- Email: [email protected]
The seismicity in Kenya (East Africa) for the period 1906–2010: A review
Journal of African Earth Sciences 89 (2014) 72–78
Contents lists available at ScienceDirect
Journal of African Earth Sciences journal homepage: www.elsevier.com/locate/jafrearsci
The seismicity in Kenya (East Africa) for the period 1906–2010: A review J.K. Mulwa a,b,⇑, F. Kimata b, S. Suzuki c, Z.N. Kuria a a
University of Nairobi, Department of Geology, P.O. Box 30197-00100, Nairobi, Kenya Research Center for Seismology, Volcanology and Disaster Mitigation (RSVD), Graduate School of Science, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8601, Japan c Tono Geoscience Center, 959-31, Jorinji, Izumicho, Toki-shi, Gifu 509-5102, Japan b
a r t i c l e
i n f o
Article history: Received 15 February 2012 Received in revised form 18 September 2013 Accepted 18 October 2013 Available online 26 October 2013 Keywords: Review Seismicity Historical records Kenya
a b s t r a c t Kenya has had a seismic station since 1963 as part of the World Wide Standardized Seismograph Network (WWSSN). In 1990, the University of Nairobi in collaboration with GeoForschungsZentrum (GFZ) started to build up a local seismological network, the Kenya National Seismic Network (KNSN), which operated for about ten years between 1993–2002. This, however, experienced a myriad of problems ranging from equipment breakdown, vandalism and lack of spares. Kenya is seismically active since the Kenya rift valley traverses through the country from north to south bisecting the country into eastern and western regions. In the central part, the Kenya rift branches to form the NW-SE trending Kavirondo (Nyanza) rift. The Kenya rift valley and the Kavirondo (Nyanza) rift are the most seismically active where earthquakes of local magnitude (Ml) in the order of 62.0–5.0 occur. Furthermore, historical records show that earthquakes of magnitudes of the order of Ml P 6.0 have occurred in Kenya. Such large magnitude earthquakes include the January 6, 1928 Subukia earthquake (Ml 7.1) and an aftershock (Ml 6.2) four days later, as well as the 1913 Turkana region earthquake (Ml 6.2). Since early 1970’s, numerous seismic investigations have been undertaken in Kenya in order to understand the formation and structure of the Kenyan part of the East African rift valley. Earthquake data from these studies is, however, rather disorganized and individual datasets, including that acquired during the period 1993–2002, cannot furnish us with comprehensive information on the seismicity of Kenya for the past 100 years. The purpose of this paper is, therefore, to review the seismicity in Kenya for the period 1906–2010 by utilizing data and results from different sources. The general seismicity of Kenya has been evaluated using historical data, data recorded by local seismic networks, the United States Geological Survey catalogue as well as earthquake data from the numerous seismic investigations by different individuals and research groups. On the basis of earthquake data from these sources, the entire N–S trending Kenya rift valley and the NW-SE trending Nyanza (Kavirondo) rift are characterized by a high rate of seismicity, and the USGS network has been effective in detecting local M > 3.0 earthquakes. A peculiar trend is exhibited by earthquakes of Ml P 5.1 in that these occur along the N-S and NW-SE trending Kenya rift valley and the Kavirondo (Nyanza) rift zone respectively. Earthquake data from the various sources for the period 1906–2010 is complete for Ml P 4.4 earthquakes with a b-value of 0.79 which is characteristic of tectonic active regions like rifts. There is need to revive and extend the KNSN for a greater coverage and effective seismic monitoring in Kenya. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction Kenya is located on the eastern part of African continent and borders Ethiopia to the north, Republic of Somalia to the east and northeast, Indian Ocean to the southeast, Tanzania to the southwest, Uganda to the west and the Republic of Southern Sudan to the northwest (Fig. 1). Earthquake monitoring in Kenya dates from 1963 when the first seismic station was installed at the University of Nairobi’s Chiromo ⇑ Corresponding author at: University of Nairobi, Department of Geology, P.O. Box 30197-00100, Nairobi, Kenya. Tel.: +254 20 4445896; fax: +254 20 4449539. E-mail address: [email protected] (J.K. Mulwa). 1464-343X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jafrearsci.2013.10.008
campus at coordinate location 1°160 2200 S and 36°800 400 E by United States Geological Survey (USGS) as part of the World Wide Standardized Seismograph Network (WWSSN). The WWSSN seismic station at the University of Nairobi’s Chiromo campus was the first and major seismic station in Kenya. The seismic station consisted of a 3 component short period (SP) recording system (Benioff Seismometers with a period of 1 s) and a 3 component long period system (Sprengnether seismometers with a period of 15 s). The sensors were ‘‘founded’’ on volcanic tuffs which, in Nairobi region, constitute part of 122 m thickness of volcanics which unconformably overlie metamorphic rocks of the Mozambique Belt. This seismic station initially operated by Mines and Geology Department of the Ministry of Environment and Natural Resources was handed
J.K. Mulwa et al. / Journal of African Earth Sciences 89 (2014) 72–78
73
Fig. 1. Map of Kenya showing the NNE–SSW Kenya rift valley and NW–SE trending fault zones; MALRZ – Muglad Anza Lamu Rift Zone (Modified after Ashley et al., 2004).
over to Geology Department of the University of Nairobi in 1964. Since its installation in 1963, analogue data was recorded on photographic paper up to 1990. Between 1990 and 1995, analogue data from the station was recorded using a hot stylus system on heat sensitive paper. In September 1995 the WWSSN analogue station was upgraded to an IRIS digital seismic station with broadband seismometers and moved to Kilimambogo hill (coordinates 1.127°S and 37.252°E) about 70 km to the east of Nairobi city. For close to 30 years when it was first installed, this seismic station was the only permanent station in Kenya. Thousands of local and regional earthquakes were recorded by this station but they could not be properly located as no recordings from other stations were available for most of these events. In 1990, the University of Nairobi in collaboration with and funding from GeoForschungsZentrum (GFZ) started to build up a local seismological network, the Kenya National Seismic Network (KNSN) consisting of five short period seismic stations, which operated for about ten years between 1993 and 2002. This seismic network, however, experienced a myriad of problems ranging from equipment breakdown, vandalism and lack of spares especially after the funding organization (GFZ) pulled out.
Table 1 and Fig. 2 show the seismic stations distribution and the current setup of the Kenya National Seismic Network (KNSN). It is evident from Fig. 2 that the station configuration for the KNSN is rather skewed omitting much of the northeastern, central and western parts of the country. This therefore implies that the magnitude threshold of Ml 6 3.0 is only detectable at distances of up to 350 km from the center of the network. 2. Tectonic setting The East African Rift System (EARS), which is a classical example of an active intra-continental ridge system comprising an axial rift zone dissects Kenya longitudinally into almost two halves. The eastern arm of EARS (Figs. 1 and 2) extends from Afar Triangle in the north through Djibouti, Ethiopia, Kenya, and Tanzania to Mozambique in the south. The eastern arm of EARS is variously referred to as Eastern, Gregory (Gregory, 1921; Sikes, 1926; Saggerson, 1991) or the Kenya rift valley, and its central and southern parts form a 50 to 80 km wide rift valley with high escarpments on one side and en echelon fault steps on the opposite (Ibs-von Seht et al., 2008).
74
J.K. Mulwa et al. / Journal of African Earth Sciences 89 (2014) 72–78
Table 1 Localities and operation times of Kenya National Seismic Network stations. Name
Code
Duration of operation
Latitude
Longitude
Elevation (m)
Type of instrument
Kilimambogo Magadi Kibwezi Kibwezi (IOTEWS) Langata Nairobi Nairobi (Africa Array) Lodwar (IOTEWS)
KMBO MAG KIB KIBK LAN NAI NAI LODK
1995(09) – Current 1994(02) – 2001(12) 1994(04) – 2001(09) 2011 – Current 1994(02) – 1995(01) 1963–2002(12) 2010 – Current 2011 – Current
1.127°S 1.918°S 2.340°S 2.3591°S 1.377°S 1.274°S 1.274°S 3.4219°N
37.252°E 36.287°E 38.046° 38.0433°E 36.773°E 36.804°E 36.804°E 35.3616°E
1950 660 775 790 1707 1713 1713 665
STS-2 LE-3D LE-3D STS-2 LE-3D LE-3D Trillium compact STS-2
Fig. 2. The current Kenya National Seismic Network (KNSN) consisting of four broadband and one short period seismic stations.
