1-1
PART 1
Seismotectonic Models for South Africa:
Synthesis of Geoscientific Information, Problems, and
the Way Forward
Mayshree Bejaichund,1 Andrzej Kijko,2 and Ray Durrheim3
1Council for Geoscience, South Africa
2Benfield Natural Hazards Centre, University of Pretoria, South Africa
3
University of Witwatersrand and Council for Scientific and Industrial Research, South, Pretoria, South Africa

Published in: Seismological Research Letters, 2009, Volume 80, Pages 65-33
Introduction

In South Africa the demand for energy resources, water, and infrastructure has grown
significantly in recent years. Furthermore, many coal reserves that are currently being
exploited will be depleted within the next 20 years. Consequently, plans to provide
alternative sources of energy are underway. Energy providers are slowly moving away
from traditional coal-fired stations to gas-powered facilities, nuclear power plants, and
portable pebble bed modular reactor (PBMR) units. Several dams within the country have
also been constructed to accommodate the growing demand for water. In South Africa, no
regulatory guidelines for seismic design of such critical facilities exist; hence engineers
make use of international guidelines such as Regulatory Guide 1.208, published by the
U.S. Nuclear Regulatory Commission (U.S. Nuclear Regulatory Commission 2007).
Engineers also need to assess the seismic risk to formulate emergency evacuation
procedures and for insurance assessment purposes.
The first step in assessing the seismic hazard and risk for any site is to develop a
seismotectonic model. The area under investigation is divided into smaller zones/regions
of similar tectonic setting and similar seismic potential (Cornell 1968). These zones are
then used in a seismic hazard assessment model to determine return periods of certain

1-2
levels of ground motion at a given site in the area in question. For example, U.S. Nuclear
Regulatory Guide 1.208 (U.S. Nuclear Regulatory Commission 2007) states that regional
seismological and geological investigations should be undertaken to identify seismic
sources and describe the Quaternary tectonic regime. The investigations should include a
comprehensive literature review (including topographic, geologic, aeromagnetic, and
gravity maps, as well as aerial photos), plus focused geological reconnaissance based on
the results of the literature study. Once the regions of active faults have been identified,
more detailed explorations such as geologic mapping, geophysical surveying, borings,
and trenching should be undertaken. Finally, the Quaternary history should be reviewed;
surface and subsurface investigations of the orientation, geometry, sense of displacement,
and length of ruptures should be conducted; and the possibility of multiple ruptures ought
to be assessed.
Seismotectonic models have not yet been developed for South Africa. The
delineation of seismotectonic zones of Africa as part of the Global Seismic Hazard
Assessment Program (GSHAP) in 1999 was based on an analysis of the main tectonic
structures and a correlation with present-day seismicity. Because of the large scale of the
GSHAP project, only regional structures were accounted for in the preparation of the
source zones. Our study was initiated to remedy this knowledge gap. We have completed
four steps, which are described in this paper.
1. Compilation of a catalog of earthquake activity that has been documented in
historical records or instrumentally recorded.
2. Synthesis of geological mapping, magnetic, and gravity surveys, and evidence of
neotectonic activity.
3. Correlation of the seismicity data with the geological, geophysical, and neotectonic
data.
4. Identification of any other data that could help to better define the boundaries of
seismotectonic provinces.

1-3
Earthquake Catalog

South African National Seismological Database

We used earthquake records from the South African National Seismological Database
(SANSD) to map seismicity. The SANSD is a compilation of seismological data from the
South African National Seismograph Network (SANSN), operated by the Council for
Geoscience (CGS). Historical data originates largely from the work of Fernandez and
Guzman (1979) and De Klerk and Read (1988), updated recently by Brandt et al. (2005).
Instrumental data recorded by the SANSN has been published in regular seismological
bulletins since 1977.
Figure 1 shows the distribution of earthquakes above magnitude 3 in the SANSN
database to June 2008. Note that earthquakes in the database are limited to South Africa
and Lesotho. There are in excess of 27,000 earthquakes in the database from 1620 to
June 2008, ranging from ML 0.2 to ML 6.3, with varying levels of completeness relating to
the different stages of development and detection capabilities of the network.

Seismicity Clusters

Fernandez and Guzman (1979) first identified seismicity patterns in South Africa. The
most prominent clusters are:
Historical Earthquakes in the Cape Town area?Earthquakes in this cluster were compiled
from diaries, journals, and newspapers written from 1620 to 1902. The locations are
given as Cape Town because this is where the effects were felt, but the actual
epicenter could be 100 km away or more (Brandt et al. 2005). No earthquakes have
been located in the Cape Town area since instrumental recording began in 1972.
Ceres Cluster?Earthquakes of ML 1?3 are recorded in this region six times per month, on
average. The well-known 1969 Ceres earthquake (ML 6.3) occurred on the western
termination of the Kango-Bavianskloof fault (KBF). Ongoing research in this area has
shown that it is possible that an Mw 7.0 earthquake occurred some 10,000 years ago
in this region.

