|
|
 |
|||||||||||||||||
|
|||||||||||||||||
 |
|||||||||||||||||
| JOURNAL HOME | HELP | CONTACT PUBLISHER | SUBSCRIBE | ARCHIVE | SEARCH | TABLE OF CONTENTS |
The city of Karachi, Pakistan (population 14 million), sits close to a plate boundary and within reach of earthquakes on numerous tectonically active structures surrounding the city. One can draw parallels—geologic as well as demographic—with another megacity for which seismic hazard is known to be high: Los Angeles, California (figure 1). Yet with a short historical record, limited instrumental seismic data, and little geological or geodetic constraint on slip rates, seismic hazard in Karachi is poorly characterized. In this report we present a critical review of the historical record as well as an overview of potential earthquake sources in and around Karachi.
Prior to 1800, the history of Karachi, the current capital of Sindh (one of Pakistan's four provinces), is indistinguishable from that of many fishing villages on the northern shores of the Arabian Sea. It was known to Arab and Portuguese traders who sometimes stopped at the village on the way to the Malabar coast. Colonial trade with Sindh was limited in the 18th century. Exploratory surveys starting in 1808 led to the annexation of Sindh to British rule in 1843. By 1901 Karachi had grown from a village to a town with a population of fewer than 140,000 people. The population grew to 500,000 by the time Pakistan became an independent nation in 1947, at which time the city grew dramatically with the influx of a million refugees from India. By 1960 its population exceeded 2 million, a figure that had doubled by 1980 and more than doubled again to 10 million by 2000. With an estimated annual growth rate of 3.75%, its population will exceed 30 million within two decades.
Karachi lies approximately 150 km east of the triple junction between the Arabian, Indian, and Asian plates (figure 2). The western and north-trending arms of the triple junction sustain convergent and transcurrent rates of 28-33 mm/yr respectively (Apel et al. 2006). A recently discovered active fault, the Sonne fault, indicates that the Arabian plate has been fragmented across the southwest corner of the triple junction defining a triangular plate, the Ormara plate (Kukowski et al. 2000) whose velocity relative to the Arabian plate increases subduction velocities by a few millimeters per year compared to the rate to the west. In addition to these clearly defined plate boundaries, two other active structural zones have produced damaging earthquakes that have been felt in the city in the past 200 years: a thrust-and-fold belt extending northward parallel to the transform fault separating India from Asia, and the Kachchh fault system trending westward toward the city.
Although residents of Karachi felt shaking from the 1945 Makran and 2001 Bhuj earthquakes, and occasional shaking from M 4-5 earthquakes on faults north and northwest of the city, no earthquake has ever produced documented damage in Karachi. The question currently faced by earthquake engineers is whether Karachi truly enjoys an aseismic setting or whether the absence of damaging earthquakes is only due to Karachi's short and incomplete history. A review of the known historical data on earthquakes within 500 km of the city shows that the historical record prior to 1800 is limited and unreliable.
 |
MISLOCATED AND POORLY LOCATED HISTORICAL EARTHQUAKES |
|---|
 TOP MISLOCATED AND POORLY LOCATED... MAKRAN SUBDUCTION AND THE...ACTIVE FAULTS NORTH AND...CONCLUSIONSACKNOWLEDGMENTSREFERENCES |
|---|
|
|
|
It was reported from the province of Tatta, that the village of Samawani in the jurisdiction of the port of Lari, had sunk down with 30,000 residents, owing to an earthquake.
Perkins in Elliot (1857) states:
On the 1st Zí-l hijja, 1078 A.H. (3rd May, 1668), the intelligence arrived from Thatta that the town of Samájí had been destroyed by an earthquake; thirty thousand houses were thrown down.
and Oldham (1883) citing Biblioteca Indica 66, 74 (1874), reproduces it as follows:
At this time (between the 1st and 10th Zí hajja, 1078 A.H) a report was received from the Soobah of Tattah that the town of Samawani (or Samanji) which belongs to the Parganah of Láhori had sunk into the ground with 30,000 houses, during an earthquake.
