Achievements

IDENTIFYING SOURCES AND MECHANISM OF GROUNDWATER RECHARGE IN THE CAPE FLATS, SOUTH AFRICA: IMPLICATIONS FOR SUSTAINABLE RESOURCE MANAGEMENT

Updated :10,17,2012

*Adelana M.A. Segun1, Xu Yongxin2 & Adams Shafick3

1 Earth Sciences Dept., University of the Western Cape, Bellville 7535, South Africa. (e-mail: sadelana@uwc.ac.za)

2 Earth Sciences Dept., University of the Western Cape, Bellville 7535, South Africa. (e-mail: yxu@uwc.ac.za)

3 Earth Sciences Dept., University of the Western Cape, Bellville 7535, South Africa. (e-mail: shadams@uwc.ac.za)

* Corresponding author

 

Abstract: The Cape Flats represents a region of broad coastal sand between the Cape Peninsula and the mainland. The sand covers an area of approximately 630 km2, extending in a northerly direction along the west coast.  It is a highly productive area with intense water use and with multi-stakeholder interest in water. Due to the location of springs close to the urban centre the water resource system here has high community and aesthetic values. Therefore, to properly manage groundwater resources, in relation with the available surface sources, accurate information about the inputs (recharge) and outputs (natural discharge and pumpage) within the groundwater basin is needed so that the long-term behaviour of the aquifer and its sustainable yield can be estimated or reassessed. To assist in this effort, a study of this nature was undertaken to compile and reassess the various recharge estimates for the Cape Flats aquifer in the Western Cape of South Africa. This study provides some highlights and attempts to elucidate the path towards sustainable use of groundwater in the region as a whole. Although the focus in this research is on the Cape Flats aquifer, the results and analysis presented are believed to be characteristic of the entire Coastal Plain sands in the Western Cape. As a first step towards this effort, this research work highlights some key groundwater recharge methods in the Cape area at different scales, such as regional soil-water budget and groundwater modelling studies, region-scale groundwater recharge studies on the TMG aquifer, as well as field-experimental local studies, including some original new findings, with an emphasis on assumptions and limitations as well as on environmental factors affecting recharge processes.

Keywords: Soil-water budgets, Groundwater recharge, Water management, Cape Flats aquifer, South Africa

 



Introduction

Estimating recharge is essential in any analysis of groundwater systems and the impacts of withdrawing water from the aquifer. Identification of all the recharge mechanisms and the estimation of the magnitude of the different components of recharge are now recognised as one of the most important aspects of groundwater resource studies. In water resource investigations, groundwater models are often used to simulate the flow of water in aquifers, and, when calibrated, may be used to simulate the long-term behaviour of an aquifer under various management scenarios. Better estimates of recharge and its spatiotemporal distribution can increase the accuracy of ground-water-flow models and contaminant transport (French et al. 1996) and predictions of water-level trends.  Thus, without a good estimate of recharge, the impacts of withdrawing groundwater from an aquifer cannot be properly assessed, and the long-term behaviour of an aquifer under various management schemes cannot be reliably estimated (Sophocleous 2004).

In the Western Cape, agricultural sector is one of the largest users of water resources. But rapid economic development and population growth is generating increased pressure on water supplies; that may soon lead to greater dependence on groundwater. The growth in urban water demand in the Greater Cape Town Metropolitan Area was projected to increase from 243 million m3 in 1990 to 456 million m3 in 2010; whereas for irrigation water demand the increase is from 56 million m3 in 1991 to 193 million m3 in 2010 (NINHAM SHAND 1994). The total urban and irrigation demand was estimated at 470 million m3 in 1998 and is expected to reach 664 million m3 by 2010. Over 60% of the annual urban demand and 90% of the irrigation demand occurs in summer. Moreover, the shortage of surface water, which is fully used in the study area, makes groundwater a potential source for development. Therefore, the shortage of water resources could become a serious restraining factor for economic development of the Western Cape if proper management schemes are not in place. The most prominent and, may be, the less developed primary aquifer in the Western Cape underlies most part of the metropolitan area of the City of Cape Town and its suburbs. Due to the increasing tendency to develop and exploit this aquifer for municipal water supply the knowledge of the recharge rates and characteristics becomes a significant issue. Therefore, to properly manage groundwater resources, in relation with the available surface sources, accurate information about the inputs (recharge) and outputs (natural discharge and pumpage) within the groundwater basin is needed so that the long-term behaviour of the aquifer and its sustainable yield can be estimated or reassessed. To assist in this effort, a study of this nature is being undertaken to compile and reassess the various recharge estimates for the Cape Flats aquifer in the Western Cape of South Africa. This paper provides some highlights and attempts to elucidate the path towards sustainable use of groundwater in the region as a whole.

