B.T. SHUMBA, D.J . HLATYWAYO AND V. MIDZI 381

IntroductionOn 15th March 2008, at 0221 Hrs (UTC), a smallmagnitude earthquake (mb = 4.3), occurred in theNyamandlovu area, approximately 135 km northwest ofBulawayo City. The epicenter of the event was 19.79°Sand 27.39°E (Figure 1). The parameters were calculatedat Goetz observatory, the national seismic data centre forZimbabwe. The earthquake was felt in the entire city ofBulawayo, which has a population of over 1 millionpeople. No injuries to the local population or destructionto structures and property were reported after the event.This is the largest earthquake on record from the area.The event was recorded as far as the Antarctica region.The Nyamandlovu area has been experiencing shakingfrom small to medium-size magnitude earthquakes (of undetermined mechanism) for some time now (Table 1). This study sought to establish the mechanismresponsible for the occurrence of these events and inferthe type of faulting in the area.

SeismotectonicsCompared to other seismically active parts of Zimbabwe,the seismicity of the Nyamandlovu area is low with fewmedium-size events that are generally believed to becaused by changes in pore pressure in the underlyingrock formations (Hlatywayo and Midzi, 2005). A smallfraction of Bulawayo city’s water supply is obtainedfrom underground water that is pumped from theNyamandlovu aquifer. Surface lineaments trend north

easterly (Hlatywayo, 1992). The geology of theNyamandlovu area comprises mostly Kalahari Aeoliansands with white and red sandstones overlain by basalticlavas approximately 700 m thick (Maufe, 1924; Vail,1967; Reeves and Hutchins, 1975). To the south-east ofthe area, towards Bulawayo city, basaltic Karoos andShamvanian metasediments dominate. Karoo sandstonescover the remainder of the area. The extreme east of thearea under consideration comprises complex gneissstructures that are part of a well known gold-bearingbelt. To the north of the area under study, the Karoosediments of the Zambezi Basin extend south-westwardsthrough the Hwange area into Botswana (Hlatywayo,1992). Seismic event data for the Nyamandlovu area were taken from the Bulawayo Seismological Bulletins for the period 1970 to 2008. Events in the area (Figure 2) appear randomly distributed. Events that havepreviously occurred in the area were poorly recorded,making it very difficult to carry out meaningful focalmechanism studies based on polarity data. One suchevent, magnitude, mb = 4.2, occurred on the 26th of June2004. It was felt widely over the whole of Bulawayo cityand the surrounding areas (Hlatywayo and Midzi, 2005).However, there was no reported damage to property orbuildings. In addition to the two events of 26 June 2004and 15 March 2008, three more events of magnitude 4.0 occurred in the area over the period 1970 to 2008, allof which were poorly recorded.

FOCAL MECHANISM SOLUTION OF THE 15TH MARCH 2008,NYAMANDLOVU EARTHQUAKE

B.T. SHUMBAGoetz Observatory, P.O.Box AC65, Ascot, Bulawayo, Zimbabwee-mail: shumbamajachani@yahoo.co.uk

D.J. HLATYWAYOPhysics Department, National University of Science and Technology, P.O.Box AC 939, Ascot,Bulawayo, Zimbabwee-mail: djhlatywayo@gmail.com

V. MIDZISeismology Unit, Council for Geoscience, P. Bag X112 Pretoria 0001 South Africae-mail: vmidzi@geoscience.org.za

© 2009 December Geological Society of South Africa

ABSTRACT

The focal mechanism solution of the 15th March 2008 earthquake (mb = 4.3) that occurred in the Nyamandlovu area, northwest of

Bulawayo City, Zimbabwe, has been determined from P-wave first motion polarities. Results show normal oblique left lateral

faulting. The earthquake mechanism bears a signature that is almost identical to those of events recorded in the Zambezi

Branch of the East African Rift Zone. Synthetic seismograms compared with observed data from regional stations were employed

for depth determination. This event had a shallow depth of 5 km. Earthquakes in the area tend to occur either after a

significant drought or a wet season of considerable length. The area is underlain by thick Karoo sandstones that form an

aquifer of high potential water storage. These events are therefore most probably induced by pore – pressure differentials in the

underlying rock.