The date of the earthquake (2-11 May 1668) is imprecise since it is inferred to have occurred between contiguous entries for which dates are provided (Ambraseys 2004). Samawani has been translated with the following variations: Jamawani (Le Gentil c. 1770, compiled by his son in 1820 and reproduced in Gole 1988), Semadany (Gladwin 1835), Samaji or Samanji (Sarkar 1947), Samawadi (Elliot 1857) and Summawati (Habib 1982).
The various translations of the original Persian text give rise to inconsistencies in interpretation. Thatta was the name of both the province (subah) of Thatta and one of its five judicial subdivisions (pargannah), as well as a town of the same name (figure 3). As described in the A'in-i Akbari of Fazl-i-'Allami (Sarkar 1978), the coastal port of Bandar Lahori (long since abandoned) was not a jurisdiction, but the largest town within the jurisdiction of Thatta. Samawani, however, lay in the jurisdiction of Nasarpur (Nassirpur, Nasrpur, or Nassirpoor) north of the jurisdiction of Chachgan (figure 4).
|
|
The size of Samawani in 1596 can be judged from its Mughal taxation revenues assessed in dams, the coinage of the time, which exceeded the combined revenues of the next two largest villages in the Nasarpur jurisdiction, Nasarpur itself and Umarkot (Jarrett 1891). As of 1596, among 53 Mughal villages, Samawani ranked as the fifth-largest town in Sindh, with roughly half the revenue of the main port of Bandar Lahori (figure 4). Towns with similar revenues in Multan were able to muster 100-1,000 cavalry and 1,000-20,000 infantry troops, suggesting a sizable contributing population. (The A'in-i Akbari does not list conscription levels for the Thatta district.) Thus although its size may have declined by the time of the earthquake, as the largest town in the Nasarpur jurisdiction in 1596, Musta'idd Khan's 1710 estimate of the number of dwellings that sank in the Samawani earthquake (30,000), although obviously an approximation, may not have been significantly exaggerated.
The magnitude of the 1668 earthquake cannot be estimated from the single
brief report, although this has not prevented previous authors from assigning
it values in the range 6.5 < M < 7.6 (for references see
Ambraseys 2004). If Samawani
were on the banks of the mile-wide Indus, damage could have been caused by the
lateral spreading of soils, slumping, liquefaction, or even a tsunami.
Ambraseys points out that the earthquake is likely to have been modest because
damage to neither Bandar Lahori nor Thatta was reported in 1668. With Samawani
located 100 km north-northeast of Thatta, however, one would need to look
elsewhere for collateral damage. The temple structures of Thatta survived 
100 km from the 7.6 < M < 8 Allah Bund 1819 earthquake, and with
minor damage 200 km from the Mw = 7.6 Bhuj 2001
earthquake. Thus the absence of damage in Bandar Lahori and Thatta does no
more than place an upper limit (e.g. M 7) on the magnitude of
the 1668 earthquake. News of the earthquake was reported not from the town of
Thatta but from the province of Thatta, and no other towns are mentioned. In
particular no damage was reported from nearby Nasarpur, which was abandoned
five decades after the earthquake because of the slow subsequent avulsion of
the Indus.
Allah Bund 1819
On 16 June 1819 a severe earthquake in the northern Rann of Kachchh caused
the deaths of 2,000 people (anonymous
1820; McMurdo
1823; Oldham 1883,
1926). The earthquake created
a 30 x 20-km basin around Fort Sindri
(figure 5), south of a
10-km-wide elongated region of uplift known as the Allah Bund
(Burnes 1834;
Baker, 1846; Oldham,
1898,
1926). The vertical scale is
increased erroneously by a factor of two in the copy of Oldham
(1898) reproduced posthumously
by Montessus de Ballore
(1924). The extent of the
subsidence was immediately apparent because within an hour the depression had
flooded with sea water brought in by an ocean tsunami. The Allah Bund
earthquake has the distinction of being the first in India for which geodetic
leveling data (Baker 1846) can
be used to constrain aspects of the rupture
(Bilham 1998). The earthquake
resulted in local uplift (the Allah Bund) that raised the bed of the river. At
the time of the earthquake the river was dry due to the diversion of its
waters by an artificial dam upstream. The channel remained dry for seven years
until in 1826 a flood burst the upstream dam, and water formed a temporary
lake north of the raised channel of the Allah Bund. The waters eventually
overtopped the lowest point of the channel and incised a path through it,
displacing the saline waters of the basin surrounding Fort Sindri with fresh
water (Burnes 1833).