1-4
Koffiefontein Cluster?An ML 6.2 earthquake occurred here in 1912. Paleoseismic studies
show that a large earthquake (Mw 8) occurred here some 50,000 years ago (Visser
and Joubert 1990; Joubert et al. 1991). Thermal springs are found here.
Lesotho Cluster?The rate of seismicity toward the north of Lesotho increased significantly
after the impoundment of the Katse Dam (Brandt 2000). Toward the west and south
of Lesotho the seismicity is of natural origin.
Witwatersrand Basin Cluster?South Africa has a number of mining regions located in and
around the country (gold, manganese, platinum, diamond, and coal mines). "Mine
tremors", rock bursts or mine collapse events are recorded on seismograph
instruments in exactly the same manner that earthquakes are recorded. If the
instruments are sensitive enough and there are enough instruments set up in the
mining district, it is possible to distinguish between real earthquakes due to natural
causes and "mine tremors" With the national network, until recently; the number of
instruments around the mining regions is now increasing. Researchers can then aim
to determine if events recorded was a natural earthquake or if it was the result of a
mine collapse, which is not within the scope of this work. Answers to these types of
questions have important implications for insurance companies and mining
companies when liability is an issue, particularly if workers are injured and equipment
is damaged. Within this basin, the clusters can further be classified into the Welkom,
Klerksdorp, Carletonville, West Rand, Central Rand, East Rand, and Evander gold
fields. Seismicity in these areas differs due to the different tectonic faults affecting the
regions and differences in mining activities (Singh and Pule 2007; Gay et al. 1995).
Richardson and Jordan (2001) and Finnie (1999) categorized mining events into two
types, A and B. Type A event records are tightly clustered in space and time, occur
within 100 m of the active mining face with an upper magnitude cut-off at 0.5
magnitude units. Type B events are distributed throughout the mining region and are
?friction-dominated? ruptures that occur on existing faults or weak geological
structures with a lower magnitude cut-off at 0 magnitude units. Similarly, McGarr and
Simpson (1997) looked at classifying mining events in terms of ?induced? and
?triggered? events.

To better understand earthquakes of tectonic origin, one needs to know to what extent the
earthquakes recorded in the SANSN are mining related. In site-specific hazard
investigations, care should be taken to carefully analyze the earthquake database. The

1-5
history of mining in the area should be documented, and possible correlations made with
mining-related tremors. One needs to research the active mining areas, obtain blasting
schedules and records of mining tremors encountered, and correlate them with the
database. Similar studies were done by Kgaswane (2002) for the mining regions. This is
beyond the scope of this study and is merely stated as a precautionary measure to be
noted before drawing conclusions on the tectonic origins of earthquakes.

Largest Earthquakes

The earliest recorded large earthquake in South Africa occurred on 4 December 1809
(Von Buchenr?der 1830). Many houses in Cape Town suffered minor damage, and
liquefaction features and fissures in the ground were observed in nearby Blauweberg
Valley. An ML 6.2 earthquake that occurred near Koffiefontein on 20 February 1912 was
felt all over South Africa. The ML 6 Cape St. Lucia event of 31 December 1932 (Krige and
Venter 1933) was located in the sea offshore the Zululand coast. Shocks were reported in
Port Shepstone, Kokstad, Koster, and Johannesburg (some 500 km away). The nearest
point on land to the epicenter was Cape St. Lucia, where Modified Mercalli Intensity (MMI)
of IX was assigned on the evidence of sand boils and cracks in the surface, but the
damage in this area was small because it was uninhabited. The MMI in Pretoria and in
Pietermaritzburg was III and V, respectively. In the severely shaken areas, poor-quality
houses (built of unburned or half-burnt bricks or other low-quality materials) were severely
damaged. In well-built houses, small cracks were occasionally seen but the structures did
not suffer major damage. The phenomenon of site effects was clearly displayed in the
observations of the after effects of this event. Structures built on thick sand were
undamaged, while those built on alluvial sands suffered severe damage. Changing rock
types in the area also had a strong influence on the attenuation of the seismic wave. From
evidence of its effects, Krige and Venter (1933) argue that this earthquake was probably
caused by slip along a fault in the sea striking in a SSW-NNE direction parallel to the
coast.
The strongest and most devastating earthquake to occur in the 20th century was
the Tulbagh earthquake of 29 September 1969, with ML 6.3. This earthquake was widely
felt over the Western Cape, especially in Ceres, Tulbagh, and Wolseley. Serious damage
occurred to certain buildings in the area (amounting to a total of U.S. $24 million). The
damage varied from almost total destruction of old and poorly constructed buildings to

1-6
large cracks in the better-built ones. Nine people were killed and many more were injured.
Green and Bloch (1971) studied the event?s aftershocks, which formed a linear plane that
had little correlation with mapped faults in the area. No surface expression of the fault was
found. The Groenhof fault, originally suspected to be the locus of the earthquake, showed
no evidence of recent displacement. Therefore it is probable that the fault associated with
the earthquake did not intersect the surface.
An Mw 7.0 earthquake occurred on 23 February 2006 in the western province of
Manica in Mozambique. Fenton and Bommer (2006) provide a detailed description of this
earthquake and its associated effects. The earthquake was felt throughout Mozambique,
as well as in parts of the neighboring countries of Swaziland, Zambia, Zimbabwe, and
South Africa, but caused surprisingly little damage and a very small number of casualties.
This is partly because the area immediately affected by the earthquake is sparsely
inhabited. The earthquake was felt as far as Durban, where people were evacuated from
high-rise buildings.
Earthquakes with magnitudes exceeding 5 are listed in Table 1. The most striking
feature of this list is that no earthquake exceeding magnitude 6.3 has been recorded since
1969, the start of the instrumental network. This could be due to a number of different
reasons: 1) the magnitudes of historical earthquakes have been overestimated; 2) the ML
scale of the network is underestimating earthquake magnitude (note that ML saturates at
larger magnitudes, and other magnitudes like Ms or Mw should be used); 3) crustal stress
might have been released in the historical period; or 4) because South Africa is in a stable
continental region (SCR) setting, the return period for earthquakes of ML > 6 is longer than
the approximately 50 years of observation, and stresses need to accumulate before the
onset of another large earthquake. None of these possibilities have been explored in this
study, but an understanding of this issue is crucial in determining the seismic hazard
potential for the country.