Study Area Location and Description

The Cape Flats falls within the region of the Western Cape and belongs to a semi-arid climatic zone. It falls under the City of Cape Town Water Management Area (CMA). Presently, most of the area underlain by the Cape Flats aquifer is built up, from the City main bowl to the northern and southern suburbs. The City of Cape Town is on the western edge of the Cape Flats with part of the city and its suburbs underlain by Cenozoic sands. The Cape Flats covers a surface area of 630 kmand is positioned between 33˚30΄ and 34˚10΄ south-latitude and between 18˚20΄ and 19˚00΄ east-longitude.  According to physiognomy and natural conditions, the Cape Flats is a coastal plain sand and essentially lowland with the highest dune only at 65 m above sea level. The geographical location and general features of the sand-covered coastal plain known as theCape Flats are shown in Figure 1. The Coastal Plain is composed of coarser grained materials, such as sand and gravel, with areas dominated by silts and lenses of clay. Furthermore, being flat and low-lying areas, the Cape Flats (like most Low Plains) is difficult to drain, and could therefore lead to soil salinization, which decrease crop yields and limits farming to salt-tolerant crops. The climate of Cape Town reveals annual precipitation of the Cape Flats area varies between 400 and 500 mm, with an arid period from November to April, and that the mean annual temperature is approximately 17˚ C. The Cape Flats area is bordered to the north by a region characterized by an extremely low amount of annual precipitation, decreasing from the interior to less than 250 mm along the northern coast (Schalke 1973).  Rainfall is often insufficient to meet the water demand of crops in the whole growing period or part of it and irrigation is therefore crucial for obtaining high yields.

The Cape Flats is within the sphere of the major catchments within greater Cape Town area (Fig. 1), where runoff is mostly generated in the mountain ranges in the southeast and the Table Mountain and CapePeninsula Mountains in the southwest. Sandy lowlands with minimal runoff and a high water-table extend over the central area. The drainage towards the south takes place by the Eerste River and by the Zeekoevlei into False Bay, whereas to the north the Salt River and Diep River flow into Table Bay (Fig.1). A feature of the Western Cape is the high river flow that is experienced during the winter (about 85% of the total). Land use within the greater Cape Town consists of urban and peri-urban areas, industry, farming (vineyards, fruit, vegetables, wheat and livestock), forestry and nature conservation. Future land use analyses indicate that residential land will increase by 25 % and business/commercial land by 125 % (CCTM 2005). Residential areas vary in density, and are both formal and informal. Informal settlements take the form of shacks and squatter camps, although the city management strives to reduce this yearly with formal housing units. The total population of the Cape Town area was estimated at 3.2 million in 2001, with the highest density occurring on theCape Flats. The average population growth rate in 2001 was about 3 % per annum. However, the impact of HIV/AIDS and tuberculosis is expected to reduce this growth rate to 1.2 % in 2010 (CCTM 2005). Cape Townhas a strong and diverse economy dominated by industry (textile and clothing, metal and steel, petroleum, glass) and other activities like tourism and agriculture.



 

 Figure 1: The Cape Metropolitan Area with major catchments and settlements over Cape Flats sands



Geology of the Cape Flats

Regional geology of the Western Cape reveal that flat low-lying areas occur in three regions, namely the Cape Flats, the Sandveld region of the West Coast, and the undulating landscape of the north-north-eastern parts of the CMA. The first two are overlain by significant depths of Quaternary calcareous sand while the third region undulating, north-north-eastern area of the CMA is underlain primarily by shale. A variety of soil types occur. They are generally nutrient rich, fertile and amenable to agriculture.

The large undulating sandy area connecting the hardrock of the Cape Peninsula with the mainland is known in the literature as the Cape Flats (Schalke 1973).  It is a component of the ‘Late-Tertiary and Recent sands’ unit, which are in places up to 50 m thick (i.e. below the present sea level). Although this cover is rather thin in relation to its wide lateral extent, practically no outcrops occur. On the Cape Flats (and along the coastal plain between Cape Town and Saldanha) are essentially sediments of Quaternary age that blankets the Neogene deposits (Theron et al. 1992).

The basement of the Cape Flats is composed of Precambrian and Palaeozoic rocks belonging to the Cape granite, the Malmesbury Formation, and the Table Mountain Sandstone (Schalke 1973, Malan 1987, Theronet al. 1992). The Cape Flats is assumed to have been developed after the closure of the ‘Cape Strait’, which at one time united False Bay with Table Bay, by lowering of the sea-level and a probable rise of the basement (Walker 1952, Schalke 1973). The sands are derived from two sources: (1) weathering followed by deposition, under marine conditions, of the quartzite and sandstones of the Malmesbury Formation and Table Mountain Group; (2) the beaches in the areas, from where Aeolian sand was deposited as dunes on the top of the marine sands.  The marine sands were deposited in accordance with the prevailing sea level and the sand body is horizontally stratified with several lithostratigraphic units identified (Theron et al. 1992). The process of sedimentation was initiated in a shallow marine environment, subsequently progressing into intermediate beach and wind-blown deposits, and finally to Aeolian and marshy conditions, which led to the formation of peaty lenses in the sands. According to Henzen (1973), portions of the area of the Cape Flats (particularly along the False Bay coast between Muizenberg and Macassar) are covered by calcareous sands and surface limestone deposits. Silcrete, marine clays and bottom sediments of small inland vlei deposits also occur sporadically (Hartnadyand Rogers 1990). The detailed geology of the Cape Flats has been described in Henzen (1973) and the full description of the various lithostratigraphic units presented in Theron et al. (1992).