Tectonics and volcanism since the Tertiary have concentrated on the region encompassing the present Kenya rift valley which is one of the better known parts of the EARS. The geologic evolution of the Kenya rift valley has been described by a number of authors (e.g. Baker, 1987; Baker and Wohlenberg, 1971; Baker et al., 1972, 1988; Logatchev et al., 1972; King and Chapman, 1972), and reviewed recently by Smith (1994) and Smith and Mosley (1993). Chorowicz (2005) describes in detail the overall morphology of the East African rift system and notes that it resembles that of mid-ocean ridges with the central rift valleys acting as depositional features mainly resulting from volcanic eruptions. The thicknesses of this volcanic material within the Kenya rift floor vary between 2 and 5 km and are characterized by low density of 2.3–2.4 g/cc (Mulwa et al., 2009, 2010; Mariita, 2003; Simiyu and Keller, 2001). The formation of the Kenya rift started about early Miocene in the north around Lake Turkana and migrated southwards being active from about middle to late Miocene in the central segment (Baker, 1986; Baker et al., 1972, 1988; Smith and Mosley, 1993). The formation of the rift started by up doming and volcanism on the crest of uplift and followed by faulting to form a half graben. The formation of a full graben occurred during the early Pleistocene. Basaltic and trachytic lavas were extruded, and the flows
are intercalated with tuffs (Gregory, 1921; Sikes, 1926; Saggerson, 1991). Subsequently, sheet trachytes were grid faulted with dominant north–south closely spaced faults. Baker (1987) identified three major stages in the tectonic development of the Kenya rift: (i) the pre-rift stage from 30 to 12 Mya with the formation of a depression and minor faulting, (ii) half-graben stage from12 to 4 Mya with the formation of the main boundary faults, and (iii) the graben stage (<4 Mya) with an increase and inward migration of faulting. All stages were accompanied by intense Quaternary volcanism mostly of alkaline nature. Apart from Ol Doinyo Lengai in northeastern Tanzania, all the rift volcanoes are classified as extinct. However, some postvolcanic hydrothermal activity is evident in a number of the rift volcanoes (Schlüter, 1997). Late Quaternary volcanic activity within the rift valley was concentrated in the axial region of the rift valley and resulted in formation of caldera volcanoes and volcanic cones. Fig. 3 shows some of the eastern rift volcanoes. There is strong evidence to suggest that the development of the Cenozoic Kenya rift and the present-day framework of fault lines were influenced by pre-existing basement structures (Braile et al., 1995). Particularly, the central and southern parts of the generally NNE trending Kenya rift developed inside a broad, NW
J.K. Mulwa et al. / Journal of African Earth Sciences 89 (2014) 72–78
Fig. 3. Volcanic centers aligned along the Kenya rift valley (red stars) with some off-rift volcanoes (blue triangles) after Mulwa, 2011. The red circles show areas characterised by occurrence of hot springs, geysers and deposition of silica sinter but where there has not been any volcanic eruption in the past. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
trending zone that is related to the suture between the Archean Tanzania Craton in the south-west and the Proterozoic Mozambique belt in the north-east. The craton margin zone is accompanied by several first-order NW trending and second order NNW-SSE trending lineaments, characterized by ductile and brittle shears and well documented in the Proterozoic rocks east and west of the rift. Smith and Mosley (1993) have interpreted the shear zones as major zones of weakness that were controlling rift propagation and accommodation of tectonic stress within several rift sectors. The course and faulting style of the rift often changes at the intersections with the shear zones. On a regional scale, the rift sections influenced by the reworked craton margin are located roughly between Lake Baringo in the north and Lake Magadi in the south. However, on a local scale, NW–SE and NNW–SSE trending fractures presumed to be under basement control are evident. Such zones have been described and referred to by Smith and Mosley (1993) as ‘accommodation zones’ and include the Kerio– Bogoria–Marmanet zone in the Lake Bogoria/Lake Baringo area and the Engorika–Magadi–Lembolos zone in Lake Magadi area. Baker et al. (1972) and Atmaoui and Hollnack (2003) have further discussed oblique shear zones crossing the Kenya rift.
75
trend of the East African rift system (Fig. 4). Usually small magnitude earthquakes are common on the eastern branch (Kenya rift) of EARS and these have been described as earthquake swarms by Ibsvon Seht et al. (2008), who have further given an overview of the occurrence of earthquake swarms in the Kenya rift. Historical earthquake records show that, in comparison to the whole of the EARS, the Kenya rift shows a relatively low seismic activity. Except for the January 6, 1928 earthquake (Ms 6.9) and aftershock (Ms 6.0) four days later in the central part of Kenya rift, as well as the 1913 Turkana region earthquake (Ms 6.0) in northern Kenya, no M > 5 earthquakes in Kenya have been reported since 1928. However, numerous local earthquake studies by Rykounov et al. (1972), Pointing et al. (1985), Maguire et al. (1986, 1988), Young et al. (1991), Tongue et al. (1992, 1994), Tongue (1992) and Ibs-von Seht et al. (2001) have established a high microearthquake activity in the Kenya rift. Ibs-von Seht et al. (2001) reported a constant rate of 10 M < 3 earthquakes per day for the southernmost part of the Kenya rift, close to the Kenya-Tanzania border. The background seismicity of the Kenya rift in this southern part was restricted to depths of 10 km whereas the earthquake swarms were generally shallow with most of the hypocenters between 0 and 6 km deep (Ibs-von Seht et al., 2001). For the central Kenya rift, Young et al. (1991) reported more than 500 M < 2.7 earthquakes during a three months period in Lake Bogoria region. Tongue et al. (1994) observed a short lasting (less than one day) M < 2 earthquake swarm in Lake Baringo region. The swarm earthquakes were found to form a narrow, elongated cluster in the center of the rift at 5 km depth. Young et al. (1991) and Tongue et al. (1994) attributed the earthquake swarm activity to emplacement of dikes. These dikes have been mapped using a fairly dense network of recent gravity data by Mulwa (2010), Mulwa et al. (2009). Some results of historical seismicity in central Kenya have been applied in the search for geothermal energy along the Kenya rift valley. The United States Geological Survey carried out seismic studies at Lake Bogoria and Olkaria areas in the central Kenya rift valley (Hamilton et al., 1973). Their results show that earthquakes of magnitude 2 or less that were located are restricted mainly within the fields along fault zones, and time distance plots indicated Lake Bogoria and Olkaria areas are underlain by a three layer volcanic sequence of about 3.5 km thick. This sequence is in turn underlain by a layer with a P-wave velocity of 6.3 km/s. Their
3. Seismicity 3.1. Observations of seismicity from 1906 to 1993 Since the late 1960s, a number of seismic investigations, including the operation of temporary seismic networks, have been carried out in Kenya by Wohlenberg (1968), Rykounov et al. (1972), Hamilton et al. (1973), Pointing et al. (1985), Maguire et al. (1986, 1988), Fairhead and Stuart (1982), Young et al. (1991), Tongue et al. (1992, 1994), Tongue (1992) and most recently by Ibs-von Seht et al. (2001). The results from these investigations document the presence of local centers of seismic activity. The results show that Kenya and the surrounding regions, especially northern Tanzania, are characterized by a moderate level of seismicity which is mainly controlled by the structural
Fig. 4. Seismicity in Kenya for the period 1906–2010 for local magnitudes 0.0 < Ml 6 7.1 (After Mulwa, 2011). All focal depths are within the crust (630 km).