Baker's 1844 profile shows a maximum elevation of the crest of the Allah
Bund of 6.2 m above Lake Sindri (figure
6), a number that has been used as a measure of coseismic uplift
assuming an absence of preseismic topography. However, there is an ambiguity
in Baker's data, in that if the bed of the river (DA) were raised, its bank
should also have been raised (EC dashed line). Baker leveled the last 31 km
south to Lake Sindri in a single day (11 July 1844), which may have been
responsible for a diminution in accuracy
(Yule and Maclagan, 1882). If
we assume the measured bed profile (DA) is not caused by bank collapse we can
obtain an independent estimate of coseismic uplift of >4 m and not greater
than 6 m, since if the lip of the channel were higher than point F
(figure 6) the stream would
have chosen a path around the Bund rather than through it
(Wynne 1872). However, the
measured bank profile (EC solid line) should also have been raised to mimic
the bed profile (DA). In places the measured bank profile is more than a meter
too low, suggesting possible bank collapse or erosion between 1819 and 1844.
Reid (1911) summarizes the
controversy, but no measurements of the Bund were made until Rajendran and
Rajendran (2001) measured
several profiles across it. They report variable relief along strike and a
maximum elevation of the crust of the Bund of 5.3 m relative to the present
Lake Sindri shoreline to the south. They also describe terraces that they
interpret as evidence for multiple uplift events, with uplift of the most
recent terrace by no more than 4.3 m. However, these terraces have not been
confirmed by subsequent visitors. Rajendran and Rajendran
(2001) also provide evidence
for multiple earthquakes in the region in the form of dated materials
associated with paleoliquefaction features in pits excavated north of the
Allah Bund. We note that these need not necessarily have been formed by former
earthquakes repeatedly raising the Allah Bund. Liquefaction could have been
caused by any large earthquake within 100 km of the Allah Bund. The Bhuj
earthquake for example, 70 km to the southwest demonstrated that regional
liquefaction occurred throughout the Rann of Kachchh
(Tuttle et al.
2002).
A dip of 68° NNE for the rupture was inferred from the ratio of uplift
to subsidence using Baker's data (Bilham
1998). However, uncertainty as to the maximum subsidence is caused
by the truncation of Baker's profile close to the shores of Lake
Sindri—it is unclear whether his maximum depth of 3 m sampled the lake
floor or the river channel. This area is now a deltaic platform extending
several hundreds of meters into Lake Sindri. A minimum length to the rupture
( 50 km) may be inferred from the east-west basin dimension of Lake
Sindri, and the maximum length has been inferred from the morphology of the
northern edge of the Rann of Kachchh as 80-150 km by Oldham (1926).
|
The 1819 earthquake was accompanied by aftershocks that decayed in intensity over the following year, but sporadic earthquakes were reported during the next 50 years. Few of the half dozen earthquakes that were felt prior to 1900, however, were described in sufficient detail to assign a magnitude or even a location. An earthquake in about 1846 occurred with numerous aftershocks, a small tsunami, and apparent ground deformation near Lakhpat. Although the true dates of these events are confused in different accounts (Bilham 1998), the 20-km region of uplift reported near Sunda involved at least 60 cm of uplift, which would probably require Mw > 6.5 given simple assumptions and a reverse mechanism.
An earthquake struck the Bhuj region in January 1856, at a time when a geodetic survey was underway. A masonry pillar at Kararho (23°5'N, 70°13'E) supporting the theodolite was damaged sufficiently to cause anxiety among the survey team (Burrard 1890). The documented effects suggest a locally damaging earthquake, perhaps around Mw 6. Anything larger would surely have been reported from Bhuj or Anjar. The most damaging earthquake near Bhuj in the region prior to 2001 occurred in Anjar in 1956 (Mw = 6.1), and this was accompanied by > 50 cm of vertical deformation east of Anjar.
Bhuj 2001
The 26 January 2001 Mw = 7.6 earthquake occurred
approximately 100 km east of the Allah Bund and resulted in 18,500 deaths due
to collapsed buildings in the Bhuj-Anjar region
(Bendick et al.