Mining-related Events

A large seismic event occurred in the Welkom region in 1976, causing damage to surface
and mine infrastructure, most notably the collapse of a six-story block of apartments. In
1989, a second earthquake caused widespread damage on the surface. Observed and
documented displacements on the President Brand fault in a nearby mine demonstrated
Ceres

1-7
that the origin of this event was local. On 7 March 1992 an event of ML 4.7 occurred near
Carletonville in the Far West Rand area. Damage to structures was observed as far as
Johannesburg and Pretoria. Newspapers reported that high-rise buildings in
Johannesburg swayed a few times. It was strange that this event caused so much
damage, because an ML 4.8 event occurred in this region in 1972, and a review of the
press coverage showed no reports of damage on that occasion. It would seem that the
attenuation for this ML 4.8 event was much more rapid than other mine-induced seismic
events.
Another notable event that caused significant damage to buildings and mine
infrastructure occurred in 1999 in the Free State district, known as the Matjabeng
earthquake. It was associated with the Dagbreek fault, which extends across kilometers of
active mining (Durrheim et al. 2007). Dor et al. (2001) observed total displacement of 44
cm on the Dagbreek fault. Dor et al. (2001) list large earthquakes in the Welkom region
since 1972. These earthquakes occurred on faults like the Erfdeel, President Brand, and
Saaiplaas, with ML in the range of 4.7?5.2 and observed displacements of 150?440 mm.
The largest earthquake that occurred in the mining areas was the Stillfontein ML 5.3 event
of 9 March 2005 (Saunders et al. 2007), which caused significant damage to surface
buildings and mine infrastructure. The cause of this earthquake was of national interest.
More information on the investigation of this event can be found in Durrheim et al. (2007).

Assessment of the Earthquake Catalog

We repeatedly encountered the following fundamental problems when looking at the data
provided by the SANSN for the purpose of better understanding the seismotectonic origins
of earthquakes in the region:
? The current number of stations and their configuration allows for a very limited
detection capability of the network. As a result the location of events can be poor,
and the ability to detect micro-earthquakes on active structures is rather limited.
? A comprehensive study is required to distinguish mining-related earthquakes from
earthquakes of natural tectonic origin in the database. Furthermore, analytical
techniques should be adopted in the SANSN to distinguish these events as they
are reported.
? An accurate assessment of the depth of earthquakes should be made.

1-8
? Focal mechanisms of earthquakes should be obtained.

A denser monitoring network is required to better understand earthquake occurrence. This
network should be concentrated in areas with active seismicity such as Ceres and
Koffiefontein.

Geological Provinces and Geophysical Observations
Topography and Major Geological Provinces

As the relief map shows (Figure 2), the South African interior is surrounded in the west,
south, and east by a cornice of mountains. This chain, consisting of many individual
mountain ranges, is known as the Great Escarpment. In the east, in the area of the
Drakensberg of KwaZulu-Natal and in the Kingdom of Lesotho, it reaches heights of
almost 4,000 m. In the south and west, the highest peaks reach 2,000 m. In front of the
escarpment is a mostly narrow coastal strip. Inland of the escarpment, the central high
plateau of South Africa reaches elevations of 1,000?1,700 m. The plateau slopes slowly
toward the Kalahari basin in the north. There is a relatively greater concentration of
seismicity along the Great Escarpment.
A brief geological description of the geologic units is provided below; refer to
Johnston et al. (2006) for more detail. A schematic representation of the geological
provinces superimposed on the seismicity recorded for the country is shown in Figure 3.
The Kaapvaal craton (KC) is of Archean age. It is the foundation upon which the
geological formations of South Africa have subsequently developed. A zone of
metamorphic sediments on the northern marginal zone of the KC separates the KC from
the Zimbabwe craton (ZC), which is of similar age and composition. It is thought that this
zone, referred to as the Limpopo belt, was formed as a result of a collision between the
KC and the ZC. The oblique nature of this collision is believed to have initiated or re-
activated major transcurrent fault systems, resulting in important structures such as the
Thabazimbi-Murchison lineament, which prepared the craton for the development (2,600?
2,100 million years ago) of the Transvaal and Griqualand West basins. The Bushveld
igneous complex intruded the KC at about 2,000 million years ago. Tectonic activity on the
KC ceased about 1,800 million years ago. Proterozoic fold and thrust belts up to 400 km

1-9
wide were added to the KC on the south (Namaqua-Natal mobile belt, NNMB) and west
(Gariep-Kaoko). A complex tectonic history 1,750?1,200 million years ago of rifting, basin
development, oceanic basin development, subduction, and plate collision is recorded in
the rocks of this period (Wilson 2005). The Pilanesberg alkaline complex and the Premier
diamond pipe intruded around 1,300 million years ago.
The rocks in the Cape fold belt (CFB) were laid down as sediments in a coastal
delta environment upon the Malmesbury unconformity in the Ordovician (450 million years
ago) period, with the folding subsequently occurring in the Carboniferous and Permian
periods during the merging of the supercontinent Pangaea.
The Karoo supergroup is the largest geological feature in southern Africa, covering
almost two thirds of the present land surface, including some parts of western and eastern
Cape provinces, almost all of Free State, western KwaZulu-Natal, much of southeast
Gauteng Province, Zambia, Zimbabwe, and Malawi. Its strata, mostly shales and
sandstones, record an almost continuous sequence of marine, glacial to terrestrial
deposition from the Late Carboniferous to the Early Jurassic, a period of about 100 million
years. Extensive basic and acid lavas of the Lebombo and Drakensberg groups cap the
Karoo supergroup, and their extrusion preceded the fragmentation of Gondwana.
South Africa began breaking away from Australia in the northeast around 200
million years ago, and this breakup proceeded southward and then westward until the
proto-Atlantic was formed about 120 million years ago. This was accompanied and
followed by widespread anorogenic alkaline magmatism of the kimberlitic, carbonatitic,
and ring-complex types (Wilson 2005).
Geologically younger deposits, ranging in age from Cretaceous to recent times,
include the Kalahari group sediments; coastal, shallow marine and lagoonal sediments;
and present and ancient river terraces (Schl?ter 2006).
Figure 3 shows that mining-related seismicity is prevalent in and around the
Witwatersrand basin. Given that the Karoo supergroup covers a large portion of the land
surface, it becomes crucial that a depth parameter be included when reporting on
earthquake occurrences. This would then provide a better understanding of whether the
earthquakes originated from the KC, the NNMB, or the Karoo. With the current monitoring
capabilities of the seismic network and the rather diffuse pattern of seismicity spanning the
country, it is rather difficult to ascertain any correlations between recorded seismicity and
the respective geological provinces.