Hydrogeology

The general hydrogeology of the Cape Flats aquifer was described by Henzen (1973) and Gerber (1976) as regionally unconfined, although interbedded clay and peat layers produce semi-confined conditions in places.  The Cape Flats aquifer pinches out against impermeable boundaries in the east, west and north while the southern boundary is defined by the coastline extending along the False Bay, between Muizenberg andMacassar. Aquifer permeability and specific yield have been calculated on the basis of 20 major abstraction and recovery tests (Gerber 1981).  The aquifer sands are well sorted with hydraulic conductivities of 30-40 m/d in the central area and 15-50 m/d in the eastern portion (Gerber 1981). Specific yield range from 0.02 to 0.25 while transmissivity values are from 50-650 m2/d but typical values between 200 and 350 m2/d. The effective porosity was of the order of 0.10 to 0.12 but values of 0.25 are found over large areas (Gerber 1981). Recharge volume in the study area is approximately 3.6 x 107 m3 per annum. The groundwater flowpaths are generally west to southeast towards the coastline, except for saline waters that tends to permeate landwards at the periphery of the coast and estuarine shoreline. Actual evapotranspiration in the study area after Turc’smethod (Domenico and Schwartz 1998), is 491.7 mm/a while the other method based on White (1932) results in an average of 119.6 mm/a.  Potential evapotranspiration for the study area has been estimated using the well known empirical formula of Thornthwaite & Mather (1957) to be 836.3 mm/a and the monthly mean is 69.7 mm.

The Cape Flats aquifer is recharged principally from precipitation within the basin. Precipitation data from 1841-2005 measured at the Cape Town Astronomical Observatory were analyzed. The mean of yearly rainfall over this period is 619 mm and variability of annual distribution pattern illustrated with the mean of ten years (Figure 2).  Annual means of rainfall for other stations around Cape Town are as follows: SommersetWest (576.1 mm), University of the Western Cape (414 mm), Kirstenbosch (1381.9 mm).  The average values are tabulated in Table 1.  Hydrographs of precipitation, potential and actual evapotranspiration are shown in Fig. 3.  The monthly values of precipitation decrease in the same direction except for Kirstenbosch, located at the foot of the mountains. The annual difference can be attributed to local climatic differences. The rainfall is largely controlled by topography although to the north of the Western Cape, this climate regime grades into semi-desert whereas to the east the climate becomes less seasonal and tends towards sub-tropical on the coast.



Table 1: Rainfall stations within Cape Town Municipal used in this study

Station

Latitude

Longitude

Altitude (m)

Period of

Observation

No. of years

Annual Rainfall (mm)

Max.

Min.

Mean

Cape Town(SAAO)*

33.97

18.6

42

1841-2005

164

1037.7

229.4

619.1

SommersetWest

34.15

18.85

10

1934-2005

71

1191.1

104.4

576.1

Kirstenbosch

34.00

18.43

160

1914-2005

91

2295.9

893.6

1381.9

§UWC, Bellville

33.93

18.62

42

2002-2005

4

662

393.

414

*SAAO = South African Astronomical Observatory; §UWC = University of the Western Cape (Test site)



Infiltration from rainfall is the main source of recharge to the aquifer. The groundwater levels are deepest during the early March (average 4.8 m bgl) and begin to rise during June/July. During July and August, the water levels rise at a constant rate. This rise accelerates in August/September, and about 60% of the annual water level rise occurs within this month. There is little or no rain from October to March accounting for <5% of the total annual rainfall and makes no contribution to the groundwater recharge.



Figure 2: Variability of annual rainfall in Cape Town from a long-term record (1841-2005)


Figure 3: Average monthly values of precipitation, potential and actual evapotranspiration in Cape Town Airport (1933-2005)      



Theory and Methods

There are a number of techniques used in groundwater recharge estimation. Generally, these are divided into five categories, with a range of techniques and variations within each category:

(1) Groundwater hydrograph methods. These methods infer recharge from groundwater response to rainfall. In general, water tables need to be shallow and hydrogeological parameters, mostly specific yield, are required as input. Comparison between different land uses is possible with a well set-up piezometer network and monitoring the difference over a number of years. A transect of piezometers can be used to infer longer term recharge to the system.

(2) Soil tracer methods. These provide point estimates of mean annual drainage. In general, water tables need to be deep. Comparison between different land uses is possible provided there are long term land management practices in adjacent areas. Possible soil tracers include artificially applied tracers (e.g. bromide, tritium), tracers from historical events (e.g. tritium, chlorine-36 from nuclear test fallout) and natural tracers such as chloride and stable isotopes of water molecule. The type of tracer to be used depends mainly on the recharge rates with artificial tracers being useful at rates of > 100 mm/year, bomb tracers at rates greater than10 mm/year and chloride over the range.