2001). The earthquake is of interest in that its magnitude was
large for its comparatively small rupture area—a 20 x 20-km
reverse fault below 9-km depth (Jade
et al. 2003; Wallace
et al. 2006). The earthquake did not occur on the same
fault that produced the 1819 event, but both faults are within the
east-west-trending Kachchh fault system. The small rupture area of the Bhuj
earthquake suggests that several such earthquakes may have occurred in the
past (or may do so in the future) between the 1819 and 2001 ruptures, or in an
extended region of the same structure to the east or west.
Ground motions from the Bhuj earthquake were strong enough in Karachi to cause doors to open and close, and some buildings were reportedly "cracked" (see Hough et al. 2002). Of concern for Karachi is the prospect of similar events on extensions of the Kachchh fault system toward the west. It is not clear if and how these fault structures extend to the west away from the Kachchh mainland. The Allah Bund appears to trend northwest or west at its most westerly expression, and Sawar (2004) argues that this trend extends beneath the Indus west toward Karachi and to the north toward the Himalaya. Stein et al. (2002) extend the Rann of Kachchh fault zone to the west-southwest, but conclude that the eastward extension of the Kachchh zone curves toward the northwest following a weak line of microseismicity. They interpret the large Rann of Kachchh earthquakes and the scattered seismicity through the deserts of Rajasthan and Sindh as defining a fragment of the Indian plate moving at a velocity on the order of 3 mm/yr relative to India.
|
MAKRAN SUBDUCTION AND THE TRANSFORM BOUNDARY |
|---|
|
TOPMISLOCATED AND POORLY LOCATED...MAKRAN SUBDUCTION AND THE...ACTIVE FAULTS NORTH AND...CONCLUSIONSACKNOWLEDGMENTSREFERENCES |
|---|
Ormara 1700-1800
An earthquake is alleged to have caused a landslide on the Makran coast 150
km west of Karachi at some time in the 18th century. Ambraseys and Melville
(1982) list the approximate
year as 1765. The source of this information is the director of the Makran
telegraph line, who wrote a note to the Government of Bombay indicating that
"a smart shock of an earthquake" had been felt in Gwadar at 00:45
on the morning of 25 August 1864 and speculated that large earthquakes, if
they occurred along the Makran coast, could jeopardize telegraph
communications (Walton 1864).
From its brief mention in telegrams from Gwadar but not from other telegraph
posts along the Makran coast, the 1864 earthquake was presumably quite small
(M < 5). Walton's letter, however, continues with the description of
a possible earthquake remembered by local people:
As the entire coast of Mekran [sic] is volcanic, I often enquired of the Baluchees regarding the occurrence of earthquakes, and the only phenomenon of this sort, of which I could obtain any information, was said to have happened about 100 years ago, when, as my informant assured me, an entire hill, with men and camels on it, disappeared into the sea. I imagine this must have been a landslip caused by some submarine disturbance. The spot was pointed out to me and is known as Ras Koocheree on the chart.
(Walton's assessment of the Makran coast as volcanic was based on the mud volcanoes found along the coast.)
The precise identification of 1765 as the year of a great earthquake is obviously inappropriate based on the vague mention of the timing of the earlier event. Byrne et al. (1992) identify this earthquake as a great rupture beneath the leading edge of Asia at the easternmost end of the Makran subduction zone. However, a landslide could have also been triggered by a large strike-slip earthquake on the southernmost Ornach-Nal fault system. It is even possible that no large earthquake occurred around 1765; the landslide could have been spontaneous, or the result of unusually heavy rain.
Byrne et al. (1992) also state that a great earthquake may have occurred on the western Makran subduction zone in 1851; however, we find no evidence for this interpretation. An earthquake apparently did occur along the Makran coast in 1851; however, while Byrne et al. (1992) cite Quittmeyer and Jacob (1979) as a reference regarding this event, the latter paper includes only brief mention that an earthquake was reported at Gwadar on 19 April 1851, in turn citing Oldham (1883), who notes that Merewether (1852) records that several houses collapsed in three shocks at 5 p.m. that day. Quittmeyer and Jacob (1979) do not include the earthquake on their list of documented south-central Asian earthquakes that produced maximum intensity of VIII or greater. One can further recall Walton's letter from Gwadar, written just 13 years after 1851. Had a major earthquake occurred along the Makran coast in 1851, surely local people would have been aware of it.