76
J.K. Mulwa et al. / Journal of African Earth Sciences 89 (2014) 72–78
interpreted model shows a structure with velocities higher than the average upper crustal velocities within the rift. Further, fewer events were recorded in Lake Bogoria compared to Olkaria. 3.2. Observations of seismicity from 1993 to 2010 As discussed in the introductory section, since 1993 when a network of five seismic stations was installed, earthquake monitoring in Kenya has mainly been carried out by the University of Nairobi, Department of Geology. The management, data collection and analysis are carried out by the University of Nairobi’s seismological group which falls under Applied Geophysics and Seismology thematic area within the Department of Geology. The seismological group is also responsible for seismic data analysis, management and archiving. In 2010, a three component broadband Africa Array seismic station was installed in Chiromo campus (1.274°S, 36.804°E) of the University of Nairobi. Further between 2010 and early 2011, two broadband seismic stations were installed in Kibwezi and Lodwar as part of the Indian Ocean Tsunami Early Warning System (IOTEWS). These newly installed seismic stations together with the IRIS seismic station (KMBO) installed by United States Geological Survey (USGS) in September 1995 at Kilimambgo comprise the current Kenya National Seismic Network. For the purpose of this review, four earthquake datasets for the recent seismicity have been used: (i) data from the temporary seismic network (Magadi network) consisting of 15 stations installed and operated for eight months by Ibs-von Seht et al. (2001), (ii) the current Kenya National Seismic Network (KNSN), (iii) earthquake data from the catalogue of Hollnack and Stangl (1998), and (iv) USGS/NEIC earthquake catalogue. Arrival times of seismic phases from (i) have been relocated (Kuria et al. (2010), and from (ii) located using SEISAN program (Havskov, 1997) and the crustal velocity model derived from the results of the KRISP 1994 experiment (Prodehl et al., 1997). However, earthquake datasets from the Hollnack and Stangl (1998) and the USGS/NEIC catalogues, as well as the historical earthquake catalogue have been used as they are. The original earthquake data for the Magadi network operated from November 1997 and June 1998 as well as the Hollnack and Stangl (1998) catalogue were obtained in a CD from the Federal Institute for Geosciences and Natural Resources (BGR) courtesy of Ibs-von Seht. The review of the seismicity in Kenya from 1906 to 2010 (Fig. 4) shows that the entire N-S trending rift system, including the E–W trending Kavirondo (Nyanza) rift are characterized by a high rate of seismicity. Even though the occurrence of large earthquakes in Kenya is rather low, historical records based on macroseismic information, supplemented by re-examination of instrumental reports to re-evaluate the position and size of major earthquakes, as well as recent deployments of seismic monitoring stations and networks by the United States Geological Survey (USGS) since 1963 as part of the World Wide Standardized Seismic Network (WWSSN), University of Nairobi’s Department of Geology under the auspices of Kenya National Seismic Network (KNSN), and most recently in 2002 by the International Monitoring System (IMS) show a somewhat increase in the number of earthquakes in Kenya. However, this increase in the number earthquakes especially between the period 1990–2008 does not necessarily imply an increase in seismicity but rather is as a result of good detection capability following installation of the local Kenyan National Seismic Network (KNSN) since 1995. Fig. 4 shows that Kenya has experienced strong earthquakes of magnitude Ml P 6.0 between the period 1906–1930. The earthquake records for the period 1906–1989 have been obtained from published macroseismic information by Shah (1986), Ambraseys (1991), and Ambraseys and Adams (1991) while records for the period 1973–2010 have been obtained from the Kenyan National
Seismic Network and the USGS/National Earthquake Information Center (NEIC) earthquake catalogues. It is worthwhile to note that earthquake magnitudes in Fig. 4 are the local magnitudes. Consequently, magnitude homogenization for earthquakes from Ambraseys (1991), Ambraseys and Adams (1991) and the USGS/NEIC catalogues has been undertaken. However, owing to lack of a defined criteria for magnitude homogenization in Kenya, we adopted magnitude homogenization criteria by Hussein et al. (2008) to convert body wave magnitudes (mb) from the USGS/NEIC catalogue, and the Gutenberg and Richter (1956) criteria to convert surface wave magnitudes (Ms) from Ambraseys (1991) and Ambraseys and Adams (1991) catalogues, to local magnitude scales as shown below (Hussein et al., 2008; respectively Gutenberg and Richter, 1956):
Ml ðIPRGÞ ¼ ð0:69 0:07Þmb ðISCÞ þ 1:67 0:29 for 3:8 6 mb 6 6:1 ð1Þ Ms ¼ 1:27ðMl 1Þ 0:016M 2l
ð2Þ
Eq. (1) is applicable to a fair degree of accuracy since in Kenya, only earthquakes of body wave magnitudes 3.8 6 mb 6 6.1 have been reported in the USGS/NEIC catalogue. Figs. 5 and 6 show plots of the total number of earthquakes (n) per year and cumulative number of earthquakes (LOG N) against magnitude for the period 1906–2010. Using the Gutenberg and Richter (1944) relation,
LOGN ¼ a bM
ð3Þ
Fig. 5. Plot of number of earthquakes (n) and LOG N (log10) of cumulative number of earthquakes against magnitude for the Kenya earthquake catalogue for the period 1906–2010.
Fig. 6. Gutenberg and Richter (1944) relation of the Kenyan earthquake catalogue for the period 1906–2010.
J.K. Mulwa et al. / Journal of African Earth Sciences 89 (2014) 72–78
the unified earthquake catalogues for the period 1906–2010 can be considered rather complete for earthquakes having Ml P 4.4. The Gutenberg and Richter (1944) relation gives a b-value of 0.79 for Kenya (Fig. 6). Hollnack and Stangl (1998) obtained a b-value of 0.8 for the Kavirondo (Nyanza) rift and a catalogue completeness value of Ml > 3.2 for the southern part of the Kenya rift using only the Kenyan catalogue for a limited period between the years 1993–1996. The magnitude range for their catalogue was 2.0 6 Ml 6 5.0. Consequently, our values for the catalogue completeness (Ml P 4.4) and b-value (0.79) are representative for the whole country and do not significantly differ from their results, and the observed b-value is characteristic of tectonic active regions like rifts.
4. Discussion and conclusion A review of the seismicity in Kenya has been undertaken using earthquake data from different sources for the period 1906–2010 e.g. Kenya National Seismic Network from 1995 to 2005, Shah (1986) from 1906 to 1979, Ambraseys (1991), Ambraseys and Adams (1991) from 1906 to 1930, Hollnack and Stangl (1998) and USGS/NEIC from 1973 to 2010. However, the seismicity of Kenya as presented in Fig. 4 can somewhat be considered to give a biased picture of the earthquake activity in Kenya for the period 1906–2010. This is mainly because of several factors. First, the configuration of the Kenya National Seismic Network (KNSN) which was initially centered around the city of Nairobi in the central part of Kenya (Hollnack and Stangl, 1998). This is further compounded by the fact that the local seismic network time of observation was limited due to equipment breakdown, and therefore not all the stations were operating throughout the whole period (1995–2005). Furthermore, the magnitude threshold of M P 3.0 has been reported by USGS/NEIC since 2000. This means that no earthquakes of M < 3.0 are reported by the USGS/NEIC but this does not necessarily imply complete absence of microearthquakes (M 6 3.0) in Kenya. The station distribution of the Kenya National Seismic Network (1995–2005) therefore does seem to influence the pattern of microearthquake activity, which is concentrated in the areas of the recording stations (Hollnack and Stangl, 1998). For the larger earthquakes, the USGS network and the current Kenya National Seismic Network has been effective in detecting M > 3.0 earthquakes. Such earthquakes do show a general pattern in that most of them occur along the Kenyan rift valley as well as along the northeast–southwest trending Kavirondo (Nyanza) rift. Fig. 4 shows that the epicentral distribution of earthquakes in Lake Victoria region along the Kavirondo (Nyanza) rift is oriented in northeast-southwest direction. This was also noted by Hollnack and Stangl (1998), and led to the assumption that the seismic activity along the Kavirondo (Nyanza) Rift continues through the lake still orientated in northeast-southwest direction. The high seismicity along the Kenyan rift valley in southern Kenya-northern Tanzania close to the Kenya–Tanzania border near Ol Doinyo Lengai (Fig. 4) is attributed to the earthquake swarm activity which occurred in July 2007 just before Ol Doinyo Lengai erupted. From historical records, the strongest earthquakes in Kenya are the 1928 Subukia earthquake (Ml 7.1) and the aftershock (Ml 6.2) which occurred four days later, as well as the 1913 (Ml 6.2) Turkana region earthquake (Fig. 4). This earthquakes show a high influence on Fig. 5 (number of earthquakes against magnitude) and therefore make the curve skewed. The epicenters of these earthquakes are located in areas where the Kenya rift valley intersects the NW–SE trending Aswa-Nyangia fault zone and Anza rift zone respectively (Fig. 1). According to Mulwa (2011), the 1928 Subukia earthquake and May 20, 1990 southern Sudan earthquake
77
are aligned in NW–SE direction along the Aswa-Nyangia fault zone. Giardini and Beranzoli (1992) and Mulwa (2011) have determined that the focal mechanism for the May 20, 1990 southern Sudan earthquake is purely due to strike slip movement. This therefore raises the question whether the three strongest earthquakes in Kenya were due to normal faulting along the Kenya rift valley or due to strike slip movements along the Aswa-Nyangia fault zone (for the 1928 Subukia earthquake) and Anza rift zone (for the 1913 Turkana region earthquake). From the data presented in this paper, we conclude the following about the seismicity in Kenya: (i) The largest magnitude earthquakes in Kenya for the period 1906–2010 are the 1928 Subukia earthquake (Ml 7.1) and its aftershock (Ml 6.2) four days later in central Kenya, as well as the 1913 Turkana region earthquake (Ml 6.2). (ii) All earthquakes of Ml P 5.1 show a general pattern in that the most seismic activity is along the Kenya rift valley and the Kavirondo (Nyanza) rift zone. (iii) The epicentral distribution along the NE-SW trending Kavirondo (Nyanza) rift crosses Lake Victoria and this implies a southwest continuation of this seismically active rift zone. (iv) The Rift Valley shows more seismic activity in northern Tanzania than in Kenya and this is partly attributed to seismic swarm activity around Ol Doinyo Lengai volcano which erupted in July 2007. (v) Despite the bias by the configuration of the KNSN, there is generally high microearthquake activity in the Kenyan rift valley which warrants close monitoring. (vi) Contrary to Hollnack and Stangl (1998), it is debatable whether the main seismic energy release is due to strong earthquakes in the Kenyan Rift Valley or due to strike slip movement along the NW-SE trending fault zones such as Aswa-Nyangia fault zone and the Anza rift zone. (vii) There is need therefore to extend the KNSN for a greater coverage and effective seismic monitoring in Kenya.
