveld and the concentration increases to about 150mg/l in the Lowveld.

Best quality water is found in the basement aquifers of the highveld while the worst is in the Lowveld. This is because of the groundwater stagnant conditions in the Lowveld while there is groundwater movement in the Highveld and Middleveld due to gradients resulting from the mountainous topography towards discharge points by springs.

On the average groundwater quality meets the WHO recommended drinking water guidelines. However, there are isolated areas with high concentrations of fluorides and nitrates. Groundwater with high concentration of salts are found in the Lowveld where evapotranspiration rates are high and the rate of ground water recharge is low due to low rainfall in this region. Fluoride concentration ranges from 0.1 mg/ℓ to 18.4 mg/ℓ and nitrate concentrations are up to 45 mg/ℓ.

4.3.8 Usutu Drainage Basin

This study focuses on assessing climate change impacts on the Great Usutu river. This basin is of great socio-economic importance to the country because about three quarters of the population of Swaziland living within and being supported through it. Apart from this fact this catchment was selected for the impact assessment because it is fairly representative of a large part of the country. The current water demand in the catchment is estimated to be 266400 m3 per day which is equivalent to 0.0222 mm/day.

4.3.8.1 Data Requirement

The assessment of the impact of climate change in water resources requires the utilisation of a various data types. The required input data are hydrological (stream flow, sediment load discharge etc.) and meteorological (precipitation, air temperature, wind speed, evaporation, humidity, air pressure, solar radiation, sunshine hours etc.)

Stream flow data are available for a period more than 30 years in many of the gauging stations. The data series is not continuous throughout the base period due to data gaps arising from siltation of the stilling well and as a result of destruction in 1984 by tropical cyclone Domoina. Likewise gaps also exist in the Meteorological data series. The Hydrological and Meteorological data were obtained from the Water Resources Branch and the Department of Meteorological Services respectively.

The stream flow gauging station at Siphofaneni (GS 6) was selected as the catchment outlet for the Usutu river basin. Data from this station covers the period 1958 to 1982 with two years discontinuity after cyclone Domoina. The other data (precipitation and potential evapotranspiration) were made to cover the same period.

The rainfall information that was used in this project was from the following stations: Siphofaneni, Mankayane, Malkerns and Mbabane. A representative station for the Usutu catchment was developed using the arithmetic mean method.

Potential Evapotranspiration Data from Malkerns weather station was used as a representative station for the Usutu catchment. Data, gaps were patched using the method of interpolation. A regression analysis between potential evapotranspiration and average temperature was developed (1964 to 1982). The regression equation that was obtained was used to determine the potential evapotranspiration values for the period from 1963 to 1969. This was done in order to increase the data record length (1963 to 1982). Potential evapotranspiration values could not be extended to 1958 due missing average temperatures.

4.3.9 Methodology

The greenhouse gases effect is expected to cause global warming which in turn will cause changes in average annual precipitation. Generally it is expected that floods now considered rare would occur more frequently in certain regions while drought related and competing water use issues will intensify in other regions (Miller, 1989; Schaake, 1989).

General circulation models provide physically based predictions of the way climate might change as a result of increasing concentrations of atmospheric carbon dioxide and other trace gasses. The GCM's are mathematical representatives of the earth's climate system, and they simulate atmospheric processes at a field of grid points that cover the surface of the earth. The outputs of these models are: temperatures and precipitation values.

4.3.9.1 Hydrologic Models

A model is a conceptualization of a real system that retains the essence of that system for a particular purpose. Every model is an attempt to capture the essence of the complex nature in hydrologic modeling in a manageable way but it is important to recognise that this conceptualization also involves a considerable degree of simplification. Anderson and Burt (1990) contend that, “all models seek to simplify the complexity of the real aspects of a system at the expense of incidental detail”.

A model must remain simple enough to understand and use , but complex enough to be representative of the system being studied. There are many hydrologic models in the literature (Singh, 1995; Anderson and Burt, 1990; Schulze, 1984; Hughes and Sami, 1994; Pitman, 1973).

For the purpose of evaluating the impact of climate change on water resources, the models that are in use usually operate in simulation mode. A riverbasin-monthly water balance model is recommended as the primary approach for assessing climate change impacts on river runoff (IPCC, 1996).