Makran 1945
In contrast to the vague information about 18th- and 19th-century Makran
earthquakes, the 28 November 1945 Mw 8.1
subduction zone earthquake is well-constrained. This event occurred 250-350
west of Karachi (figure 2) and
was recorded globally. Walton's 1864 concerns about submarine cables were
borne out when this earthquake occurred, as cables broke in eight places due
to submarine slides. To reconcile observed shoreline uplift with the mechanism
of the earthquake, Byrne et al.
(1992) argue for a rupture on
a thrust plane with a 5° dip, extending 70-90 km inland to 10-30 km
offshore. Based on the distribution of intensities, morphological changes, and
the sparse distribution of recorded aftershocks, the rupture is inferred to
have propagated south-southeast with a duration of 56 s, suggesting a rupture
dimension of 80-150 km extending eastward from Pasni. Numerous mud volcanoes
erupted inland and offshore, creating four transient islands of stiff clay
with 200-300 m dimensions in 10-m water depths coastward of a line
connecting the horsts at Pasni and Ormara. The coastline at Pasni subsided and
the tombola at Ormara was raised 2 m. Numerous aftershocks were recorded and
locally felt (Ambraseys and Melville
1982; Byrne et al.
1992).
Important unresolved problems attend the interpretation of the tsunami generated by the 1945 earthquake. The tsunami was reported as a 0.5-m wave in the Seychelles 3,400 km to the southeast; it broke mooring ropes at 13°N on the Malabar coast and was noticed at Muscat on the Oman coast. At Pasni a small wave arrived soon after the mainshock, but according to Pendse (1948), who does not cite his source, the damaging tsunami did not arrive until 3.25 hours later. Pendse (1948) also describes the damaging 1.5-m wave at Karachi as having followed three earlier, smaller waves during the previous two hours. Ambraseys and Melville (1982) indicate that two damaging waves arrived at Pasni 90 and 120 minutes after the mainshock. Although accounts conflict to some extent, it appears that the largest tsunami wave did not arrive at Pasni until several hours following the mainshock. There are many possible explanations for this delay, the most likely of which is that it was caused by one or more submarine landslides triggered by the earthquake.
GPS Deformation Rates and Current Slip Potential, Makran
Modern global positioning system (GPS) studies indicate that Arabia
approaches the Asian plate at a velocity of 28-30 mm/yr along the Makran coast
(Apel et al. 2006).
Relative motion between the Ormara plate and Arabia increases this velocity to
32-35 mm/yr (Kukowski et al.
2000). Assuming complete seismic coupling, a maximum slip deficit
of 
2m has developed along the 1945 rupture zone. The seismic potential
of the subduction zone to the east of the 1945 rupture is less
well-constrained. As discussed above, it is possible that a large earthquake
occurred to the east of the 1945 rupture zone at some time during the 18th
century. If this is the case, a slip deficit of 6-9 m has developed along this
segment of the subduction zone. Were this earthquake to occur today, we
estimate its potential magnitude to be as high as
Mw 8.2. The slip deficit along the 1945 rupture
zone could produce an earthquake with Mw 7.8 if the event
occurred today.
We note, however, that the magnitude estimates above assume complete
seismic coupling. Kukowski et al.
(2000) conclude from the
offset of accretionary ridges by the Sonne fault that offshore locking is
strong, but it is unlikely that complete seismic coupling extends throughout
the subduction interface. Also, in contrast to the 70 km inland
preseismic locking line inferred for down-dip rupture termination of the 1945
earthquake by Byrne et al.