Acknowledgements We would like to express our gratitude to Japan International Cooperation Agency (JICA) for granting the principal author a scholarship so as to pursue a nine months’ training on earthquake, tsunami and volcano eruption observation systems at the Research Center for Seismology, Volcanology and Disaster mitigation, Nagoya University in Japan. We are also indebted to thank the principal authors’ employer (University of Nairobi) and the Kenyan Ministry of State for Public Service for granting him study leave. Finally, we wish to express our sincere gratitude to Dr. Malte Ibs-von Seht of the Federal Institute for Geosciences and Natural Resources in Germany for availing to us the earthquake catalogue for the period 1997–2001. References Ambraseys, N.N., 1991. Earthquake hazard in the Kenya Rift: the Subukia earthquake 1928. Geophys. J. Int. 105, 253–269. Ambraseys, N.N., Adams, R.D., 1991. Reappraisal of major african earthquakes, south of 20 N, 1900–1930. Nat. Hazards 4, 389–419. Ashley, G.M., Mworia, J.M., Muasya, A.M., Owen, R.B., Driese, S.G., Hover, V.C., Renaut, R.W., Goman, M.F., Mathai, S., Blatt, S.H., 2004. Sedimentation and recent history of a freshwater wetland in a semi-arid environment: Loboi swamp, Kenya, East Africa. Sedimentology 51, 1–21. Atmaoui, N., Hollnack, D., 2003. Neotectonics and extension direction of the Southern Kenya rift, Lake Magadi. Tectonophysics 364, 71–83. Baker, B.H., 1986. Tectonics and volcanism of the southern Kenya rift valley and its influence on sedimentation in the African rifts. In: Frostick, L.E., Renaut, R.W., Reid, I., Tiercelin, J.J. (Eds.), Geol. Soc. Lond. Spec. Publ. 25, pp. 45–57.
78
J.K. Mulwa et al. / Journal of African Earth Sciences 89 (2014) 72–78
Baker, B.H., 1987. Outline of the petrology of the Kenya rift alkaline province. Geol. Soc. Lond. Spec. Publ. 30, 293–311. Baker, B.H., Wohlenberg, J., 1971. Structure and evolution of the Kenya rift valley. Nature 229, 538–542. Baker, B.H., Mohr, P.A., Williams, L.A.J., 1972. Geology of the eastern rift system of Africa. Geol. Soc. Am. Spec. Pap. 136, 1–67. Baker, B.H., Mitchell, J.G., Williams, L.A.J., 1988. Stratigraphy, geochronology and volcano-tectonic evolution of the Kedong–Naivasha–Kinangop region, Gregory rift valley, Kenya. J. Geol. Soc. Lond. 145, 107–116. Braile, L.W., Keller, G.R., Wendland, R.F., Morgan, P., Khan, M.A., 1995. The East African rift system. In: Olsen, K.H. (Ed.), Continental Rifts: Evolution, Structure, and Tectonics. Developments in Geotectonics, vol. 25, pp. 213–231. Chorowicz, J., 2005. The East African rift system. J. Afr. Earth Sci. 43, 379–410. Fairhead, J., Stuart, G., 1982. The seismicity of the EARS and comparison with other continental rift. In: Palmason, G. (Ed), Geodynamics Series AGU, vol. 8. Washington, DC, pp. 41–61. Giardini, D., Beranzoli, L., 1992. Waveform modeling of the May 20, 1990 Sudan earthquake. Tectonophysics 209, 105–114. Gregory, J.W., 1921. The Rift Valleys and Geology of East Africa. Seeley Service, London. Gutenberg, B., Richter, C.F., 1944. Frequency of earthquakes in California. Bull. Seismol. Soc. Am. 34, 185–188. Gutenberg, B., Richter, C.F., 1956. Magnitude and energy of earthquakes. Ann. Geofis. 9, 1–15. Hamilton, R.M., Smith, B.E., Knapp, F., 1973. Earthquakes in geothermal areas near Lake Naivasha and Lake Hannington. Unpubl. Rept. by United States geological survey for the UNDP/EAPL, p. 37. Havskov, J., 1997. The SEISAN Earthquake Analysis Software for IBM PC and SUN. Ver. 6.0., Institute of Solid Earth Physics, University of Bergen, Norway. Hollnack, D., Stangl, R., 1998. The seismicity related to the southern part of the Kenya rift. J. Afr. Earth Sci. 26, 477–495. Hussein, H.M., Abou Elenean, K.M., Marzouk, I.A., Peresan, A., Korrat, I.M., Abu ElNader, E., Panza, G.F., El-Gabry, M.N., 2008. Integration and magnitude homogenization of the Egyptian earthquake catalogue. Nat. Hazards 47, 525– 546. Ibs-von Seht, M., Blumenstein, S., Wagner, R., Hollnack, D., Wohlenberg, J., 2001. Seismicity, seismotectonics and crustal structure of the southern Kenya rift – new data from the Lake Magadi area. Geophys. J. Int. 146, 439–453. Ibs-von Seht, M., Plenefisch, T., Klinge, K., 2008. Earthquake swarms in continental rifts – a comparison of selected cases in America, Africa and Europe. Tectonophysics 452, 66–77. King, B.C., Chapman, G.R., 1972. Volcanism of the Kenya Rift Valley. Phil. Trans. R. Soc. A271, 185–208. Kuria, Z.N., Woldai, T., van der Meer, F.D., Barongo, J.O., 2010. Active fault segments as potential earthquake sources: Inferences from integrated geophysical mapping of the Magadi fault system, southern Kenya Rift. J. Afr. Earth. Sci. 57 (4), 345–359. Logatchev, N.A., Beloussov, V.V., Milanovsky, E.E., 1972. East African Rift development. Tectonophyiscs 15, 71–81. Maguire, P.K., Cooke, P.A.V., Laffoley, N.d’.A., Evans, J.R., 1986. Earthquake recording in the central part of the Kenya Rift Valley. EOS 67, 1103.
Maguire, P.K., Shah, E.R., Pointing, A.J., Cooke, P.A., Khan, M.A., Swain, C.J., 1988. The seismicity of Kenya. J. Afr. Earth Sci. 7, 915–923. Mariita, N.O., 2003. An Integrated Geophysical Study of the Northern Kenya Rift Crustal Structure: Implications for Geothermal Energy Prospecting for Menengai Area. Ph.D. Dissertation, University of Texas at El Paso, USA. Mulwa, J.K., 2010. Integrated geophysical study of Lake Bogoria basin, Kenya: Implications for geothermal energy prospecting. Ph.D. dissertation, University of Nairobi, Kenya. Mulwa, J.K., 2011. Determination of source parameters for the May 20, 1990 southern Sudan earthquake by inversion of teleseismic body-waves. In: JICA Training Program, Operating management of Earthquake, Tsunami and Volcano Eruption observation system, Fiscal year 2010. Research Center for Seismology Volcanology and Disaster Mitigation, Japan, pp. 15–31. Mulwa, J.K., Fairhead, D., Barongo, J.O., Mariita, N.O., 2009. Heat source mapping and evaluation of geothermal resource potential in Lake Bogoria basin, Kenya. SEG Expand. Abstr. 28, 1294–1299. Pointing, A.J., Maguire, P.K.H., Khan, M.A., Francis, D.J., Swain, C.J., Shah, E.R., Griffiths, D.H., 1985. Seismicity of the northern part of the Kenya rift valley. J. Geodyn. 3, 23–37. Prodehl, C., Ritter, J., Mechie, J., Keller, G.R., Khan, M.A., Fuchs, K., Nyambok, I.O., Obel, J.D., 1997. The KRISP 94 lithospheric investigation of southern Kenya-the experiments and their main results. In: Fuchs, K., Altherr, R., Muller, B., Prodehl, C. (Eds.), Stress and Stress Release in the Lithosphere, Tectonophysics, vol. 278, pp. 121–148. Rykounov, L.N., Sedov, V.V., Savrina, L.A., Bourmin, V.J., 1972. Study of microearthquakes in the rift zones of East Africa. Tectonophysics 25, 123–130. Saggerson, E.P., 1991. Geology of the Nairobi area. Geol. Surv. Kenya, Rep. No. 98. Schlüter, T., 1997. Geology of East Africa. Beiträge zur Regionalen Geologie der Erde, vol. 27. Gebrüder Bornträger, Berlin-Stuttgart, 484pp. Shah, E., 1986. Seismicity of Kenya. Ph.D. Thesis, University of Nairobi, Kenya. Sikes, H.L., 1926. The structure of the eastern flank of the Rift valley near Nairobi. Geogrl. J. 68. Simiyu, S.M., Keller, G.R., 2001. An integrated geophysical analysis of the upper crust of the southern Kenya rift. Geophys. J. Int. 147, 543–561. Smith, M., 1994. Stratigraphic and structural constraints on mechanisms of active rifting in the Gregory Rift, Kenya. In: Prodehl, C., Keller, G.R., Khan, M.A. (Eds.), Crustal and Upper Mantle Structure of the Kenya Rift. Tectonophysics, vol. 236, pp. 3–22. Smith, M., Mosley, P.N., 1993. Crustal heterogeneity and basement influence on the development of the Kenya Rift, East Africa. Tectonics 12, 591–606. Tongue, J.A., 1992. Tomographic study of local earthquake data from the Lake Bogoria region of the Kenya Rift Valley. Geophys. J. Int. 109, 249–258. Tongue, J.A., Maguire, P.K.H., Young, P.A.V., 1992. Seismicity distribution from temporary earthquake recording networks in Kenya. Tectonophysics 204, 71–79. Tongue, J., Maguire, P., Burton, P., 1994. An earthquake study in the Lake Baringo basin of the central Kenya Rift. Tectonophysics 236, 151–164. Wohlenberg, J., 1968. Remarks on the seismicity of East Africa between 4°N–12°S and 23°–40°E. Tectonophysics 8, 567–577. Young, P.A.V., Maguire, P.K.H., Laffoley, d.’.A., Evans, J.R., 1991. Implications of the distribution of seismicity near Lake Bogoria in the Kenya Rift. Geophys. J. Int. 105, 665–674.