(3) Soil physical techniques. This includes a range of techniques such as lysimeters, neutron moisture meters, zero flux method, and hydraulic gradient methods. At high recharge rates, lysimeters are perhaps the most technique and while having been used in semi-arid and arid environments successfully. Neutron moisture methods are used as part of water balance techniques, including meteorological station and modeling. Zero flux and hydraulic gradient methods are based on accurate determination of hydraulic conductivity as a function of water content or pressure head, which is often difficult to estimate (Sophocleous 1991).

(4) Groundwater tracer methods. These can be used to infer the mean annual recharge to the groundwater system, as a whole, or parts of the aquifer. The type of tracer depends on the mean residence time of the aquifer, i.e. the recharge rate. The commonly used tracers are carbon-14, CFCs and tritium. These should be part of a whole groundwater study that includes hydrogeological investigations and full hydrochemicalanalyses (Walker et al. 2002).

(5) Modelling techniques. Modelling generally augments other techniques and can not be used by itself to estimate recharge. Simply, there are too many parameters involved in this process. In some cases there may not be enough previous measurements to constrain the modelling. Modelling techniques are widely varied including water balance modelling, crop modelling and groundwater modelling (Morgan & Henrion 1990, Bekesi& McConchei 1999).

According to Walker et al. (2002), having varied techniques available, the best technique to use depends on a number of factors. These include: spatial scale of interest, time scale of interest, magnitude of the recharge, accuracy required, cost and access to facilities, time lags associated with processes, whether variability is required and whether predictions of impacts are required. In South Africa the methods previously applied have been summarized in a schematic presentation (Bredenkamp et al. 1995, Xu & Beekman 2003) to provide a logical structure and have been grouped into categories relating to the following:

(i) The unsaturated zone which includes lysimeters studies, soil moisture flow and balances, use of tritium profiles, chloride profiles in the soil overlying an aquifer.

(ii) The saturated zone which includes an analysis of groundwater hydrographs, water balances of delineated aquifers, the analysis of spring flow, the saturated volume fluctuation method, and the cumulative rainfall departure method (CRD).

(iii) Modelling of groundwater flow and the water balance, incorporating the determination of recharge, storativity and transmissivity by inverse solution techniques, the direct parameter estimation method involving a multiple linear regression (inverse) fit of water balance parameters, hydrological models based on conceptual hydrological interrelationships.

(iv)       Steady state flow approximations which involves applying Darcy’s law, incorporating the flow through a cross-section of the aquifer.

(v) Rainfall-recharge relationship expressed by a regression type simulation of the groundwater recharge in accordance with some conceptual logic built into the formulae.

(vi)       Natural radioisotopes used to reveal mixing and transient flow within an aquifer system.

(vii)       Natural stable isotopes such as 18O and 2H are commonly used to reveal groundwater characteristics and to distinguish between waters of different origin.

The most important recharge studies carried out in South Africa have indicated the essentials of different methods of analyzing a groundwater system. Recharge estimates (by whatever methods) are normally subject to large uncertainties and errors (Simmers 1988). In addition, the determination of recharge variability in space and time is often high and can create a number of unresolved problems or requiring additional investigations (Sophocleous 1991). The various techniques applied in the quantification of recharge estimates for the study area are discussed in the following subsections.

Water-table Recharge

Water table recharge, often referred to by many authors as water-table fluctuation (WTF) method (Gerhart 1986, Healy & Cook 2002), is based on the premise that rises in groundwater levels in an unconfined aquifer is due to recharging water arriving at the water table. The spatial and temporal variations in the water table have been established from regular head measurements over a two-year period and analysis of a 6-day rain event in the study area. Typical response time between rainfall periods and water table rise is 3-4 days, with rising limbs on hydrographs showing increases of 3-15 cm (Fig. 4). Overall, the trend in hydrograph patterns in the Cape Flats aquifer is very similar, and this has been attributed to the high degree of connectivity between the sand bodies. Water levels within the Cape Flats aquifer show an immediate and marked response to rainfall (typically >30 cm rise within a week), as well as gradual decreases (0.1-1.0 m) associated with longer periods (2-4 months) of  little or  no rain. Between 30 March and 7 April, 1979 there was a pronounced 13 mm rise in the water table and up to 40 mm after recharge events (between 21 March and 21 June). However, between June and September only a rise of 6 mm was recorded. The falling limbs may sometimes take 2-4 months to return to initial levels; perched water bodies were a common occurrence after rain events in places.

Recharge in this way is calculated using the following relationship:

R = Sydh/dt = Sy∆h/∆t

where Sis specific yield, h is water-table height, and t is time.