1-10
Geophysical Observations and Subsurface Geology

Geophysical investigations (e.g., gravity, magnetics, and seismic anisotropy) provide a
better understanding of the subsurface geology of the Earth. Regional aeromagnetic and
gravimetric maps of South Africa are shown in Figures 4 and 5. The seismicity data and
boundary of major geological provinces are superimposed on these maps. Note that in
both these figures the patterns formed by the geophysical signatures correlate to a large
extent with the geological provinces beneath the Karoo sediments.
Gravity data provide information about densities of rocks underground. Generally,
gravity highs indicate the presence of relatively dense rocks, and magnetic anomalies are
caused by rocks with abundant magnetite in them. Very high-intensity anomalies (more
than 50 milligals or more than 200 gammas) typify major changes in rock type, usually (but
not always) in basement rocks. Most sedimentary rocks (with the exception of banded
ironstones) contain little magnetite, so generally we are dealing with igneous and
metamorphic rocks (Gibson 2007).
Weckmann et al. (2007) used magnetotelluric and seismic imaging to investigate
two major geophysical anomalies: the Beattie magnetic anomaly (BA) and the Southern
Cape conductive belt (SCCB), which occur in the NNMB and extend across the continent
from east to west (refer to Figure 4). The maximum of the Beattie anomaly coincides with
a narrow zone of high conductivity at 8?15 km depth and a zone of high seismic reflectivity
and high P-wave velocity. The results show that the crustal structure of the NNMB is
complex, consisting of conductive material in the upper to mid-crustal sections, with the
lower crust containing reflectors. The Cape Fold belt is characterized by low electrical
conductivities and high P-wave velocities. In contrast, the Mesozoic/Cenozoic Kango and
Oudtshoorn Basins of the Cape fold belt appear as regions of high electrical conductivity
and low P-wave velocities. The BA and SCCB show no distinct correlation with
earthquakes recorded in the region.
A magnetic lineament and a mafic pluton known as the Trompsburg complex (Mare
and Cole 2006) can be found 100 km north of the Koffiefontein cluster. No magnetic
anomalies exist in the vicinity of the Ceres cluster. Toward the west of the country, a
cluster of events occurs on a linear magnetic fabric on the NNMB. Similarly, this occurs on
the east in KwaZulu-Natal and in the north within a dyke swarm in the Limpopo Province.
Earthquakes in the Cape fold belt are diffuse and relatively low in number. No statement

1-11
can be made about the correlation between earthquakes and magnetic anomalies in
Lesotho because of the lack of magnetic data coverage in this area.
Figure 5 shows that gravity highs delineate the northern boundary of the NNMB,
the limbs of the BC, the Kheis graben in the west, and the Trompsburg complex near
Koffiefontein. A significant gravity low occurs in the CFB, corresponding with the trace of
the reactivated CKBF. No obvious correlations are visible between the seismicity patterns
and gravity anomalies. A moderate gravity high is visible within the Ceres cluster area that
is worth investigating in detail.
Other geophysical projects aimed to image the deeper earth structure beneath
South Africa include the Kaapvaal Project, which was conducted in the late 1990s.
Broadband seismic stations were deployed in an array that extended across South Africa.
Recordings of teleseismic earthquakes by this array provided an opportunity for scientists
to learn more about the deeper earth structures spanning the continent. Nguuri et al.
(2001) found that the Archean crust (KC, ZC) is typically thin (~35?40 km) in contrast to
the Proterozoic belts and post-Archean regions (e.g., Bushveld complex) where the crust
tends to be relatively thick (~ 45?50 km). A study of seismic wave anisotropy (Fouch et al.
2004), found clear evidence that mantle structures mimic the surface geology. A thick
mantle keel (or root) exists underneath the Archean cratons (at 250?300 km), while there
is no evidence for similar structures beneath the adjacent, younger Proterozoic mobile
belts. Reduced mantle seismic velocities are evident underneath the Bushveld Complex
(James et al. 2001; Fouch et al. 2004).
Current projects that are providing information on geophysical properties in the
lithosphere of southern Africa include the SAMTEX project (Hamilton et al. 2006) and the
AfricaArray project (Shen and Nyblade 2006). These observations can contribute toward
better modeling of the seismic structure and the characterization of large structural
seismotectonic domains. Only through a denser network of seismic monitoring stations
can one better correlate the geophysical investigations with the earthquake record.

1-12
Geodynamics: African Plate Motion and Regional Stress
Regimes

The motion of the African plate is coupled with intraplate motions. Zoback (1992) showed
that large regions in the interiors of plates are characterized by uniform compressive
stress orientations, which are produced by forces acting on the plate boundaries (e.g.,
ridge push). In these areas the maximum principal stresses are horizontal. This stress
regime is clearly seen in the southwestern part of South Africa, where an east-west
horizontal stress regime prevails, dominated by compressional forces originating in the
Mid-Atlantic Ridge. This regime is manifested in strike-slip faulting along major east-west-
trending fracture systems (e.g., Worcester and the KBF). A different stress regime prevails
in the East African rift system (EARS). Dominant buoyancy forces (swell-push) are
generated by the upwelling of the asthenospheric material and the thinning of the
lithosphere. The maximum principal stresses are in the vertical direction, resulting in
predominant normal faulting. In the area between the above two regions, there is an
intermediate zone of elevated topography, extending from eastern and southern Africa
into the surrounding oceans, known as the African superswell (Nyblade and Robinson
1994). Here the maximum principal stress remains vertical, and minor principal stress
assumes a NW-SE orientation. This feature has been named the ?Wegener stress
anomaly? (Andreoli et al. 1996; Bird et al. 2005). This anomaly is best explained in terms
of the combined effects of ridge-push and swell-push, with the latter dominant.
Quaternary Faults, Seismotectonic Faults, and Neotectonic
Activity