Contents lists available at ScienceDirect
Journal of African Earth Sciences journal homepage: www.elsevier.com/locate/jafrearsci
The seismicity in Kenya (East Africa) for the period 1906–2010: A review J.K. Mulwa a,b,⇑, F. Kimata b, S. Suzuki c, Z.N. Kuria a a
University of Nairobi, Department of Geology, P.O. Box 30197-00100, Nairobi, Kenya Research Center for Seismology, Volcanology and Disaster Mitigation (RSVD), Graduate School of Science, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8601, Japan c Tono Geoscience Center, 959-31, Jorinji, Izumicho, Toki-shi, Gifu 509-5102, Japan b
a r t i c l e
i n f o
Article history: Received 15 February 2012 Received in revised form 18 September 2013 Accepted 18 October 2013 Available online 26 October 2013 Keywords: Review Seismicity Historical records Kenya
a b s t r a c t Kenya has had a seismic station since 1963 as part of the World Wide Standardized Seismograph Network (WWSSN). In 1990, the University of Nairobi in collaboration with GeoForschungsZentrum (GFZ) started to build up a local seismological network, the Kenya National Seismic Network (KNSN), which operated for about ten years between 1993–2002. This, however, experienced a myriad of problems ranging from equipment breakdown, vandalism and lack of spares. Kenya is seismically active since the Kenya rift valley traverses through the country from north to south bisecting the country into eastern and western regions. In the central part, the Kenya rift branches to form the NW-SE trending Kavirondo (Nyanza) rift. The Kenya rift valley and the Kavirondo (Nyanza) rift are the most seismically active where earthquakes of local magnitude (Ml) in the order of 62.0–5.0 occur. Furthermore, historical records show that earthquakes of magnitudes of the order of Ml P 6.0 have occurred in Kenya. Such large magnitude earthquakes include the January 6, 1928 Subukia earthquake (Ml 7.1) and an aftershock (Ml 6.2) four days later, as well as the 1913 Turkana region earthquake (Ml 6.2). Since early 1970’s, numerous seismic investigations have been undertaken in Kenya in order to understand the formation and structure of the Kenyan part of the East African rift valley. Earthquake data from these studies is, however, rather disorganized and individual datasets, including that acquired during the period 1993–2002, cannot furnish us with comprehensive information on the seismicity of Kenya for the past 100 years. The purpose of this paper is, therefore, to review the seismicity in Kenya for the period 1906–2010 by utilizing data and results from different sources. The general seismicity of Kenya has been evaluated using historical data, data recorded by local seismic networks, the United States Geological Survey catalogue as well as earthquake data from the numerous seismic investigations by different individuals and research groups. On the basis of earthquake data from these sources, the entire N–S trending Kenya rift valley and the NW-SE trending Nyanza (Kavirondo) rift are characterized by a high rate of seismicity, and the USGS network has been effective in detecting local M > 3.0 earthquakes. A peculiar trend is exhibited by earthquakes of Ml P 5.1 in that these occur along the N-S and NW-SE trending Kenya rift valley and the Kavirondo (Nyanza) rift zone respectively. Earthquake data from the various sources for the period 1906–2010 is complete for Ml P 4.4 earthquakes with a b-value of 0.79 which is characteristic of tectonic active regions like rifts. There is need to revive and extend the KNSN for a greater coverage and effective seismic monitoring in Kenya. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction Kenya is located on the eastern part of African continent and borders Ethiopia to the north, Republic of Somalia to the east and northeast, Indian Ocean to the southeast, Tanzania to the southwest, Uganda to the west and the Republic of Southern Sudan to the northwest (Fig. 1). Earthquake monitoring in Kenya dates from 1963 when the first seismic station was installed at the University of Nairobi’s Chiromo ⇑ Corresponding author at: University of Nairobi, Department of Geology, P.O. Box 30197-00100, Nairobi, Kenya. Tel.: +254 20 4445896; fax: +254 20 4449539. E-mail address: [email protected] (J.K. Mulwa). 1464-343X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jafrearsci.2013.10.008
campus at coordinate location 1°160 2200 S and 36°800 400 E by United States Geological Survey (USGS) as part of the World Wide Standardized Seismograph Network (WWSSN). The WWSSN seismic station at the University of Nairobi’s Chiromo campus was the first and major seismic station in Kenya. The seismic station consisted of a 3 component short period (SP) recording system (Benioff Seismometers with a period of 1 s) and a 3 component long period system (Sprengnether seismometers with a period of 15 s). The sensors were ‘‘founded’’ on volcanic tuffs which, in Nairobi region, constitute part of 122 m thickness of volcanics which unconformably overlie metamorphic rocks of the Mozambique Belt. This seismic station initially operated by Mines and Geology Department of the Ministry of Environment and Natural Resources was handed
J.K. Mulwa et al. / Journal of African Earth Sciences 89 (2014) 72–78
73
Fig. 1. Map of Kenya showing the NNE–SSW Kenya rift valley and NW–SE trending fault zones; MALRZ – Muglad Anza Lamu Rift Zone (Modified after Ashley et al., 2004).
over to Geology Department of the University of Nairobi in 1964. Since its installation in 1963, analogue data was recorded on photographic paper up to 1990. Between 1990 and 1995, analogue data from the station was recorded using a hot stylus system on heat sensitive paper. In September 1995 the WWSSN analogue station was upgraded to an IRIS digital seismic station with broadband seismometers and moved to Kilimambogo hill (coordinates 1.127°S and 37.252°E) about 70 km to the east of Nairobi city. For close to 30 years when it was first installed, this seismic station was the only permanent station in Kenya. Thousands of local and regional earthquakes were recorded by this station but they could not be properly located as no recordings from other stations were available for most of these events. In 1990, the University of Nairobi in collaboration with and funding from GeoForschungsZentrum (GFZ) started to build up a local seismological network, the Kenya National Seismic Network (KNSN) consisting of five short period seismic stations, which operated for about ten years between 1993 and 2002. This seismic network, however, experienced a myriad of problems ranging from equipment breakdown, vandalism and lack of spares especially after the funding organization (GFZ) pulled out.
Table 1 and Fig. 2 show the seismic stations distribution and the current setup of the Kenya National Seismic Network (KNSN). It is evident from Fig. 2 that the station configuration for the KNSN is rather skewed omitting much of the northeastern, central and western parts of the country. This therefore implies that the magnitude threshold of Ml 6 3.0 is only detectable at distances of up to 350 km from the center of the network. 2. Tectonic setting The East African Rift System (EARS), which is a classical example of an active intra-continental ridge system comprising an axial rift zone dissects Kenya longitudinally into almost two halves. The eastern arm of EARS (Figs. 1 and 2) extends from Afar Triangle in the north through Djibouti, Ethiopia, Kenya, and Tanzania to Mozambique in the south. The eastern arm of EARS is variously referred to as Eastern, Gregory (Gregory, 1921; Sikes, 1926; Saggerson, 1991) or the Kenya rift valley, and its central and southern parts form a 50 to 80 km wide rift valley with high escarpments on one side and en echelon fault steps on the opposite (Ibs-von Seht et al., 2008).
74
J.K. Mulwa et al. / Journal of African Earth Sciences 89 (2014) 72–78
Table 1 Localities and operation times of Kenya National Seismic Network stations. Name
Code
Duration of operation
Latitude
Longitude
Elevation (m)
Type of instrument
Kilimambogo Magadi Kibwezi Kibwezi (IOTEWS) Langata Nairobi Nairobi (Africa Array) Lodwar (IOTEWS)
KMBO MAG KIB KIBK LAN NAI NAI LODK
1995(09) – Current 1994(02) – 2001(12) 1994(04) – 2001(09) 2011 – Current 1994(02) – 1995(01) 1963–2002(12) 2010 – Current 2011 – Current
1.127°S 1.918°S 2.340°S 2.3591°S 1.377°S 1.274°S 1.274°S 3.4219°N
37.252°E 36.287°E 38.046° 38.0433°E 36.773°E 36.804°E 36.804°E 35.3616°E
1950 660 775 790 1707 1713 1713 665
STS-2 LE-3D LE-3D STS-2 LE-3D LE-3D Trillium compact STS-2
Fig. 2. The current Kenya National Seismic Network (KNSN) consisting of four broadband and one short period seismic stations.