SOUTH AFRICAN JOURNAL OF GEOLOGY, 2009, VOLUME 112 PAGES 381-386

doi:10.2113/gssajg.112.2-3.381

DataWave-form data of the 15 March, 2008 Nyamandlovuearthquake were recorded by seismic stations in SouthAfrica, Zimbabwe, Malawi, Botswana, Ivory Coast andNamibia (Figure 3). In an attempt to improve theazimuthal coverage for the data, additional waveformdata from seismic stations OPO (Madagascar), VNDA (Victoria lands) and CPUP (Paraguay) were used. To supplement the waveform data, P-wave firstmotion data recorded at other stations that could besourced from the International Seismological Centre(ISC) bulletins were incorporated. In total, data from 63 seismic stations around the world were used (Figure 3).

Hypocenter locationTo determine the hypocenter of the earthquake, weused the programme HYPOCENTER in SEISAN (Lienertet al., 1986; Lienert and Havskov, 1995). It can be usedwith all common crustal and global phases, and to locateteleseismic, regional and local events using the IASP91model as well as a local model (Midzi et al., 2010). P and S wave arrival times were picked using theSEISAN MULPLT program (Havskov and Ottemöller,2008). An epicenter located at 19.217S; 27.39E wasdetermined with error values in the latitude andlongitude of 0.2°. This compared to locations obtained by USGS and Council for Geoscience (CGS), (-19.217°S, 27.480°E) and (-19.115°S, 27.494°E)respectively. The USGS calculated a moment magnitudevalue of 4.7 compared to our local magnitude of 4.3. It must be stressed that the focal depth is difficult toconstrain given that we were only able to identify firstarrival P and S phases. In order to determine the focaldepth of the event, we computed three syntheticseismograms (Bouchon, 1981) at different depth values.The synthetic seismograms computed were thencompared with the observed seismograms to select thebest-fit seismograms. The method is based on a discrete wave number representation of the wave fields. The source is repeated periodically in space, so that integration over the k-domain is replaced by aseries.

Focal mechanismWe used programme FOCMEC (Snoke et al., 1984) in

SOUTH AFRICAN JOURNAL OF GEOLOGY

FOCAL MECHANISM SOLUTION OF THE 15TH MARCH 2008, NYAMANDLOVU EARTHQUAKE382

Figure 1. Map of Nyamandlovu area with major geological boundaries

Table 1. Earthquakes felt in the Nyamandlovu area from 1959 to

present (Data obtained from the Bulawayo Seismological centre

Date Coordinates Magnitude

Degrees Degrees

South East

1973-08-02 19.9 28.1 2.9

1975-11-01 20.4 28.5 3.0

1975-11-01 20.0 29.0 2.9

1976-04-03 19.4 27.2 3.8

1976-11-05 19.8 28.0 3.9

1978-03-06 19.3 27.7 3.5

1982-06-02 19.4 28.4 3.8

1984-12-26 20.3 29.0 3.3

2004-06-25 19.9 28.4 4.2

2008-03-15 19.7 27.3 4.3

B.T. SHUMBA, D.J . HLATYWAYO AND V. MIDZI

SOUTH AFRICAN JOURNAL OF GEOLOGY

383

SEISAN to determine the focal mechanism of the event.The programme performs an efficient systematic searchof the focal sphere and reports acceptable solutionsbased on selection criteria for the number of polarityuncertainties. The selection criteria for both polaritiesand angles allow correction or weightings for near-nodalsolutions. The velocity model by Midzi et al. (2010) wasused in determination of take off angles. The completedescription of how the programme works is found in themanual by Snoke et al. (1984). The data were weightedaccording to the nature of the polarity onset. That is ifthe P-wave polarity was clear, the weight of that polarity was given a unity value but if the onset was notvery clear, the weight was assigned to half. Data fromISC bulletins were given half weight since we cannotvouch for the quality of the actual picking of thepolarities.

ResultsThe focal mechanism solution of the earthquakeobtained from P-wave arrivals shows normal oblique leftlateral faulting (Figure 4). The first nodal plane’sparameters are: Strike = 285.7°, Dip = 27.99° and Rake =-43.22° while the second nodal plane has theparameters: strike = 55°, Dip = 48° and Rake = -95°(Figure 4). To determine the actual fault plane of the

event, we compared the solution to known existingfaults and lineaments in the area. The main lineamentsand faults are generally oriented in a north easterlydirection (Hlatywayo, 1992; 1995). The second nodalplane (strike = 55°, Dip = 48° and Rake = -95°) was thenselected as the likely solution from the comparisonresults.