The CLIRUN set of models is the standard water balance tool selected for Country Studies Programme (IPCC, 1996). The WATBALL model developed by Yates (1994) is one of the CLIRUN sets of models and was used in this study.

4.3.9.2 Watball Model

WatBall is a lumped conceptual integrated rainfall runoff model. It has two major components which are: (i) A water balance which describes the water movement into and out of a basin (ii) The computation of potential evapotranspiration (however, potential evapotranspiration can be input directly).

The water balance is written as a differential equation involving input and output, where storage is lumped as a single conceptualised bucket with the components of discharge and infiltration being dependent on the relative storage which is expressed as follows:

Where, Smax is maximum water holding capacity (mm); Pe is effective rainfall (mm/day); Rs is surface runoff described in terms of storage, precipitation over time; Rg is the ground water flow (mm/day); Rb is baseflow (mm/day); Ev is actual evaporation which is a function of potential evapotranspiration (PET), relative catchment storage (z) and time (t in days).

The model contains five parameters which are: direct runoff; surface runoff; subsurface runoff, maximum catchment water holding capacity and base flows.

WatBall accounts for changes in the soil moisture by taking into account precipitation, runoff, actual evapotranspiration while using potential evapotranspiration to derive the extraction of water from the soil strata. It has been established that, any estimate of climate change impacts on water resources depends on the ability of the model to relate changes in actual evapotranspiration to predict changes in the runoff in the stream.

WatBall has been found appropriate for the estimates of the impact of climate change on water resources because it meets the above criteria. Secondly it requires less input parameters compared to other hydrologic models.

4.3.9.3 Application of Watball Model to the Usutu Catchment

The WatBall model has been applied to the Usutu river for the evaluation of the effect of climate change on the water resources in the basin. There are two stages in the application of a rainfall runoff model, that is calibration and verification.

During the calibration stage the model parameters are adjusted by trial and error process till the model closely reproduces the observed stream flow. Ten years of monthly flow data was used during the calibration stage (January 1963 to December 1972).

Figure 4.6: Simulated and observed stream flow at GS6 during calibration (1963-1972)

Model verification on the Usutu drainage basin was performed using monthly stream flow data for the years from January 1973 to December 1982. Other input variables such as rainfall and potential evapotranspiration also covered the same period. Figures 4.6 and 4.7 show the hydrographs of simulated and observed stream flow for calibration and validation respectively.

The WatBall model was also calibrated for the wettest years, driest years and average years. That is, average daily monthly values were used in this exercise. Table 4.13 shows the optimal model parameters for wet years, dry years and the average or normal year.

Figure 4.7: Simulated and observed stream-flow at GS6 during Validation (1973 to 1982)

The above model parameters for the wet years (1969, 1971, 1972, 1973, 1976 and 1978), dry years (1964, 1965, 1968, 1970 and 1982) and the average year were used to simulate the flow given the results of the GCM models (that is predicted precipitation, temperature and thus potential evapotranspiration).

The developed regression equation between temperature and potential evapo-transpiration was used to predict the potential evapotranspiration for year 2075 given the predicted temperatures by GCM models.

4.3.10 Results of the effect of climate change on water resources

The response of the Usutu river to climate change has been evaluated using GCM models which are; the Geophysical Fluid Dynamics Laboratory (GFDL), the United Kingdom Transient Resalient (UKTR), and the Canadian Climate Change Equilibrium (CCC-EQ).

The selected models were used to simulate the temperatures for Swaziland. All the models well simulated the observed temperature values. They therefore, have been found the ideal choice for simulating future climate scenarios for the country.

Results of these GCM models (temperature, rainfall changes and potential evapotranspiration for year 2075) were used as input to the calibrated WatBall model to forecast stream flow for Usutu for the wet years, dry years and the average years for the year 2075 without taking into consideration water abstractions.

Figure 4.8: Simulated and observed stream flow for Usutu river (Dry years)

The results of the runoff simulations are shown in Figures 4.8, 4.9 and 4.10 for dry, normal and wet years respectively. It can be seen from Figure 4.8 that the forecasted flows with inputs derived from the GCMs for the months of October to December are higher than the current observed flows. There after all the forecasted flows are lower than the current observed stream flows for the rest of the months.