Tectonics and volcanism since the Tertiary have concentrated on the region encompassing the present Kenya rift valley which is one of the better known parts of the EARS. The geologic evolution of the Kenya rift valley has been described by a number of authors (e.g. Baker, 1987; Baker and Wohlenberg, 1971; Baker et al., 1972, 1988; Logatchev et al., 1972; King and Chapman, 1972), and reviewed recently by Smith (1994) and Smith and Mosley (1993). Chorowicz (2005) describes in detail the overall morphology of the East African rift system and notes that it resembles that of mid-ocean ridges with the central rift valleys acting as depositional features mainly resulting from volcanic eruptions. The thicknesses of this volcanic material within the Kenya rift floor vary between 2 and 5 km and are characterized by low density of 2.3–2.4 g/cc (Mulwa et al., 2009, 2010; Mariita, 2003; Simiyu and Keller, 2001). The formation of the Kenya rift started about early Miocene in the north around Lake Turkana and migrated southwards being active from about middle to late Miocene in the central segment (Baker, 1986; Baker et al., 1972, 1988; Smith and Mosley, 1993). The formation of the rift started by up doming and volcanism on the crest of uplift and followed by faulting to form a half graben. The formation of a full graben occurred during the early Pleistocene. Basaltic and trachytic lavas were extruded, and the flows
are intercalated with tuffs (Gregory, 1921; Sikes, 1926; Saggerson, 1991). Subsequently, sheet trachytes were grid faulted with dominant north–south closely spaced faults. Baker (1987) identified three major stages in the tectonic development of the Kenya rift: (i) the pre-rift stage from 30 to 12 Mya with the formation of a depression and minor faulting, (ii) half-graben stage from12 to 4 Mya with the formation of the main boundary faults, and (iii) the graben stage (<4 Mya) with an increase and inward migration of faulting. All stages were accompanied by intense Quaternary volcanism mostly of alkaline nature. Apart from Ol Doinyo Lengai in northeastern Tanzania, all the rift volcanoes are classified as extinct. However, some postvolcanic hydrothermal activity is evident in a number of the rift volcanoes (Schlüter, 1997). Late Quaternary volcanic activity within the rift valley was concentrated in the axial region of the rift valley and resulted in formation of caldera volcanoes and volcanic cones. Fig. 3 shows some of the eastern rift volcanoes. There is strong evidence to suggest that the development of the Cenozoic Kenya rift and the present-day framework of fault lines were influenced by pre-existing basement structures (Braile et al., 1995). Particularly, the central and southern parts of the generally NNE trending Kenya rift developed inside a broad, NW
J.K. Mulwa et al. / Journal of African Earth Sciences 89 (2014) 72–78
Fig. 3. Volcanic centers aligned along the Kenya rift valley (red stars) with some off-rift volcanoes (blue triangles) after Mulwa, 2011. The red circles show areas characterised by occurrence of hot springs, geysers and deposition of silica sinter but where there has not been any volcanic eruption in the past. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
trending zone that is related to the suture between the Archean Tanzania Craton in the south-west and the Proterozoic Mozambique belt in the north-east. The craton margin zone is accompanied by several first-order NW trending and second order NNW-SSE trending lineaments, characterized by ductile and brittle shears and well documented in the Proterozoic rocks east and west of the rift. Smith and Mosley (1993) have interpreted the shear zones as major zones of weakness that were controlling rift propagation and accommodation of tectonic stress within several rift sectors. The course and faulting style of the rift often changes at the intersections with the shear zones. On a regional scale, the rift sections influenced by the reworked craton margin are located roughly between Lake Baringo in the north and Lake Magadi in the south. However, on a local scale, NW–SE and NNW–SSE trending fractures presumed to be under basement control are evident. Such zones have been described and referred to by Smith and Mosley (1993) as ‘accommodation zones’ and include the Kerio– Bogoria–Marmanet zone in the Lake Bogoria/Lake Baringo area and the Engorika–Magadi–Lembolos zone in Lake Magadi area. Baker et al. (1972) and Atmaoui and Hollnack (2003) have further discussed oblique shear zones crossing the Kenya rift.
75
trend of the East African rift system (Fig. 4). Usually small magnitude earthquakes are common on the eastern branch (Kenya rift) of EARS and these have been described as earthquake swarms by Ibsvon Seht et al. (2008), who have further given an overview of the occurrence of earthquake swarms in the Kenya rift. Historical earthquake records show that, in comparison to the whole of the EARS, the Kenya rift shows a relatively low seismic activity. Except for the January 6, 1928 earthquake (Ms 6.9) and aftershock (Ms 6.0) four days later in the central part of Kenya rift, as well as the 1913 Turkana region earthquake (Ms 6.0) in northern Kenya, no M > 5 earthquakes in Kenya have been reported since 1928. However, numerous local earthquake studies by Rykounov et al. (1972), Pointing et al. (1985), Maguire et al. (1986, 1988), Young et al. (1991), Tongue et al. (1992, 1994), Tongue (1992) and Ibs-von Seht et al. (2001) have established a high microearthquake activity in the Kenya rift. Ibs-von Seht et al. (2001) reported a constant rate of 10 M < 3 earthquakes per day for the southernmost part of the Kenya rift, close to the Kenya-Tanzania border. The background seismicity of the Kenya rift in this southern part was restricted to depths of 10 km whereas the earthquake swarms were generally shallow with most of the hypocenters between 0 and 6 km deep (Ibs-von Seht et al., 2001). For the central Kenya rift, Young et al. (1991) reported more than 500 M < 2.7 earthquakes during a three months period in Lake Bogoria region. Tongue et al. (1994) observed a short lasting (less than one day) M < 2 earthquake swarm in Lake Baringo region. The swarm earthquakes were found to form a narrow, elongated cluster in the center of the rift at 5 km depth. Young et al. (1991) and Tongue et al. (1994) attributed the earthquake swarm activity to emplacement of dikes. These dikes have been mapped using a fairly dense network of recent gravity data by Mulwa (2010), Mulwa et al. (2009). Some results of historical seismicity in central Kenya have been applied in the search for geothermal energy along the Kenya rift valley. The United States Geological Survey carried out seismic studies at Lake Bogoria and Olkaria areas in the central Kenya rift valley (Hamilton et al., 1973). Their results show that earthquakes of magnitude 2 or less that were located are restricted mainly within the fields along fault zones, and time distance plots indicated Lake Bogoria and Olkaria areas are underlain by a three layer volcanic sequence of about 3.5 km thick. This sequence is in turn underlain by a layer with a P-wave velocity of 6.3 km/s. Their
3. Seismicity 3.1. Observations of seismicity from 1906 to 1993 Since the late 1960s, a number of seismic investigations, including the operation of temporary seismic networks, have been carried out in Kenya by Wohlenberg (1968), Rykounov et al. (1972), Hamilton et al. (1973), Pointing et al. (1985), Maguire et al. (1986, 1988), Fairhead and Stuart (1982), Young et al. (1991), Tongue et al. (1992, 1994), Tongue (1992) and most recently by Ibs-von Seht et al. (2001). The results from these investigations document the presence of local centers of seismic activity. The results show that Kenya and the surrounding regions, especially northern Tanzania, are characterized by a moderate level of seismicity which is mainly controlled by the structural
Fig. 4. Seismicity in Kenya for the period 1906–2010 for local magnitudes 0.0 < Ml 6 7.1 (After Mulwa, 2011). All focal depths are within the crust (630 km).
76
J.K. Mulwa et al. / Journal of African Earth Sciences 89 (2014) 72–78
interpreted model shows a structure with velocities higher than the average upper crustal velocities within the rift. Further, fewer events were recorded in Lake Bogoria compared to Olkaria. 3.2. Observations of seismicity from 1993 to 2010 As discussed in the introductory section, since 1993 when a network of five seismic stations was installed, earthquake monitoring in Kenya has mainly been carried out by the University of Nairobi, Department of Geology. The management, data collection and analysis are carried out by the University of Nairobi’s seismological group which falls under Applied Geophysics and Seismology thematic area within the Department of Geology. The seismological group is also responsible for seismic data analysis, management and archiving. In 2010, a three component broadband Africa Array seismic station was installed in Chiromo campus (1.274°S, 36.804°E) of the University of Nairobi. Further between 2010 and early 2011, two broadband seismic stations were installed in Kibwezi and Lodwar as part of the Indian Ocean Tsunami Early Warning System (IOTEWS). These newly installed seismic stations together with the IRIS seismic station (KMBO) installed by United States Geological Survey (USGS) in September 1995 at Kilimambgo comprise the current Kenya National Seismic Network. For the purpose of this review, four earthquake datasets for the recent seismicity have been used: (i) data from the temporary seismic network (Magadi network) consisting of 15 stations installed and operated for eight months by Ibs-von Seht et al. (2001), (ii) the current Kenya National Seismic Network (KNSN), (iii) earthquake data from the catalogue of Hollnack and Stangl (1998), and (iv) USGS/NEIC earthquake catalogue. Arrival times of seismic phases from (i) have been relocated (Kuria et al. (2010), and from (ii) located using SEISAN program (Havskov, 1997) and the crustal velocity model derived from the results of the KRISP 1994 experiment (Prodehl et al., 1997). However, earthquake datasets from the Hollnack and Stangl (1998) and the USGS/NEIC catalogues, as well as the historical earthquake catalogue have been used as they are. The original earthquake data for the Magadi network operated from November 1997 and June 1998 as well as the Hollnack and Stangl (1998) catalogue were obtained in a CD from the Federal Institute for Geosciences and Natural Resources (BGR) courtesy of Ibs-von Seht. The review of the seismicity in Kenya from 1906 to 2010 (Fig. 4) shows that the entire N-S trending rift system, including the E–W trending Kavirondo (Nyanza) rift are characterized by a high rate of seismicity. Even though the occurrence of large earthquakes in Kenya is rather low, historical records based on macroseismic information, supplemented by re-examination of instrumental reports to re-evaluate the position and size of major earthquakes, as well as recent deployments of seismic monitoring stations and networks by the United States Geological Survey (USGS) since 1963 as part of the World Wide Standardized Seismic Network (WWSSN), University of Nairobi’s Department of Geology under the auspices of Kenya National Seismic Network (KNSN), and most recently in 2002 by the International Monitoring System (IMS) show a somewhat increase in the number of earthquakes in Kenya. However, this increase in the number earthquakes especially between the period 1990–2008 does not necessarily imply an increase in seismicity but rather is as a result of good detection capability following installation of the local Kenyan National Seismic Network (KNSN) since 1995. Fig. 4 shows that Kenya has experienced strong earthquakes of magnitude Ml P 6.0 between the period 1906–1930. The earthquake records for the period 1906–1989 have been obtained from published macroseismic information by Shah (1986), Ambraseys (1991), and Ambraseys and Adams (1991) while records for the period 1973–2010 have been obtained from the Kenyan National
Seismic Network and the USGS/National Earthquake Information Center (NEIC) earthquake catalogues. It is worthwhile to note that earthquake magnitudes in Fig. 4 are the local magnitudes. Consequently, magnitude homogenization for earthquakes from Ambraseys (1991), Ambraseys and Adams (1991) and the USGS/NEIC catalogues has been undertaken. However, owing to lack of a defined criteria for magnitude homogenization in Kenya, we adopted magnitude homogenization criteria by Hussein et al. (2008) to convert body wave magnitudes (mb) from the USGS/NEIC catalogue, and the Gutenberg and Richter (1956) criteria to convert surface wave magnitudes (Ms) from Ambraseys (1991) and Ambraseys and Adams (1991) catalogues, to local magnitude scales as shown below (Hussein et al., 2008; respectively Gutenberg and Richter, 1956):
Ml ðIPRGÞ ¼ ð0:69 0:07Þmb ðISCÞ þ 1:67 0:29 for 3:8 6 mb 6 6:1 ð1Þ Ms ¼ 1:27ðMl 1Þ 0:016M 2l
ð2Þ
Eq. (1) is applicable to a fair degree of accuracy since in Kenya, only earthquakes of body wave magnitudes 3.8 6 mb 6 6.1 have been reported in the USGS/NEIC catalogue. Figs. 5 and 6 show plots of the total number of earthquakes (n) per year and cumulative number of earthquakes (LOG N) against magnitude for the period 1906–2010. Using the Gutenberg and Richter (1944) relation,
LOGN ¼ a bM
ð3Þ
Fig. 5. Plot of number of earthquakes (n) and LOG N (log10) of cumulative number of earthquakes against magnitude for the Kenya earthquake catalogue for the period 1906–2010.