(1992), our recent GPS
measurements indicate that the locking line must be close to the coast or
offshore. The data are derived from continuous GPS receivers installed in
Ormara and Karachi and indicate Ormara's velocity is 21.5 ± 3 mm/yr
south-southwest relative to India. Assuming that aseismic slip occurs downdip
on a planar subhorizontal dislocation, and that Asia/Ormara plate convergence
is 33mm/yr, elastic theory requires that the current locking line must be
offshore, close to the seaward termination of rupture in 1945. If further GPS
measurements along the Makran coast confirm the high rate of slip at Ormara,
it would suggest that little or no potential slip is accumulating. This result
would be of importance to Karachi since it would imply that no great
earthquake is pending, and/or that the renewal time for 1945-type events is
significantly more than the 200 years inferred from plate convergence
rates alone. Clearly, additional measurements are needed to confirm this
initial result before one can draw conclusions about the seismic potential of
the subduction zone segment to the east of the 1945 rupture zone.
|
ACTIVE FAULTS NORTH AND NORTHWEST OF KARACHI |
|---|
|
TOPMISLOCATED AND POORLY LOCATED...MAKRAN SUBDUCTION AND THE...ACTIVE FAULTS NORTH AND...CONCLUSIONSACKNOWLEDGMENTSREFERENCES |
|---|
Ornach Nal Earthquakes?
The Ornach-Nal fault (figure
2) is the southernmost of several en echelon strike faults
collectively termed the Chaman fault system
(Yeats et al. 1979)
that define the western edge of the Indian plate. Its slip rate is estimated
from geological offsets to be 20-40 mm/year
(Lawrence et al.
1992), and plate closure rates constrained by recent data suggest
a mean velocity of 26 mm/yr (Apel et
al. 2006), or 
34 mm/yr if one takes the inferred Ormara
plate velocity into account. The southernmost segment of the Ornach Nal fault
system starts 130 km due west of Karachi and extends northward for 200 km.
Based on scaling relations (e.g.,
Wells and Coppersmith 1994),
the fault could produce an earthquake with a magnitude as large as
Mw 7, or perhaps larger. Though recent faulting has been
identified (Nakata et al.
1991) the historical record contains no known earthquake on this
segment. Recent seismicity is sparse north of the plate boundary, but
increases northward away from the coast. Four hundred kilometers north of
Karachi it reaches a region of significant historical seismicity with five
damaging earthquakes starting in 1892
(Griesbach 1893) and 1909
(Heron 1911) followed by the
1931/35 Mach/Quetta earthquake sequence (West
1934,
1936), a series of three
M > 7 earthquakes that resulted in 35,000 deaths. This sequence
included large earthquakes on both the Chaman fault system and the frontal
thrust faults to the east (discussed below)
(Ambraseys and Bilham 2003). If
we assume that the Ornach Nal fault has not slipped since the 18th century, it
would now have developed 6-8 m of potential slip. If the southernmost segment
were to slip in a single event ( 200 km), it could generate a maximum
credible earthquake approaching Mw 8.0.
Blind Thrust Earthquakes near Karachi?
To the north of Karachi the well-developed Kirthar fold-and-thrust belt
verges to the east, the southernmost ranges of which are expressed both east
and west of the city (figure
2). Nakata et al.
(1991) identify a possibly
active surface scarp on the Korangi fault on the eastern outskirts of Karachi.
Shortening is driven in the north by the convergent component of strain caused
by transpressional oblique-slip of the transform boundary, but to the south
this obliquity appears to be much reduced or absent. Near Sibi the frontal
thrusts most recently slipped in 1931 (the Mach Mw = 7.1
earthquake; Ambraseys and Bilham
2003) but as the Kirthar hills are followed southward no
historical large earthquakes have been recorded, although microseismicity does
indeed occur. Estimates for the rate of shortening across this fold belt near
the latitude of Quetta vary from 5 to 11 mm/year
(Ambraseys and Bilham 2003;
Apel et al. 2006);
hence in the absence of aseismic slip or folding one might anticipate M
> 7 earthquakes to be associated with maximum renewal times of the order of
200 years. If maximum earthquake size were limited by fault segmentation to
below Mw 7, one would expect to see more frequent
moderate shocks. The paucity of such events in historical times argues against
this interpretation, suggesting instead that less-frequent, large events
should be expected.
Partitioned Convergence in the Kirthar Range?