The three simulation results show that the forecasted flows are higher during the early summer months (October to January) and lower during the late summer and winter months (February to September). This implies that the country could experience high flows during the early summer months and low flows thereafter. This is due to the fact that the expected climate change will bring high temperatures and therefore, high evapotranspiration and thus low flows during the late summer and winter months.

Figure 4.9: Simulated and observed stream flow for Usutu river (Normal years)

Figure 4.10: Simulated and observed stream flow for Usutu river (Wet years)

Figure 4.11 shows the % annual runoff change for all the GCMs under dry, normal and wet scenarios for all of the climate change scenarios (Low, Medium and High). The interpretation is that the high annual runoff change (Reduction) occurs during the dry scenario under almost all the climate change scenarios (Low, Medium and High) followed followed by the normal and the wet scenario.

Figure 4.12 shows the average of GCMs % annual runoff change for dry, normal and wet scenario for all the climate change scenarios (Low, Medium and High).

The projection therefore, shows a maximum reduction in annual runoff of 12.6% in the Usutu river under climate change conditions which is equivalent to 133.6 million cubic meters (11.35 mm per year).

When this maximum % annual runoff reduction is applied to all the catchments in the country, the predicted annual runoff reduction becomes 350 million cubic meters which is approximately the size of Maguga reservoir.

Figure 4.11: Average Annual runoff change for Usutu river under different scenarios (%)

4.3.11 Interpretation of Results

The models’ stream flow future projections point to highs and lows during the summer and winter months respectively under climate change conditions. Currently the country is experiencing flooding related problems during summer months of wettest years and drought related problems during years with low flows. Therefore, flooding and drought related problems are likely to prevail during the summer and winter months respectively.

The combined effect of high temperatures and low runoff especially during the winter months could adversely affect groundwater recharge particularly in the Lowveld. Therefore, the present salinity of groundwater could be worsened due to the reduced groundwater recharge and high evaporation rate. The low flows during the winter months have the potential to affect negatively the riverine ecological system.

Figure 4.12: Average GCMs annual runoff change for Usutu river (%)

Simulation results have shown a possible water reduction of 12.6% which is equivalent to 133.6 million m3. Therefore, if nothing is planned for the development and management of the water resource, water shortage related problems could prevail by year 2075.

The economy of the country is based on agricultural activities (agro-based economy). The enumerated changes could negatively impact on rain fed and irrigated agriculture, hydropower generation, livestock and other social economic water uses. The water shortage for domestic use will cause poor sanitation conditions and thus the outbreak of diseases. The country therefore could be forced to import food and medicine. The drought conditions will also impact negatively on the biodiversity of the country.

4.3.12 Adaptation options

The flow regime of any river is greatly influenced by human activities particularly land use. Overgrazing which leads to land degradation is emerging as a problem in the country. Poor farming practices also lead to land degradation.

The time horizon of the change that might occur (increased or reduced precipitation) is similar to the time required for planning, approval, funding, construction and economic life of water resource projects (dams, irrigation canals, drainage systems etc, (Schaake, 1989)). Therefore, mitigation strategies should make sense regardless of the direction and magnitude of change.

It has been established that Swaziland could experience a reduction in stream flows under all scenarios (wet, dry and average year) given climate change. Therefore, in view of this the vision for water resources planning, development, operation and management demands the development of policies and strategies that will promote for the growth of water in the future. Adaptation strategies should be directed at developing robust water resource systems as well as techniques to incorporate climate change uncertainties into the long-term planning.

A reduction in annual runoff of the order of 134 million cubic meters has been established in this study for the Usutu catchments and 350 million cubic meters for the whole country. Inter-basin water transfers in the country is not a viable option due to over commitment of the available water resource within each basin.

4.3.12.1 Water Conservation

It has been established that there could be a reduction in runoff under climate change conditions. Therefore, water use sectors will have to adapt to the meager resource that will be available. It has been assumed here that there will be no significant water savings from industrial and domestic water use, currently at 4% of the total water demand. The major consumer of water in the country is irrigation and at 96%. It is expected that large savings in water will come from efficient use of irrigation water.

Currently the land that is under irrigation in the Usutu catchment is estimated at 15000 hectares. The acreage that is under furrow, centre pivot and drip system presently is assumed to average to sprinkler system. The water demand for sugar cane under sprinkler irrigation system is 1400mm per year per hectare.