Return periods as long as 10,000 years are considered when assessing the risk posed by
large events to critical structures such as nuclear reactors. This is where the field of
neotectonics and paleoseismology becomes crucial in determining if and when a large
earthquake has occurred in the recent geological past, i.e., 500,000 years ago for
intraplate regions. McCalpin (1996) provides guidance on how to identify these features.
Many authors, in their quest to understand the relation between seismicity and tectonics,

1-13
have identified paleoseismic and neotectonic features in South Africa. We briefly review
these here, with key features shown in Figure 3.
? In the Kaapvaal craton, natural seismicity can be associated in the north with a
lineament defined by the Murchison greenstone Belt, the Zebediela fault, and the
Thabazimbi fault. The Thabazimbi and Zebediela faults are related to subsidence of
the Bushveld basin by as much as 400 m. The Zebediela fault is associated with a
number of thermal springs.
? Earthquakes occur seaward of the Great Escarpment, mainly on mountains of the
Cape fold belt.
? Earthquakes in the Koffiefontein cluster occur near the Lithani/Matigulu thrust in the
amphibolitic Mzumbe terrain. Andreoli et al. (1996) discovered a recent fault zone
reaching the surface 10 km southwest of Bultfontein. The linear feature appears as
a flat-bottomed furrow 30 cm deep and 0.5 m wide which could represent a belt of
ground depressed as a result of extensional faulting.
? Earthquake occurrence in the Lesotho cluster follows the Caledon River.
? Earthquakes in Lesotho occur near the Cedarville fault and Cedarville Flats alluvial
deposits that are located on the inland flank of the Ciskei/Swaziland axis of
upwarping. Thermal springs in KwaZulu-Natal and Mpumulanga also occur along
this axis (Kent 1981).

The Griqualand-Transvaal axis in the continental interior is related to the subsidence of
the Kalahari basin ( T Partridge, personal communication 2007). Small movements along
this axis led to disruption of drainage networks and development of new drainage lines.
The Saldanha-Agulhas axis of warping takes the form of a hinge line over a distance of
300 km in the southwestern Cape. Near Cape Agulhas, fluvial terraces, probably of
Neogene age, are uparched across this axis. Andreoli et al. (1996) state that neotectonic
joints, faults, and breccias cut consolidated and semiconsolidated Late Pliocene to
Pleistocene calcarenites near Gansbaai, Quoin Point, Cape Agulhas, and Gouriqua.
The CKBF in the Cape Province has reactivated fault scarps that are, in some
places, between 2?4 m high. The Worcester fault lies south of the CKBF, with similar
strike and orientation as the CKBF, and extends toward the Ceres cluster. Although there
are some correlations with seismicity along the Worcester fault, there is no recorded
evidence of reactivation similar to that found on the CKBF.

1-14
Andreoli et al. (1996) also pointed out widespread reactivation of Precambrian
faults from the Wesselsbron panneveld 60 km north of Bultfontein. Late Pleistocene to
Holocene faults are well-exposed at Port Durnford near Richards Bay, extending
northward through the St. Lucia lakes and the northern KwaZulu-Natal coastal plain into
southern Mozambique.
Near Johannesburg, the Rietfontein fault system runs from Edenvale in the east to
beyond Krugersdorp in the west. A series of landslides are found that could be related to
seismic events along the fault. Other evidence is becoming available that supports the
suggestion that this fault system may be the source of localized distress in buildings and
may also be the locus of low-level seismic events (Barker 2004).
Although evidence of neotectonic activity is clearly present, systematic recording
and mapping is required. The Quaternary sediments need to be mapped. If they are
absent, this should be indicated. Similarly, the evidence, or lack thereof, of paleoseismic
and neotectonic activity should be recorded spatially. Where there is evidence of such
activity, this needs to be investigated using various techniques now available (see
McCalpin 1996). A useful seismotectonic model requires: (i) the characteristics of the fault,
(ii) its earthquake history, and (iii) recurrence properties.

Conclusions

Building a seismotectonic model involves several geoscientific disciplines. Although we
have made progress in each discipline, we still need to extend and fine-tune our work to
build such a model. Milestones yet to be reached are:
Seismology:
? A denser network of seismic monitoring stations is required to improve the
sensitivity and location accuracy of recorded earthquakes.
? The earthquake database needs to be revisited to distinguish between earthquakes
of natural origin and those that are mining-related.
? Depths and focal mechanisms of earthquakes need to be determined and routinely
published.
? Microseismic monitoring needs to be undertaken of active regions such as the
Ceres and Koffiefontein areas and active fault regions in the CFB.

1-15
Geology:
? Quaternary sediments, especially those providing evidence of neotectonic and
paleoseismicity, need to be dated and mapped across the country.

In an intraplate region like South Africa, there may be little correlation between seismicity
and faults. Nevertheless, these knowledge gaps need to be addressed.

Acknowledgments

This work is funded by the Council for Geoscience (CGS) and the National Research
Foundation as part of the AfricaArray Project, which is hosted by the University of
Witwatersrand, South Africa, and Pennsylvania State University, U.S.A. We are grateful to
colleagues Mr. Dirk Grobbelaar for GIS assistance, Mr. Ian Saunders for GIS assistance
and useful insights into the earthquake database, and Professor Artur Cichowicz for
stimulating discussions and criticism. The CGS Geophysics Department provided the
geophysical information and the Seismology Unit provided the seismological information.