This method has been applied in several previous studies (Meinzer & Sterns 1929, Rasmussen & Andreasen 1959, Gerhart 1986, Healy & Cook 2002, Xu & Beekman 2003) and the conclusion is that it’s best applied over short period (say, of few hours or days) in regions having shallow water tables that display sharp rises and declines in water levels. Ideally, water-level fluctuations occur in response to spatially averaged recharge (Scanlon et al. 2002). Hence, this method can also be applied over longer period or time intervals (either seasonal or annual) to produce an estimate of change in groundwater storage, sometimes referred to as “net” recharge (Healy and Cook 2002). In order to achieve this, ∆h is set equal to the difference between the peak of the water level rise and the low point of the extrapolated antecedent recession curve at the time of the peak.  According to Healy and Cook (2002) the antecedent recession curve is the trace that the well hydrograph would have followed in the absence of the rise-producing precipitation.


Figure 4(a): Hydrograph for monitoring wells DC182, DC183, DC184 with average monthly precipitation


Figure 4(b): Hydrograph for monitoring well BA081 with average monthly precipitation

 

Figure 4©:  Hydrograph for monitoring well BA083 with average monthly precipitation


Figure 5: Hydrograph of average water level in a borehole and bar graph of weekly rainfall in the study area



A consistent static water level measurement on 6-hourly basis of a well in the study area, with longer record, was averaged weekly and used to plot the water level hydrograph in (fig. 5). The figure shows the average water level in a borehole (BA80) and total rainfall on a weekly basis for April 21, 1979 through March 21, 1981. Water levels are highest in late winter and early spring. Precipitation was not evenly distributed even though rainfall occurrence was almost nearly weekly. Recharge was calculated on a monthly basis using the above equation. As described previously, Sy was estimated to be 0.26 for the study area (Gerber 1976) while ∆h was taken as the cumulative rise in water level for each month (i.e. the sum of all rises that occurred). In order to account for drainage from the water table that takes place during the rises in water levels, water level prior to each rise was extrapolated to the expected position had there been no rainfall event. The rise was then estimated as the difference between the peak level and the extrapolated antecedent level at the time of the peak. The results of the monthly estimates of groundwater recharge for the period March 1979 through February 1981 are shown in Table 2. The water level and rainfall for this site (BA80) was looked into more closely using the daily record. Figure 7a shows rainfall and depth to water table for 22 days during August and September 1979.



Table 2: Monthly change in water level and estimated groundwater recharge for the study area (March 1979-February 1981)

Month

Change in water level,

∆H (cm)

Groundwater recharge,

R (cm)

1979

 

 

April

26.0

6.8

May

15.0

3.9

June

31.0

8.1

July

21.0

5.5

August

19.0

4.9

September

29.0

7.5

October

27.4

7.1

November

28.8

7.5

December

39.9

10.4

1980

 

 

January

31.0

8.1

February

30.5

7.9

March

31.0

8.1

April

17.5

4.6

May

4.0

1.0

June

9.0

2.3

July

4.0

1.0

August

26.8

7.0

September

30.5

7.9

October

0.0

0.0

November

11.5

3.0

December

6.0

1.6

1981

 

 

January

9.0

2.3

February

4.8

1.2

March

9.0

2.3

 


Figure 6a: Rainfall and depth to water table for 22 days during August and September 1979.


Figure 6b: Rainfall and water-level at the same site for parts of July and August 1980.



Figure 6 (a & b) shows rainfall and depth to water table for two successive rainy seasons in site BA80. The water level is in decline for the most of the period, with the exception of a sharp rise that occurred on 23 August in response to less than 2 cm of rainfall. The difference between the peak and the trace of the recession curve is about 0.32 metres. Using the same approach described above recharge for the 6-day period is then estimated to be 0.25 m, which is close to 50% of the total amount of rainfall. The shallower water table and permeable sediments all contributed to the rapid water-level rise at this site. All of the water arriving at the water table this time probably went into storage, at least for a short period of time. Because the rise in the first few days (in Fig 6a) was long and gradual, some water arriving at the water table was likely lost toevapotranspiration or baseflow prior to the time of water level. These losses would not be reflected in the estimated recharge rate. The longer rainless days before each rain event, the higher contribution to rising water level.  But, each rain event does not contribute to the same ratio of the rising water level for a precipitation.

The effective infiltration rate of rain would depend upon several factors such as no rainy days before a rain event, density of rains fall, and precipitation amount as well as the size of reservoir. The expected recharge rate from groundwater level fluctuation could be about 9% based on the slope of the relationship on the figure 7 according to the Bredenkamp (1990) as follow:

RE=r(Py-My)                      (1)

Where, RE is annual recharge, r is annual recharge rate and Pand My are annual precipitation and threshold rainfall, respectively.



 

Figure 7: The relationship between changed groundwater level and precipitation of each rain event.