The city of Karachi is constructed on the southernmost folds of the Kirthar
range with several named faults (Surjan, Lakhni, Jhimpir) within 125 km of the
city. Schelling (1999) argues
that activity on the easternmost frontal faults of the range near and north of
Hyderabad appears to have ceased. However, at least two faults have been
mapped near and northwest of Karachi: the Hab and the Pab faults. These faults
are thought to be active, although neither slip-rate estimates nor local
earthquake recordings are available for either of them. We note, however, that
evidence for seismicity 50-110 km northeast of Hyderabad (possible M
7 earthquakes in Brahmanabad circa 980 and Samawani in 1668) suggests
either that activity on the western-most Kirthar thrusts continues beneath the
Indus sediments, or that other structures are tectonically active in this
region.
The only instrumental seismic data for events in and around Karachi is from the National Earthquake Information Center (NEIC), which is incomplete for Mw below approximately 5. Events recorded since the early 1960s reveal few events immediately around the city of Karachi, with a small number of M 4-5 shocks that were large enough to be felt in the city (figure 2). Several events are located close to the Hab and Pab faults, including an mb 4.6 earthquake on 8 September 1986 and an mb 4.5 earthquake on 29 September 1998. Both of these earthquakes were felt in Karachi.
Seismic Hazard
Our present, imperfect understanding of earthquake hazard in Karachi
reveals an apparent paradox. On the one hand, Karachi sits close to an active
plate boundary and is literally surrounded by active faults. We note a
striking parallel between its setting and that of another well-known megacity:
Los Angeles, California (figure
1). On the other hand, in contrast to Los Angeles, Karachi has
experienced no damaging earthquakes in the past 150 years and few events large
enough to be felt.
Having summarized what is known about tectonics in and around Karachi, it is clear that key questions still remain regarding the seismogenic potential of a number of regional faults. At best the hazard in Karachi might be low, if, for example, the southern Chaman fault is creeping and/or the eastern Makran subduction zone has a shallow locking depth. At worst, however, the hazard in Karachi could be roughly comparable to that in Los Angeles, or perhaps even worse in Karachi given its proximity to the subduction zone, for which Los Angeles has no analog. Considering the number of known active faults that menace Karachi from almost every direction, however, it seems possible if not probable that hazard is higher than that assigned by recent national and global hazard maps.
A consideration of relative seismic risk in Karachi and Los Angeles leads to even more alarming conclusions. Seismic building provisions, including the Field Act to protect public school buildings, were first adopted in California following the 1933 Long Beach earthquake. The building codes have been updated and strengthened since that time. Notably, codes were strengthened in the early 1970s after the 1971 Sylmar earthquake revealed previously unsuspected vulnerabilities of non-ductile concrete buildings—the likes of which were ubiquitous throughout the region. The code has never included mandatory retrofitting, so these older and highly vulnerable buildings are a serious concern. The building code provisions were strengthened again in 1997, so buildings built between the early 1970s and 1997 are also not constructed according to current standards. Unexpected vulnerabilities of relatively modern structures were revealed as recently as 1994, when the M 6.7 Northridge earthquake caused unexpected damage to steel welds.
Although concern thus remains for the adequacy of building codes in southern California, efforts to assess hazards and implement effective codes have far outpaced comparable efforts for Karachi. Two efforts have been made in recent years to develop a national seismic hazard map for Pakistan. The first of these was the analysis done by the Global Seismic Hazard Assessment Program, or GSHAP (Giardini et al. 1999). The GSHAP map for Pakistan (figures 7 and 8) was based by necessity on global earthquake catalogs and reveals, to the trained eye, apparent strengths as well as weaknesses. One possible strength is that this map reveals fairly high hazard in southeasternmost Pakistan, a consequence of high rates of recent activity primarily across the border in India, but to some extent also a consequence of activity on the Pakistani side of the border. An obvious limitation of the GSHAP map, however, is the "bull's-eye" around the location of the 1935 Quetta earthquake and the low hazard both north and south of this location. One possible reprieve to future seismicity here is that like the San Andreas system, part of the Chaman system may be creeping. It was initially thought that the Chaman fault might be creeping to the north of Quetta (Szeliga et al. 2006), but a recent analysis of the interferograms interprets this aseismic slip as local afterslip following an M 5 earthquake on the fault (M. Furuya & S. P. Satyabala, personal communication, 2007). In general, the map reveals the inevitable limitation of a seismicity-based map in a region where the historical record is clearly much shorter than the length of the earthquake cycle. This limitation is further exacerbated by the absence of other information to constrain source models, for example GPS data and slip-rate information.