References
Andreoli, M. A. G., M. Doucoure, J. Van Bever Donker, D. Brandt, and N. J. B. Andersen
(1996). Neotectonics of southern Africa: A review. Africa Geoscience Review 3 (1),
1?16.
Barker, A. (2004). Proceedings of the Landslides and Seismicity Course, Geoscience
Africa, University of the Witwatersrand
Bird, P., Z. Ben-Avraham, G. Schubert, M. Andreoli, and G. Viola (2005). Patterns of
stress and strain rate in southern Africa. Journal of Geophysical Research 111, 1?
14.
Brandt, M. (2000). A review of the reservoir induced seismicity at the Katse Dam, Kingdom
of Lesotho, November 1995 to March 1999. Cant. Scient. diss., University of
Bergen, Norway.

1-16
Brandt, M. B. C., M. Bejaichund, E. M. Kgaswane, E. Hattingh, and D. L. R. Roblin (2005).
Seismic History of Southern Africa. Seismological Series of the Council for
Geoscience no. 37. Pretoria: Council for Geoscience.
Cornell, C. (1968). Engineering seismic risk analysis. Bulletin of the Seismological Society
of America 58, 1,583?1,606.
De Klerk, W. J., and J. Read (1988). An account of historical seismicity in southern Africa
with emphasis on the southern and eastern coastal belts. Grahamstown, Albany
Museum
Dor, O., Z. Reches, and G. Van Aswegen (2001). Fault zones of the Matjabeng
earthquake, 1999, South Africa. In Proceedings, 5th International Conference on
Rockbursts and Seismicity in Mines, ed. G. van Aswegen, R. J. Durrheim, and W.
D. Ortlepp, 109?112. Magaliesburg, South Africa. South African Institute of Mining
and Metallurgy.
Durrheim, R.J., R. L. Anderson, A. Cichowicz, R. Ebrahim-Trollope, G. Hubert, A. Kijko,
A. McGarr, W. D. Ortlepp and N. van Der Merwe. (2007). Investigation into the
Risks to Miners, Mines, and the Public Associated with Large Seismic Events in
Gold Mining Districts, vol. 2, Department of Minerals and Energy. Pretoria.
Fenton, C. H., and J. J. Bommer (2006). The M(w) 7 Machaze, Mozambique, earthquake
of 23 February 2006. Seismological Research Letters 77, 426?439.
Fernandez, L. M., and J. A. Guzman (1979). In Seismic History of Southern Africa, 1?38.
Geological Survey of South Africa Seismologic Series 9. Pretoria. Geological
Survey of South Africa.
Finnie GJ (1999). Using neural networks to discriminate between genuine and spurious
seismic events in mines. Pure and Appl. Geophysics, 154, 41-56
Fouch, M. J., D. E. James, J. C. VanDecar, S. Van Der Lee, and the Kaapvaal Seismic
Group (2004). Mantle seismic structure beneath the Kaapvaal and Zimbabwe
cratons. South African Journal of Geology 107, 33?44.
Gay, N. C., R. J. Durrheim, S. M. Spottiswoode, and A. J. van der Merwe (1995). Effect of
geology, in situ stress and mining methods on seismicity in southern African gold
and platinum mines. In Proceedings of the International Congress of Rock
Mechanics, vol. 8, ed. T. Fujii, 1,321?1,325. International Society for Rock
Mechanics
Gibson, D. (2007). Primer on Gravity and Magnetics;

1-17
http://www.gravmag.com/gmprimr.html.
Green, R. W. E., and S. Bloch (1971). The Ceres, South Africa, earthquake of September
29, 1969. I. Report on some aftershocks. Bulletin of the Seismological Society of
America 61 (4), 851?859.
Hamilton, M. P., A. G. Jones, R. L. Evans, S. F. Evans, C. J. S. Fourie, A. Mountford, X.
Garcia, J. E. Spratt, and the SAMTEX team (2006). Magnetotelluric studies across
the Kaapvaal craton in southern Africa. 49th Irish Geological Research Meeting,
University College Cork, Cork, Ireland, 24?26 February 2006.
James, D. E., M. J. Fouch, J. C. VanDecar, S. van der Lee, and the Kaapvaal Seismic
Group (2001). Tectospheric structure beneath southern Africa. Geophysical
Research Letters 28, 2,485?2,488.
Johnson, M. R., C. R. Anhaeusser, and R. J. Thomas, eds. (2006). The Geology of South
Africa, Pretoria: Geological Society of South Africa, Johannesburg/Council of
Geoscience.
Joubert, A., G. J. Beukes, J. N. J. Visser, and de H. Bruin (1991). A note on fluorite in Late
Pleistocene thermal spring deposits at Florisbad, Orange Free State. South African
Journal of Geology 94, 174?177.
Kent, L. E. (1981). The thermal springs of south-eastern Transvaal and northern Natal.
Annals of the Geological Survey of South Africa 15, 51?67
Kgaswane, E. M. (2002). The characterisation of South African seismicity April 1997?April
1999: Implications for refining models of sub-cratonic lithosphere. M.S. diss.,
University of the Witwatersrand, Johannesburg, 108 pps.
Krige, L. J. and F. A. Venter (1933). The Zululand earthquake of the 31st December 1932.
Transactions of the Geological Society of South Africa 36, 101?112.
Mare, L. P., and J. Cole (2006). The Trompsburg complex, South Africa: A preliminary
three dimensional model. Journal of African Earth Sciences 44, 314?330.
McCalpin, J., ed. (1996). Paleoseismology. San Diego: Academic Press, 588 pps.
McGarr, A. and Simpson D. (1997). A broad look at induced and triggered seismicity.
Rockburts and Seismicity in Mines-RaSiM, Balkema, Rotterdam.
Nguuri, T., J. Gore, D. E. James, S. J. Webb, C. Wright, T. G. Zengeni, O. Gwavava, J. A.
Snoke, and Kaapvaal Seismic Group (2001). Crustal structure beneath southern
Africa and its implications for the formation and evolution of the Kaapvaal and
Zimbabwe cratons. Geophysical Research Letters 28 (13), 2,501?2,504.