Fig. 6. Gutenberg and Richter (1944) relation of the Kenyan earthquake catalogue for the period 1906–2010.
J.K. Mulwa et al. / Journal of African Earth Sciences 89 (2014) 72–78
the unified earthquake catalogues for the period 1906–2010 can be considered rather complete for earthquakes having Ml P 4.4. The Gutenberg and Richter (1944) relation gives a b-value of 0.79 for Kenya (Fig. 6). Hollnack and Stangl (1998) obtained a b-value of 0.8 for the Kavirondo (Nyanza) rift and a catalogue completeness value of Ml > 3.2 for the southern part of the Kenya rift using only the Kenyan catalogue for a limited period between the years 1993–1996. The magnitude range for their catalogue was 2.0 6 Ml 6 5.0. Consequently, our values for the catalogue completeness (Ml P 4.4) and b-value (0.79) are representative for the whole country and do not significantly differ from their results, and the observed b-value is characteristic of tectonic active regions like rifts.
4. Discussion and conclusion A review of the seismicity in Kenya has been undertaken using earthquake data from different sources for the period 1906–2010 e.g. Kenya National Seismic Network from 1995 to 2005, Shah (1986) from 1906 to 1979, Ambraseys (1991), Ambraseys and Adams (1991) from 1906 to 1930, Hollnack and Stangl (1998) and USGS/NEIC from 1973 to 2010. However, the seismicity of Kenya as presented in Fig. 4 can somewhat be considered to give a biased picture of the earthquake activity in Kenya for the period 1906–2010. This is mainly because of several factors. First, the configuration of the Kenya National Seismic Network (KNSN) which was initially centered around the city of Nairobi in the central part of Kenya (Hollnack and Stangl, 1998). This is further compounded by the fact that the local seismic network time of observation was limited due to equipment breakdown, and therefore not all the stations were operating throughout the whole period (1995–2005). Furthermore, the magnitude threshold of M P 3.0 has been reported by USGS/NEIC since 2000. This means that no earthquakes of M < 3.0 are reported by the USGS/NEIC but this does not necessarily imply complete absence of microearthquakes (M 6 3.0) in Kenya. The station distribution of the Kenya National Seismic Network (1995–2005) therefore does seem to influence the pattern of microearthquake activity, which is concentrated in the areas of the recording stations (Hollnack and Stangl, 1998). For the larger earthquakes, the USGS network and the current Kenya National Seismic Network has been effective in detecting M > 3.0 earthquakes. Such earthquakes do show a general pattern in that most of them occur along the Kenyan rift valley as well as along the northeast–southwest trending Kavirondo (Nyanza) rift. Fig. 4 shows that the epicentral distribution of earthquakes in Lake Victoria region along the Kavirondo (Nyanza) rift is oriented in northeast-southwest direction. This was also noted by Hollnack and Stangl (1998), and led to the assumption that the seismic activity along the Kavirondo (Nyanza) Rift continues through the lake still orientated in northeast-southwest direction. The high seismicity along the Kenyan rift valley in southern Kenya-northern Tanzania close to the Kenya–Tanzania border near Ol Doinyo Lengai (Fig. 4) is attributed to the earthquake swarm activity which occurred in July 2007 just before Ol Doinyo Lengai erupted. From historical records, the strongest earthquakes in Kenya are the 1928 Subukia earthquake (Ml 7.1) and the aftershock (Ml 6.2) which occurred four days later, as well as the 1913 (Ml 6.2) Turkana region earthquake (Fig. 4). This earthquakes show a high influence on Fig. 5 (number of earthquakes against magnitude) and therefore make the curve skewed. The epicenters of these earthquakes are located in areas where the Kenya rift valley intersects the NW–SE trending Aswa-Nyangia fault zone and Anza rift zone respectively (Fig. 1). According to Mulwa (2011), the 1928 Subukia earthquake and May 20, 1990 southern Sudan earthquake
77
are aligned in NW–SE direction along the Aswa-Nyangia fault zone. Giardini and Beranzoli (1992) and Mulwa (2011) have determined that the focal mechanism for the May 20, 1990 southern Sudan earthquake is purely due to strike slip movement. This therefore raises the question whether the three strongest earthquakes in Kenya were due to normal faulting along the Kenya rift valley or due to strike slip movements along the Aswa-Nyangia fault zone (for the 1928 Subukia earthquake) and Anza rift zone (for the 1913 Turkana region earthquake). From the data presented in this paper, we conclude the following about the seismicity in Kenya: (i) The largest magnitude earthquakes in Kenya for the period 1906–2010 are the 1928 Subukia earthquake (Ml 7.1) and its aftershock (Ml 6.2) four days later in central Kenya, as well as the 1913 Turkana region earthquake (Ml 6.2). (ii) All earthquakes of Ml P 5.1 show a general pattern in that the most seismic activity is along the Kenya rift valley and the Kavirondo (Nyanza) rift zone. (iii) The epicentral distribution along the NE-SW trending Kavirondo (Nyanza) rift crosses Lake Victoria and this implies a southwest continuation of this seismically active rift zone. (iv) The Rift Valley shows more seismic activity in northern Tanzania than in Kenya and this is partly attributed to seismic swarm activity around Ol Doinyo Lengai volcano which erupted in July 2007. (v) Despite the bias by the configuration of the KNSN, there is generally high microearthquake activity in the Kenyan rift valley which warrants close monitoring. (vi) Contrary to Hollnack and Stangl (1998), it is debatable whether the main seismic energy release is due to strong earthquakes in the Kenyan Rift Valley or due to strike slip movement along the NW-SE trending fault zones such as Aswa-Nyangia fault zone and the Anza rift zone. (vii) There is need therefore to extend the KNSN for a greater coverage and effective seismic monitoring in Kenya.
Acknowledgements We would like to express our gratitude to Japan International Cooperation Agency (JICA) for granting the principal author a scholarship so as to pursue a nine months’ training on earthquake, tsunami and volcano eruption observation systems at the Research Center for Seismology, Volcanology and Disaster mitigation, Nagoya University in Japan. We are also indebted to thank the principal authors’ employer (University of Nairobi) and the Kenyan Ministry of State for Public Service for granting him study leave. Finally, we wish to express our sincere gratitude to Dr. Malte Ibs-von Seht of the Federal Institute for Geosciences and Natural Resources in Germany for availing to us the earthquake catalogue for the period 1997–2001. References Ambraseys, N.N., 1991. Earthquake hazard in the Kenya Rift: the Subukia earthquake 1928. Geophys. J. Int. 105, 253–269. Ambraseys, N.N., Adams, R.D., 1991. Reappraisal of major african earthquakes, south of 20 N, 1900–1930. Nat. Hazards 4, 389–419. Ashley, G.M., Mworia, J.M., Muasya, A.M., Owen, R.B., Driese, S.G., Hover, V.C., Renaut, R.W., Goman, M.F., Mathai, S., Blatt, S.H., 2004. Sedimentation and recent history of a freshwater wetland in a semi-arid environment: Loboi swamp, Kenya, East Africa. Sedimentology 51, 1–21. Atmaoui, N., Hollnack, D., 2003. Neotectonics and extension direction of the Southern Kenya rift, Lake Magadi. Tectonophysics 364, 71–83. Baker, B.H., 1986. Tectonics and volcanism of the southern Kenya rift valley and its influence on sedimentation in the African rifts. In: Frostick, L.E., Renaut, R.W., Reid, I., Tiercelin, J.J. (Eds.), Geol. Soc. Lond. Spec. Publ. 25, pp. 45–57.