The depth of the event was determined by com-paring synthetic waveforms to observed P waveformsfor three stations (Matopos (MATP, Lobatse (LBTB) andBoshof (BOSA). Waveforms from these three stationswere found to have best signal to noise ratios that wereeasier to fit. Figure 5 shows typical examples of thecomparison of the synthetic and observed seismogramsat different depths. All other parameters were fixed withonly depth varying. By comparing the synthetics atvarious depths with the observed waveforms, wededuced that the best-fit is obtained at around 5 km,which compares very well with the value of 5.8 kmdetermined by the USGS.

DiscussionThe mechanism of the 15th March 2008 event is normaloblique left lateral with one of the nodal planes strikingnortheast to southwest and the other striking northwestto southeast. Without using additional information such

Figure 2. Seismicity map of the Nyamandlovu area (Data obtained from the Zimbabwe Meteorological Services, Goetz Observatory).

as geology or aftershocks, it is difficult to determinewhich of the two possible solutions is the actual faultplane. An attempt to determine the actual fault plane isdone here by comparing the obtained solution withnotable major observed fault lineaments. Hlatywayo(1995; 1997) noted that the main lineaments in theHwange, Zambezi Basin to the north of our study area, trend in the north-easterly direction further north into the Luangwa river valley in eastern Zambia.Furthermore, Vail (1967) showed the trend of lineamentsfrom Botswana running northeasterly through Hwangeand Kariba gorge into Luangwa valley. Scholtz et al.(1976), found concentrations of epicenters ofearthquakes trending along a northeasterly directionfrom Botswana into the Deka-fault zone, Zambezi/Luangwa valleys. The focal mechanism solutionobtained is similar to other solutions which wereobtained for the Lake Kariba area which isapproximately 150 km north of Nyamandlovu (Sykes,1967; Gupta et al., 1972; Shudofsky, 1985; Hlatywayo,1995). Most of their results showed fault planes strikingto the northeast. Fairhead and Girdler (1971) noted thatthe Zambian earthquake which occurred on December 2, 1968 has a normal fault mechanism withthe faults striking north-easterly. Skyes (1967) alsoshowed that solutions for two events that occurred onLake Kariba were identical with both showing normal

SOUTH AFRICAN JOURNAL OF GEOLOGY

FOCAL MECHANISM SOLUTION OF THE 15TH MARCH 2008, NYAMANDLOVU EARTHQUAKE384

Figure 4. The focal mechanism solution of the 15th March 2008

event obtained global data using SEISAN algorithm (Havskov and

Ottemöller, 2008). The triangles shows seismic stations which has

negative onset polarity where as the circles represents seismic

stations with positive polarity

Figure 3. Red triangles represent global seismic stations whose data were used for determining the focal mechanism solution for the

2008, 15th March Nyamandlovu earthquake. The blue circle indicates the epicenter of the earthquake.

B.T. SHUMBA, D.J . HLATYWAYO AND V. MIDZI

SOUTH AFRICAN JOURNAL OF GEOLOGY

385

Figure 5. Comparison of the observed seismograms (red line) with the synthetic seismograms (blue line) using SEISAN algorithm

(Havskov and Ottemöller, 2008) at depths labeled next to each set of traces

faulting mechanisms. The results obtained in this studyare consistent with those obtained along the East AfricanRift, in particular, the mid-Zambezi rift. Hlatywayo(1995), Skyes (1967) and Scholz et al. (1976) confirmedthat fault plane solutions of earthquakes in the regionare associated with rifting and show normal faulting with tensile axis directed to the southeast. We can then attribute the seismodynamics of the events of the Nyamandlovu earthquakes to be consistent with themechanics of the rift systems.

There is a possibility of pore-pressure changes beingthe major trigger of earthquakes in this area (Hlatywayoand Midzi, 2005). The government of Zimbabwe hasdrilled many boreholes in this area to provide water toresidents in the nearby Bulawayo City. Hlatywayo andMidzi (2005) concluded that there is a possibility thatexcessive pore pressure built up by slow pore pressurechanges is the driving mechanism and causes instabilityin small, low-permeability creeping joints, resulting inseismic activity.