1-18
Nyblade, A.A., and S.W. Robinson (1994). The African Superswell, Geophys.Res.Lett. 21,
765-768.
Richardson and Jordan (2001). Some properties of gold-mine seismicity and implications
for tectonic earthquakes. Rockburts and Seismicity in Mines-RaSiM5, South African
Institute of Mining and Metallurgy, 2001.
Saunders, I., M. B. C. Brandt, and E. Kgaswane (2007). Preliminary appraisal of the effect
of a magnitude ML = 5.3 event of 9 March 2005, Klerksdorp, South Africa. Report
no. 2005-0066. Pretoria: Council for Geoscience.
Schl?ter, T. (2006) Geological Atlas of Africa, with Notes on Stratigraphy, Tectonics,
Economic Geology, Geohazards and Geosites of Each Country. Berlin: Springer,
272 pps.
Shen, Y., and A. Nyblade (2006) AfricaArray and other broadband seismic experiments in
Africa. AGU Western Pacific Geophysics Meeting, 24-27 July 2006. Beijing, China.
Singh, M., and T. Pule (2007). Maximum possible earthquakes in the Witwatersrand
Basin. In Proceedings of Conference: Seismology Workshop 2007: Neotectonics
and Mining Seismology?is a Relationship Possible? Council for Geoscience,
Pretoria, 8 May 2007.
U.S. Nuclear Regulatory Commission (2007). A Performance Based Approach to Define
Site Specific Ground Motion. Regulatory Guide 1.208, U. S. Nuclear Regulatory
Commission, March 2007.
Visser, J. N. J., and A. Joubert (1990). Possible earthquake-induced sediment liquefaction
in thermal spring deposits at Florisbad, Orange Free State. South African Journal of
Geology 93, 525?530.
Von Buchenr?der, W. L. (1830). An account of earthquakes which occurred at the Cape of
Good Hope during the month of December 1809 etc. South African Quarterly
Journal, Cape Town. 1, 18-25.
Weckmann, U., O. Ritter, A. Jung, T. Branch, and M. de Wit (2007). Magnetotelluric
measurements across the Beattie magnetic anomaly and the Southern Cape
conductive belt, South Africa. Journal of Geophysical Research 112, B05416;
doi:10.1029/2005JB003975.
Wilson, M. G. C. (2005). Simplified Geology, Selected Mines and Mineral Deposits-South
Africa, Lesotho and Swaziland. Council for Geoscience, Pretoria.
Zoback, M. L. (1992). First hand second-order patterns of stress in the lithosphere: The
World Stress Map Project. Journal of Geophysical Research 97, 11,703?11,728.

1-
19

Fi
gu
re

1
M
ap

Ea
rth
qu
ak
es

16
20
?
20
08

co
n
ta
in
ed

th
e

So
ut
h
Af
ric
an

Na
tio
n
al

Se
ism
ol
o
gi
ca
l D
at
a
ba
se

(S
AN
SD
).
Kn
ow
n

clu
st
e
rs

re
la
tin
g
to

n
at
u
ra
l
a
n
d
m
in
in
g-
in
du
ce
d
se
ism
ici
ty

ar
e

hi
gh
lig
ht
ed
.

Th
e
se
ism
ic

re
co
rd
in
g
st
at
io
n
s
ar
e
re
pr
es
en
te
d
by

tri
a
n
gl
es
.

1-
20

Fi
gu
re

2
To
po
gr
ap
hy

o
f S
ou
th

Af
ric
a
to
ge
th
er

w
ith

th
e
re
co
rd
ed

se
is
m
ici
ty

1-
21

Fi
gu
re

3
M
ajo
r
ge
ol
og
ic
al

pr
ov
in
ce
s
an
d
n
e
ot
ec
to
n
ic
fe
at
u
re
s/
fa
u
lts

So
u
th

Af
ric
a
su
pe
rim
po
se
d
w
ith

th
e
se
is
m
ici
ty

re
co
rd
ed

th
e
SA
NS
N
(ad
a
pt
e
d
fro
m

Ng
u
u
ri
et

al
.

20
01
)

1-
22

Fi
gu
re

4
R
eg
io
n
al

ae
ro
m
ag
n
et
ic
m
ap

So
u
th

Af
ric
a
su
pe
rim
po
se
d
on

th
e
m
ajo
r
ge
ol
o
gi
ca
l b
o
u
n
da
rie
s
a
n
d
th
e
se
ism
ici
ty

1-
23

Fi
gu
re

5
R
eg
io
n
al

gr
av
im
et
ric

m
ap

So
ut
h
Af
ric
a
su
pe
rim
po
se
d
o
n

th
e
m
ajo
r
ge
ol
o
gi
ca
l b
ou
n
da
rie
s
an
d
th
e
se
is
m
ici
ty

1-24
Table 1 List of Earthquakes in the national database above magnitude 5
Year Month Day Magnitude Region
1809 12 4 6.3 Cape Town Region
1811 6 2 5.7 Cape Town
1811 6 19 5 Cape Town
1850 5 21 5 Grahamstown
1857 8 14 5 Western Cape
1870 8 3 5 Harrismith
1899 9 13 5 Cape Town
1908 9 26 5 Bloemfontein
1910 10 21 5 Philipstown
1911 11 8 5 Windhoek
1912 2 20 6.2 Koffiefontein
1919 10 31 6.3 Swaziland
1921 10 9 5 Tulbagh
1922 6 23 5 Panbult Siding - Transvaal
1922 8 5 Panbult Siding-Tansvaal
1925 10 10 5 Leutwein Siding- Nambia
1932 8 9 5 Grahamstown
1932 12 31 6.3 Off Cape St. Lucia
1936 1 12 5 Mooihoek-Swaziland
1936 1 16 5 Fauresmith (Free State)
1940 11 10 5 Tzaneen (Transvaal)
1942 11 1 5.5 Port Shepstone
1950 9 14 6 Mozambique Channel
1950 9 30 5.5 Namaqualand
1952 1 27 5 Sutherland
1952 1 27 5.3 Sutherland