With technological advancement there might be more efficient irrigation systems in the future. A 20% water saving could be realised by switching from sprinkler to drip irrigation system. This water saving translates to 280mm per hectare per year. Water that will be conserved in the Usutu catchment by the use of drip irrigation system therefore amounts to 42 million m3 per year. This will leave a water deficit of 92 million m3 per year. This water deficit will have to be provided for by the construction of a water storage facilities and other related means.

4.3.12.2 Dam Construction Cost Estimates

The cost estimates for construction of a dam in Usutu catchment in order to meet water demand due to anticipated climatic changes has been derived from the cost estimates for the construction of Maguga dam. The cost for the construction of the Maguga dam is currently at 1.2 billion Emalangeni (US $150 million). The Maguga dam will impound 330 million m3 of water. The cost of a Dam in the Usutu catchment that will impound 90 million m3 could therefore cost 327 million Emalangeni ($40 million) today.

Other water resources adaptation options that are possible for Swaziland in order to deal with the effects of expected climatic changes are as follows:

4.3.12.3 Modification of the existing infrastructure

i. Supply adaptation (installing canal linings, changing location of water intakes, using grating separate reservoirs into a single system, using artificial recharge to reduce evaporation)

ii Construction of new infrastructure (reservoirs, hydro-plants, delivery systems).

iii Alternative management of existing water supply systems (change operating rules, use conjunctive surface/groundwater supply, change priority of releases, physically integrate reservoir operation system, coordinate supply/demand)

4.3.12.4 Demand adaptation

i. Conservation and improved efficiency

ii. Domestic (low-flow toilets, low-flow showers, re-use of cooking water, more efficient appliance use leak repair, commercial car washing where recycling takes place, rainwater collection for non-potable uses)

iii. Agricultural (night time irrigation, lining canals, closed conduits, improvements in measurements to find losses and apply water efficiently, drainage re-use, use of wastewater effluent, better control and management of supply network

iv. Industrial (re-use of acceptable water quality, recycling)

4.3.12.5 Technological change

i. Domestic (water efficient toilets, water efficient appliances, landscape changes, dual supply systems, recycled water for non-potable uses)

ii. Agricultural (low water use crops, high value per water use crops, drip, micro-spray, low-energy, precision application irrigation systems, salt tolerant crops that can use drain water, drainage water mixing stations)

iii. Industrial (dry cleaning technologies, closed cycle and/or air cooling, plant design with reuse and recycling of water imbedded, shift the type of products manufactured)

iv. Energy (additional reservoirs and hydropower stations, low head run of river hydropower, more efficient hydropower turbines)

v. Market/price-driven transfers to other activities

vi. Using water price to shift water use between sectors

4.3.12.6 Land Use Management

Land use is the major factor that affects the runoff in a stream. Therefore, there is a need for the implementation of good land use practices in the country in order to conserve the water resource. This will require the change of attitudes of the people in animal herding and land husbandry principles and strategies. This is very important for the conservation of the water resource for the present and future generations.

4.3.12.2.7 Promoting Regional Partnership

Almost all rivers in Swaziland are international rivers. Therefore, there is the need to establish partnership in the utilisation of international waters through:

The implementation of the protocol on Shared Watercourse Systems in the SADC Region and other bilateral agreements for the benefit of the country through the implementation of joint water resource project between Swaziland South Africa and Mozambique.

Encouraging and stimulating partnership within the context of SADC in conformity with the international Law of the Non-Navigational uses of International Watercourses.

4.3.13 Conclusions

The impact of climate change on water resources in the Usutu river basin has been evaluated using WatBall model. Model parameters were determined during the calibration stage using two sets of 10 years of daily time series data (rainfall, stream flow and potential evapotranspiration).

The results of three GCM models were used in simulating the stream flow of the Usutu catchment in year 2075 for the wet, dry and average year for the natural conditions. All the GCM models are simulating high and low stream flows during summer and winter months respectively. Simulation results in this study are in agreement with the results of other studies that have been conducted in the Southern African region (Schulze and Perks, 2000).

Results of the study suggest that the country could experience high and low flows during summer and winter months respectively under climate change conditions. The overall water deficit under climate change conditions has been estimated to be 134 million m3 per year. Water saving through efficient irrigation water application systems has been estimated to be 47million m3 per year. A water storage facility is therefore, needed to provide 87 million m3, the cost estimate of which has been estimated.

This page was last updated on 11 October 2004