Recharge From Stream Hydrograph

Groundwater and surface water in South Africa have been perceived and managed as isolated resources. There is, however, a growing recognition that rivers can receive groundwater from underlying aquifers, and this can have significant implications for river quantity and quality. The analysis of groundwater inputs into streams is critical when dealing with issues such as reliability of water supply, water allocation, design of water storages, ecosystem water requirements, contamination impacts or predicting peak stream salinities. Hence, the method of stream flow hydrograph analysis used by Meyboon (1961) and Fetters (1994) was adapted in the study area to differentiate among the various components of stream runoff and also to obtain quantitative information concerning the basic hydrologic equation:

Groundwater recharge = groundwater discharge + change in storage

The base flow recession equation is given by

Q = Qo.e-at                            (2)

Where Q is the flow at some time t after recession started, Qo is the flow at the start of recession, a is a recession constant for the basin and t is the time since recession started.

This equation shows that Qo varies logarithmically with time, t.  A plot of a stream hydrograph with discharge on a logarithmic scale and time on an arithmetic scale will therefore yield a straight line for the base flow recession. The baseflow recession starts with the first slope or the first low value of the recession graph and ends with the first flood. The complete potential groundwater runoff (Qtp) represents the runoff of a complete groundwater recession. It is calculated as follows:

Qtp = (Qo * t1)/ 2.3                             (3)

Qo = runoff at time t = 0

t1 = time of a logarithmic cycle of the recession,

The remaining potential groundwater runoff is calculated as follows:

Qt = Qtp / 10(t/t1)                          (4)

The difference between the remaining potential groundwater runoff (Qt) at the end of one recession and the complete potential groundwater runoff (Qtp) at the beginning of the next recession yields the groundwater recharge between the two recessions (Figure 8). Table 3 shows the calculated Qtp and Qt and the storage change in the study area.



 

Figure 8: Monthly runoff of Eerste River at Faure near Strand (2002-2005).

Table 3: Baseflow analysis of rivers in the study area with discharge measurements

River

Qtp (m )

Qt (m )

Recharge (m )

Period

 

Rainfall (mm)

Eerste at Faure

1.97 x 108

1.29 x 107

1.84 x 108

2002-2003

524.3-424.3

Eerste atFleurbaai

5.41 x 107

3.17 x 106

5.09 x 107

 

2002-2003

-

Lourens atStrand

2.13 x 107

5.53 x 106

1.58 x 107

 

2002-2003

589.2-456.6



Conceptual hydrogeologic model

        The first stage in the estimation of recharge in the study area involved collection and collation of existing data on potential controls of recharge, such as climate, hydrology, geormophology and geology, to develop a conceptual model of recharge into the aquifer system. The conceptual model describing the location, timing, and likely mechanisms of recharge into the Cape Flatsaquifer is illustrated in Figure 9.


Figure 9: A conceptual hydrogeologic model of the Cape Flats area.



The source and recharge processes of different types of groundwaters with their isotope and chloride data can be interpreted using Fig. 9, which represents a conceptual hydrogeologic model of the Cape Flatsarea. The model takes into account the following observations: (i) lithology as revealed through a number of boreholes in the study area, (ii) geomorphic variations, (iii) groundwater fluctuations, (iv) qualitative yields of the aquifer, (v) variation of δ 18O and chloride concentration in groundwater. The hydrogeologic model presented above has some similarities with a conceptual model of groundwater flow developed by Gerber (1981) for the Cape Flats aquifer. In this model, on the basis of two aquifer test sites where groups of observation wells were aligned in more than one direction respectively, Gerber presented that there was evidence of anisotropy in the sand deposits of the Cape Flats. In the hydrogeologic model developed in this study, it is believed that this sandy aquifer, which is connected to the vadose zone, carry recharge water laterally a few hundreds of meters to a few kilometers, which could result in ageing of the recharged water. However, in contrast to the above situation, younger ages with similar chemical and isotopic characterization are obtained. This situation may occur due to the fact that the aquifer system is generally unconfined in nature and infiltration and movement of water through the matrix result in apparent young age of groundwater. This situation can be described from the conceptual hydrogeologic model (Fig. 9) in which wells like AB-7 and BA-6 are located where the source of recharge is some distance away. In an actual field situation, the sampled groundwater wells mainly tapping from the Cape Flats aquifer represent the above hydrogeological situation. From hydrochemical and isotopic data and considering extensive fracture zones in the neighboring TMG with good yield (5–7 l/s) prevalent in the area, it is envisaged that groundwater moves laterally through the sands for sufficiently long periods and over distant areas as we have not observed any depth variation. Wu (2005), in his attempt to model the groundwater flow and travel time through the fractured zone of the Table Mountain Group (TMG), estimated groundwater recharge. According to him, recharge is nearly vertical and the travel time to the saturated zone is dependent on the thickness of the unsaturated zone and effective travel path porosity. The flow path may be in a rock matrix or in fractures. Further, a significant flux of recharge occurs through fractures under saturated or nearly saturated conditions. Then, infiltrated water is expected to travel laterally through the Cape Flats aquifer until intercepted by deep wells in the weathered shale and granite of theMalmesbury. A higher infiltration rate is assumed to take place through fractures while a lower value of infiltration moves through the sands (inter-mixed with silt, clay and peat in places). If this conceptualisation is correct, the pattern of the hydrochemical depth profiles will not necessarily imply the slow passage vertically downward of a ‘front’ of urban recharge. Instead, a given profile could be the product of a complex series of mixing ‘cells’, slowly evolving as water moves generally downdip (and occasionally cross-dip along discontinuities) in the recharge area through the sand aquifer and subsequently into the deeper fractured aquifers. This is subject to further investigations and detailed groundwater evolutional studies.