ConclusionThe scope of this paper attempts to address the focalmechanism of the 15th March 2008 Nyamandlovuearthquake and to determine the depth of this event.The results from this research are consistence with the known seismotectonics of the area. This study isnecessary in order to improve the understanding of theseismotectonics of the Nyamandlovu area and also to tryto understand the causes of the Nyamandlovu tremors.The results obtained in this study are very important infuture studies such as seismic hazard assessment.

AcknowledgementsThe analysis described here, was performed using datafrom stations of the Council of Geoscience in SouthAfrica, AfricaArray, International Seismological Centreand CTBTO. We thank these institutions for providing uswith this valuable data. We also wish to thank theanonymous reviewers for their constructive comments.

ReferencesBouchon., M. (1981). A simple method for calculating Green’s functions for

elastic layered media. Bulletin of the Seismological Society of America,

71, 959–972.

Fairhead, J.D. and Girdler R.W. (1971). The seismicity of Africa, Geophysical

Journal of the Royal Astronomical Society, 24, 271–301.

Havskov, J. and Ottemöller. L. (2008). Processing earthquake data. Book in

preparation, preliminary version at SEISAN web site fall 2008.

Hlatywayo, D.J. (1992). Seismicity of Zimbabwe during the period

1959–1990., Seismological Department, Uppsala, Sweden. Report 3–92,

98pp.

Hlatywayo, D.J. (1994). Fault plane solutions of the Deka Fault zone and

mid-Zambezi Valley, Geophysical Journal International, 120, 567–576.

Hlatywayo, D.J. (1997). Seismic hazard in central southern Arica, Geophysical

Journal Internaitonal, 130, 737–745.

Hlatywayo, D.J. and Midzi V. (2005). Report on the investigations into the

location, causes and the effects of the 25th June 2004 Bulawayo-

Nyamandlovu earthquake, Harare. Government of Zimbabwe, 13pp.

Lienert, B.R.E., Berg, E. and Frazer, L.N. (1986). Hypocenter: An earthquake

location method using centered, scaled, and adaptively least squares,

Bulletin of the Seismological Society of America, 76, 771–783.

Lienert, B.R.E. and Havskov, J. (1995). A computer program for locating

earthquakes both locally and globally, Seismological Research Letters,

66, 26–36.

Maufe H.B. (1924). An outline of the geology of Southern Rhodesia.

Southern Rhodesia Geological Survey, Short Report, 17pp.

Midzi V., Ayele A. and ESARSWG. (2010). Determination of velocity models

for the east and southern Africa region. Under review, Africa Geoscience

Review, 26pp.

Molnar, P. and Lyon-Caent, H. (1989). Fault plane solutions of earthquakes

and active tectonics of the Tibetan plateau and its margins. Geophysical

Journal International, 99, 123–155.

Reeves C.V. and Hutchins, D.G. (1975). Crustal structures in central southern

Arica. Nature, 254, 408–409.

Scholtz, C.H., Kocynski, T.A. and Hutchins, D.G. (1976). Evidence for

incipient rifting in Southern Africa. Geophysical Journal of the Royal

Astronomical Society, 44, 135–144.

Shudofsky N.G. (1985). Source mechanisms and focal depths of East African

earthquakes using Rayleigh-wave inversion and body-wave modeling,

Geophysical Journal of the Royal Astronomical Society. 83, 563–614.

Skyes L.R. (1970). Seismicity of the Indian Ocean and a possible nascent

Island arc between Ceylon and Australia, Journal of Geophysical Research,

75, 5041–5055.

Skyes L.R. (1967). Mechanism of earthquakes and nature of faulting on the

mid-ocean ridges, Journal of Geophysical Research, 72, 2131–2153.

Snoke, J.A., Munsey, J.W., Teague, A.G. and Bollinger, G. A. (1984).

A programme for focal mechanism determination by combined use of

polarity and SV-P amplitude ratio data, Earthquake Note, 55, 15pp.

Vail, J.R. (1967). The southern extension of the East Africa Rift system and

related igneous activity, Geologische Rundshau, 57, 601–614.

Editorial handling: R.J. Durrheim

SOUTH AFRICAN JOURNAL OF GEOLOGY

FOCAL MECHANISM SOLUTION OF THE 15TH MARCH 2008, NYAMANDLOVU EARTHQUAKE386