1-25
1952 1 28 5 Sutherland
1952 1 28 5.4 Sutherland
1952 6 9 5.5 Keetmanshoop District (Namibia)
1952 9 4 5 SWA (Namibia)
1952 11 8 5.2 SWA (Namibia)-Botswana Border
1953 5 1 5.8 Namaqualand
1954 2 17 5.5 Mozambique
1955 1 20 5.5 Offshore Mozambique
1955 5 20 5.1 Fauresmith District (Free State)
1957 4 13 5.5 Zastron District (Free State)
1963 8 27 5 Worcester-Ceres
1964 6 9 5 Luckhoff (Free State)
1966 6 18 5 Mokhotlong (Lesotho)
1968 1 12 5.5 Uitenhage
1968 1 14 5 Sul Do Save Prov (Mozambique)
1969 9 11 5.2 Heidelberg
1969 9 29 6.3 Tulbagh
1976 12 8 5.1 Welkom gold mines
1977 3 2 5.3 S.W. Cape Province
1977 4 7 5.2 Klerksdorp gold mines
1979 2 21 5.8 N. Cape Province SA.
1984 1 28 5.01 Klerksdorp gold mines
1985 5 8 5.22 Koffiefontein Region (Free State)
1986 10 5 5.15 Transkei
1987 9 30 5.04 Klerksdorp gold mines
1989 9 29 5 Mandileni Region (Transkei)
1991 10 31 5 Ceres Area Cape Province
1992 12 23 5.1 Namibia

1-26
1994 8 20 5 Southern Namibia
1994 10 30 5.1 Free State gold mines
1994 12 31 5.1 Brandvlei Region - Northern Cape
1996 9 15 5.1 Loeriefontein Region
1999 4 22 5.1 Free State gold mines
2001 4 6 5.2 Boesmanland Area - N. Cape
2001 7 31 5 Klerksdorp gold mines
2005 3 9 5.3 Klerksdorp gold mines
2005 10 12 5.1 Klerksdorp gold mines

Published in: Seismological Research Letters, 2009, Volume 80, Pages 65-33
Introduction

1-3
Earthquake Catalog

South African National Seismological Database

Seismicity Clusters

Largest Earthquakes

Mining-related Events

Assessment of the Earthquake Catalog

1-8
? Focal mechanisms of earthquakes should be obtained.

A denser monitoring network is required to better understand earthquake occurrence. This
network should be concentrated in areas with active seismicity such as Ceres and
Koffiefontein.

Geological Provinces and Geophysical Observations
Topography and Major Geological Provinces

1-10
Geophysical Observations and Subsurface Geology

1-12
Geodynamics: African Plate Motion and Regional Stress
Regimes

Conclusions

1-15
Geology:
? Quaternary sediments, especially those providing evidence of neotectonic and
paleoseismicity, need to be dated and mapped across the country.

In an intraplate region like South Africa, there may be little correlation between seismicity
and faults. Nevertheless, these knowledge gaps need to be addressed.

Acknowledgments

1-
19

Fi
gu
re

1
M
ap

Ea
rth
qu
ak
es

16
20
?
20
08

co
n
ta
in
ed

th
e

So
ut
h
Af
ric
an

Na
tio
n
al

Se
ism
ol
o
gi
ca
l D
at
a
ba
se

(S
AN
SD
).
Kn
ow
n

clu
st
e
rs

re
la
tin
g
to

n
at
u
ra
l
a
n
d
m
in
in
g-
in
du
ce
d
se
ism
ici
ty

ar
e

hi
gh
lig
ht
ed
.

Th
e
se
ism
ic

re
co
rd
in
g
st
at
io
n
s
ar
e
re
pr
es
en
te
d
by

tri
a
n
gl
es
.

1-
20

Fi
gu
re

2
To
po
gr
ap
hy

o
f S
ou
th

Af
ric
a
to
ge
th
er

w
ith

th
e
re
co
rd
ed

se
is
m
ici
ty

1-
21

Fi
gu
re

3
M
ajo
r
ge
ol
og
ic
al

pr
ov
in
ce
s
an
d
n
e
ot
ec
to
n
ic
fe
at
u
re
s/
fa
u
lts

So
u
th

Af
ric
a
su
pe
rim
po
se
d
w
ith

th
e
se
is
m
ici
ty

re
co
rd
ed

th
e
SA
NS
N
(ad
a
pt
e
d
fro
m

Ng
u
u
ri
et

al
.

20
01
)

1-
22

Fi
gu
re

4
R
eg
io
n
al

ae
ro
m
ag
n
et
ic
m
ap

So
u
th

Af
ric
a
su
pe
rim
po
se
d
on

th
e
m
ajo
r
ge
ol
o
gi
ca
l b
o
u
n
da
rie
s
a
n
d
th
e
se
ism
ici
ty

1-
23

Fi
gu
re

5
R
eg
io
n
al

gr
av
im
et
ric

m
ap

So
ut
h
Af
ric
a
su
pe
rim
po
se
d
o
n

th
e
m
ajo
r
ge
ol
o
gi
ca
l b
ou
n
da
rie
s
an
d
th
e
se
is
m
ici
ty