However, despite the fact that the studied area have fractured hard rock aquifer system (with a complex geologic environment comprising unsaturated rock matrix and multiple fracture network) as its immediate neighboring aquifer, the study has identified dominant recharge processes and flow mechanisms operative in the area. The possible sources of errors that would offset the conceptual data from its real value could be recharge contributions from unidentifiable subsurface regimes as well as contribution from multiple fractured basement aquifers  during pumping of wells near the west and east boundaries of the Cape Flats.

From an agricultural perspective, where irrigation water is available, such areas can be used for intensive market gardening and potato cultivation. However, it is these areas which are currently experiencing an explosion of urban development. This includes the informal, and low-cost formal housing of the Cape Flats, and the middle income development of the West Coast (Table View) as well as developments such as CenturyCity. Such areas do not place any particular constraints on construction. In fact, the nature of the sandy substrate facilitates the digging of foundations and the emplacement of infrastructure (pipes, cables). However, the low-lying nature of these areas and their proximity to the ocean implies a high local water table and, in specific areas, frequent waterlogging and localised flooding in the winter months. Landfill sites, typically sited in such areas in the past, are prone to seepage and groundwater contamination. Fruits and, in places, vines are grown. Provided they are correctly managed (contour ploughing, for example) they are not particularly vulnerable to erosion and will provide sustainable crop yields. However, construction should be undertaken with caution and only after thorough examination of the substrate.

Steps Towards Sustainable Resource Management in the Western Cape

The people of Cape Town, through their citizen committees, their local and regional agencies, and Department of Water Affairs and Forestry (DWAF) under the Integrated Environmental Management (IEM) guidelines issued by the City Council of Cape Town, are unanimous in their desire to extend and to ascertain a basis for the development of future water supplies in the Western Cape. These developments would be subject to a public participation process. This warranted the commissioning of a comprehensive study in 1989 entitled the Western Cape System Analysis (WCSA) by the DWAF in conjunction with the City Council. The Cape Town Municipal Water Plan objectives call for utilization of the Cape Flats aquifer by 2005 and implementation of enhanced water management in targeted areas. The plan also calls for achievement of sustainable management of both surface and groundwater sources within the region by 2008 (in western and eastern catchments, where precipitation and aquifer recharge are generally greater than in the central and northern areas) and for meeting minimum desirable stream flow standards (established at 20 locations regionally) at a frequency no less than the historical achievement. In the long term, it is impossible to extract more water from an aquifer than is recharged to it by seepage from precipitation or surface water bodies and flow from other aquifers. The key steps to move toward sustainable use of groundwater are basic, but a sustainable groundwater management strategy can only be as good as its potential for implementation. Several of the following steps were outlined as implemented in other places (NRC 2000, Sophocleous 2002, 2004, 2002, among others), but they can be extended and applied to the Western Cape conditions:

Improve the knowledge base.

Improve reporting and access to information.

Improve public education and better understand the public’s attitudinal motivations.

Use the ecosystem approach to manage groundwater.

Embrace adaptive management.

Further improve water efficiency and crop productivity.

Exploit the full potential of the Cape Flats aquifer along side with others.

Adopt a goal of sustainable use.

References

[1]      BEKESI, G. & MCCONCHEI, J. 1999. Groundwater recharge modelling using the Monte Carlo technique, Manawatu Region, New Zealand. J hydrol. 224:137-148.

[2]      BREDENKAMP, D.B., BOTHA L.J. & VAN RENSBURG, H.J. 1995. Manual on quantitative estimation of groundwater recharge and aquifer storativity. Rep TT 73/95. Water Research Commission, Pretoria, 419pp.

[3]      BREDENKAMP, D.B. 1990. Quantitative estimation of groundwater recharge by means of a simple rainfall-recharge relationship. In: D.N.LERNER, A.S. ISSAR & I. SIMMERS (eds.) Groundwater Recharge, IAHMemoir 8:247-256.

[4]      CATCHMENT, STORMWATER AND RIVER MANAGEMENT (CSRM) 2005. Annual Report for 2004/2005.

[5]      DOMENICO, P.A. & SCHWARTZ, F.W. 1998. Physical and chemical hydrogeology (Chapter 17), 2nd edn. John Wiley & Sons, Inc. New York, 506pp

[6]      FETTERS, C.W. 1994. Applied hydrogeology, 3rd edn, Merril, Columbus, 691p.

[7]      FRENCH, R.H., JACOBSON, R.L. & LYLES, B.F. 1996. Threshold precipitation events and potential groundwater recharge. J. Hydraul. Eng. 122:573-578.