78
J.K. Mulwa et al. / Journal of African Earth Sciences 89 (2014) 72–78
Baker, B.H., 1987. Outline of the petrology of the Kenya rift alkaline province. Geol. Soc. Lond. Spec. Publ. 30, 293–311. Baker, B.H., Wohlenberg, J., 1971. Structure and evolution of the Kenya rift valley. Nature 229, 538–542. Baker, B.H., Mohr, P.A., Williams, L.A.J., 1972. Geology of the eastern rift system of Africa. Geol. Soc. Am. Spec. Pap. 136, 1–67. Baker, B.H., Mitchell, J.G., Williams, L.A.J., 1988. Stratigraphy, geochronology and volcano-tectonic evolution of the Kedong–Naivasha–Kinangop region, Gregory rift valley, Kenya. J. Geol. Soc. Lond. 145, 107–116. Braile, L.W., Keller, G.R., Wendland, R.F., Morgan, P., Khan, M.A., 1995. The East African rift system. In: Olsen, K.H. (Ed.), Continental Rifts: Evolution, Structure, and Tectonics. Developments in Geotectonics, vol. 25, pp. 213–231. Chorowicz, J., 2005. The East African rift system. J. Afr. Earth Sci. 43, 379–410. Fairhead, J., Stuart, G., 1982. The seismicity of the EARS and comparison with other continental rift. In: Palmason, G. (Ed), Geodynamics Series AGU, vol. 8. Washington, DC, pp. 41–61. Giardini, D., Beranzoli, L., 1992. Waveform modeling of the May 20, 1990 Sudan earthquake. Tectonophysics 209, 105–114. Gregory, J.W., 1921. The Rift Valleys and Geology of East Africa. Seeley Service, London. Gutenberg, B., Richter, C.F., 1944. Frequency of earthquakes in California. Bull. Seismol. Soc. Am. 34, 185–188. Gutenberg, B., Richter, C.F., 1956. Magnitude and energy of earthquakes. Ann. Geofis. 9, 1–15. Hamilton, R.M., Smith, B.E., Knapp, F., 1973. Earthquakes in geothermal areas near Lake Naivasha and Lake Hannington. Unpubl. Rept. by United States geological survey for the UNDP/EAPL, p. 37. Havskov, J., 1997. The SEISAN Earthquake Analysis Software for IBM PC and SUN. Ver. 6.0., Institute of Solid Earth Physics, University of Bergen, Norway. Hollnack, D., Stangl, R., 1998. The seismicity related to the southern part of the Kenya rift. J. Afr. Earth Sci. 26, 477–495. Hussein, H.M., Abou Elenean, K.M., Marzouk, I.A., Peresan, A., Korrat, I.M., Abu ElNader, E., Panza, G.F., El-Gabry, M.N., 2008. Integration and magnitude homogenization of the Egyptian earthquake catalogue. Nat. Hazards 47, 525– 546. Ibs-von Seht, M., Blumenstein, S., Wagner, R., Hollnack, D., Wohlenberg, J., 2001. Seismicity, seismotectonics and crustal structure of the southern Kenya rift – new data from the Lake Magadi area. Geophys. J. Int. 146, 439–453. Ibs-von Seht, M., Plenefisch, T., Klinge, K., 2008. Earthquake swarms in continental rifts – a comparison of selected cases in America, Africa and Europe. Tectonophysics 452, 66–77. King, B.C., Chapman, G.R., 1972. Volcanism of the Kenya Rift Valley. Phil. Trans. R. Soc. A271, 185–208. Kuria, Z.N., Woldai, T., van der Meer, F.D., Barongo, J.O., 2010. Active fault segments as potential earthquake sources: Inferences from integrated geophysical mapping of the Magadi fault system, southern Kenya Rift. J. Afr. Earth. Sci. 57 (4), 345–359. Logatchev, N.A., Beloussov, V.V., Milanovsky, E.E., 1972. East African Rift development. Tectonophyiscs 15, 71–81. Maguire, P.K., Cooke, P.A.V., Laffoley, N.d’.A., Evans, J.R., 1986. Earthquake recording in the central part of the Kenya Rift Valley. EOS 67, 1103.
Maguire, P.K., Shah, E.R., Pointing, A.J., Cooke, P.A., Khan, M.A., Swain, C.J., 1988. The seismicity of Kenya. J. Afr. Earth Sci. 7, 915–923. Mariita, N.O., 2003. An Integrated Geophysical Study of the Northern Kenya Rift Crustal Structure: Implications for Geothermal Energy Prospecting for Menengai Area. Ph.D. Dissertation, University of Texas at El Paso, USA. Mulwa, J.K., 2010. Integrated geophysical study of Lake Bogoria basin, Kenya: Implications for geothermal energy prospecting. Ph.D. dissertation, University of Nairobi, Kenya. Mulwa, J.K., 2011. Determination of source parameters for the May 20, 1990 southern Sudan earthquake by inversion of teleseismic body-waves. In: JICA Training Program, Operating management of Earthquake, Tsunami and Volcano Eruption observation system, Fiscal year 2010. Research Center for Seismology Volcanology and Disaster Mitigation, Japan, pp. 15–31. Mulwa, J.K., Fairhead, D., Barongo, J.O., Mariita, N.O., 2009. Heat source mapping and evaluation of geothermal resource potential in Lake Bogoria basin, Kenya. SEG Expand. Abstr. 28, 1294–1299. Pointing, A.J., Maguire, P.K.H., Khan, M.A., Francis, D.J., Swain, C.J., Shah, E.R., Griffiths, D.H., 1985. Seismicity of the northern part of the Kenya rift valley. J. Geodyn. 3, 23–37. Prodehl, C., Ritter, J., Mechie, J., Keller, G.R., Khan, M.A., Fuchs, K., Nyambok, I.O., Obel, J.D., 1997. The KRISP 94 lithospheric investigation of southern Kenya-the experiments and their main results. In: Fuchs, K., Altherr, R., Muller, B., Prodehl, C. (Eds.), Stress and Stress Release in the Lithosphere, Tectonophysics, vol. 278, pp. 121–148. Rykounov, L.N., Sedov, V.V., Savrina, L.A., Bourmin, V.J., 1972. Study of microearthquakes in the rift zones of East Africa. Tectonophysics 25, 123–130. Saggerson, E.P., 1991. Geology of the Nairobi area. Geol. Surv. Kenya, Rep. No. 98. Schlüter, T., 1997. Geology of East Africa. Beiträge zur Regionalen Geologie der Erde, vol. 27. Gebrüder Bornträger, Berlin-Stuttgart, 484pp. Shah, E., 1986. Seismicity of Kenya. Ph.D. Thesis, University of Nairobi, Kenya. Sikes, H.L., 1926. The structure of the eastern flank of the Rift valley near Nairobi. Geogrl. J. 68. Simiyu, S.M., Keller, G.R., 2001. An integrated geophysical analysis of the upper crust of the southern Kenya rift. Geophys. J. Int. 147, 543–561. Smith, M., 1994. Stratigraphic and structural constraints on mechanisms of active rifting in the Gregory Rift, Kenya. In: Prodehl, C., Keller, G.R., Khan, M.A. (Eds.), Crustal and Upper Mantle Structure of the Kenya Rift. Tectonophysics, vol. 236, pp. 3–22. Smith, M., Mosley, P.N., 1993. Crustal heterogeneity and basement influence on the development of the Kenya Rift, East Africa. Tectonics 12, 591–606. Tongue, J.A., 1992. Tomographic study of local earthquake data from the Lake Bogoria region of the Kenya Rift Valley. Geophys. J. Int. 109, 249–258. Tongue, J.A., Maguire, P.K.H., Young, P.A.V., 1992. Seismicity distribution from temporary earthquake recording networks in Kenya. Tectonophysics 204, 71–79. Tongue, J., Maguire, P., Burton, P., 1994. An earthquake study in the Lake Baringo basin of the central Kenya Rift. Tectonophysics 236, 151–164. Wohlenberg, J., 1968. Remarks on the seismicity of East Africa between 4°N–12°S and 23°–40°E. Tectonophysics 8, 567–577. Young, P.A.V., Maguire, P.K.H., Laffoley, d.’.A., Evans, J.R., 1991. Implications of the distribution of seismicity near Lake Bogoria in the Kenya Rift. Geophys. J. Int. 105, 665–674.