[8]      GERHART, J.M. 1986. Groundwater recharge and its effect on nitrate concentrations beneath a manured field in Pennsylvania. Ground Water 24:383-389.

[9]      GERBER, A. 1976. An investigation into the hydraulic characteristics of the groundwater source in the Cape Flats. Unpublished M.Sc. thesis, University of the Orange Free State, Bloemfontein

[10]  GERBER, A. 1981. A digital model of groundwater flow in the Cape Flats. CSIR Contract Report C WAT 46, Pretoria.

[11]  HARTNADY, C.J.H. & ROGERS, J. 1990. The scenery and geology of the Cape Peninsula. Guidebook Geocon-gress ’90, Geological Society, South Africa.

[12]  HEALY, R.W.  & COOK, P.G. 2002. Using groundwater levels to estimate recharge. Hydrogeol. J. 10:91–109.

[13]  HENZEN, M. 1973. The reclamation, storage and abstraction of purified sewage effluents in the Cape Peninsula (in Afrikaans). D.Sc. Thesis, University of the Orange Free State, Bloemfontein.

 

[14]  MALAN, J.A. 1987. The Bredasdorp Group in the area between Gansbaai and Mossel Bay. South Africcan journal of Science 83, pp. 506-507.

[15]  MEYBOOM, P. 1961. Estimating groundwater recharge from stream hydrographs. Journal Geophysical Research 66:1203-1214.

[16]  MORGAN, M.G. & HENRION, M. 1990. Uncertainty.  A guide to dealing with uncertainty in Quantitative Risk Analysis and Policy Analysis.  Cambridge University Press.  ISBN 0-521-36542-2, 332 p.

[17]  NINHAM SHAND. 1994. Western Cape System Analysis: Study overview. NINHAM SHAND Consulting Engineers.

[18]  NATIONAL RESEARCH COUNCIL (NRC). 2000. Investigating groundwater systems on regional and national scales. National Research Council, National Academy, Washington DC, 143 pp

[19]  RASMUSSEN, W.C. & ANDREASEN G.E. 1959. Hydrologic budget of the Beaverdam Creek Basin, Maryland. USGS Water-Supply Paper 1472:106.

[20]  SCHALKE,H.J.W.G. 1973. The upper Quaternary of the Cape Flats area, Cape Province, South Africa. Scripta Geol. 15, pp1-37.

[21]  SCANLON, B.R., HEALY, R.W.  & COOK, P.G. 2002. Choosing appropriate techniques for quantifying groundwater recharge. Hydrogeol. J. 10:18–39.

[22]  Simmers, I. (eds) 1988. Estimation of natural groundwater recharge. NATO ASI series C 222, Reidel, Dordrecht.

[23]  SOPHOCLEOUS, M.A. 2004. Groundwater recharge and water budgets of the Kansas High Plains and related aquifers. Kansas Geol. Surv. Bull. 249:1–102.

[24]  SOPHOCLEOUS, M.A. 2002. Interactions between groundwater and surface water: the state of the science. Hydrogeol. J. 10(1):52–67

[25]  SOPHOCLEOUS, M.A. 2000. Quantification and regionalization of ground-water recharge in south-central Kansas: integrating field characterization, statistical analysis, and GIS. Spec Issue, Compass 75(2–3):101–115.

[26]  SOPHOCLEOUS, M.A. 1991. Combining the soil water balance and water level fluctuation methods to estimate natural groundwater recharge: Practical aspects. J. Hydrology 124:229-241.

[27]  THERON, J.N., GRESSE P.G., SIEGFRIED H.P. & ROGERS J. 1992. The geology of the Cape Town area. Explanation on Sheet 3318, Geological Survey, South Africa, 140p.

[28]  THORNTHWAITE, C.W. & MATHER, J.R. 1957. Instructions and tables for computing potential evapotranspiration and the water balance, Publications in Climatology, 10 (3):183–311. Laboratory of Climatology, Drexel Institute of Technology, Centerton, New Jersey, USA.

[29]  WALKER, G.R., COOK, P.G. & GILFEDDER, M. 2002. Recharge/discharge determination. In Fundamentals of Groundwater Science, Technology and Management Vol. 2 – Technology, Australian GroundwaterSchool.

[30]  WALKER, F. 1952. The geology, In: The Cape Peninsula – Maskew Miller Ltd., Cape Town, 1-13. White, W.N. 1932. A method of estimating groundwater supply based on discharge by plants and evaporation from soil. US Geological Water Supply Pap 659-A

[31]  WU, Y. 2005. Groundwater recharge estimation in Table Mountain Group aquifer systems with a case study of Kammanassaie area. PhD Dissertation, University of the Western Cape

[32]  XU, Y. & BEEKMAN, H.E. (eds), 2003. Groundwater recharge estimation in Southern Africa. UNESCO IHP Series No. 64, UNESCO Paris. ISBN 92-9220-000-3.