|
|
The new hazard map and its attendant peak ground acceleration (PGA) map for the Karachi region reduces the zoning appropriate for construction in Karachi back from zone 4 to zone 2b. Zone 2b is, for example, the current zonation of easternmost Nevada and the Rio Grande rift zone. As we have seen, the new map is based on scant historical data and influenced significantly by earthquakes that have occurred since 1800. It is not clear, however, how long the engineering community will take to adopt the new guidelines, or whether (and how) the government will enforce the new zoning for Karachi. In any case, new building codes will not remedy the known and worrisome vulnerability of existing structures in which 14 million people now work and live.
|
CONCLUSIONS |
|---|
|
TOPMISLOCATED AND POORLY LOCATED...MAKRAN SUBDUCTION AND THE...ACTIVE FAULTS NORTH AND...CONCLUSIONSACKNOWLEDGMENTSREFERENCES |
|---|
7.9 strike ruptures to the northwest, and
Mw 6 earthquakes near and possibly beneath the
city, little or no data are available to characterize return times and
probabilities for any of these events. Considerable insight may be gathered
about possible rates of seismic activity from GPS studies, but further
geological investigations are also essential for a better understanding of
active faults. An additional, potentially valuable source of data, especially
in the search for potential blind thrusts in the Karachi region and Indus
delta, may already exist in the form of unexplored seismic profiles associated
with hydrocarbon exploration in Pakistan. Tsunami hazards exist in Karachi and its contiguous coastline that we have not examined in this article. The > 1-hour delay between the mainshock and the arrival of the damaging tsunami associated with the 1945 earthquake was very probably caused by submarine slumping offshore rather than direct uplift of the coast. If this were indeed the case, even a quite modest earthquake in the Kachchh region might be sufficient to trigger a submarine slide that would endanger the Karachi shoreline.
|
ACKNOWLEDGMENTS |
|---|
|
TOPMISLOCATED AND POORLY LOCATED...MAKRAN SUBDUCTION AND THE...ACTIVE FAULTS NORTH AND...CONCLUSIONSACKNOWLEDGMENTSREFERENCES |
|---|
1 Cooperative Institute for Research in Environmental Sciences and the
Department of Geological Sciences, University of Colorado at Boulder, Boulder,
Colorado 80309-0399 USA 
2 Department of Civil Engineering, NED University of Engineering and Technology, Karachi 75270 Pakistan.
3 U.S. Geological Survey, 525 South Wilson Avenue, Pasadena, California, 91106 USA
|
REFERENCES |
|---|
|
TOPMISLOCATED AND POORLY LOCATED...MAKRAN SUBDUCTION AND THE...ACTIVE FAULTS NORTH AND...CONCLUSIONSACKNOWLEDGMENTSREFERENCES |
|---|
Ambraseys, N. N. (2004). Three little-known earthquakes in India. Current Science 86 (4),506 -508.[Web of Science]
Ambraseys, N. N., and C. P. Melville (1982). A History of Persian Earthquakes. Cambridge and New York: Cambridge University Press, 219 pps.
Ambraseys, N., and R. Bilham (2003). Earthquakes and
crustal deformation in northern Baluchistan. Bulletin of the
Seismological Society of America 93
(4), 1,573-1,600.
Ambraseys, N. N., and J. Douglas (2004). Magnitude calibration of North Indian earthquakes. Geophysical Journal International 158,1 -42.[CrossRef][Web of Science][GeoRef]
Anonymous. (1820). Account of the earthquake at Kutch on the 16th June 1819 drawn from published and unpublished letters from India. Edinburgh Philosophical Journal 3, Article XXII,120 -124.
Apel, E., R. Bürgmann, P. Bannerjee, and B. Nagarajan (2006). Geodetically constrained Indian plate motion and implications for plate boundary deformation. EOS, Transactions, American Geophysical Union 85,52 T51B-1524 Fall meeting supplement,
Baker, W. E. (1846). Remarks on the Allah Bund and on the drainage of the eastern part of the Sind basin. Transactions of the Bombay Geographical Society 7,186 -188.
Bellasis, A. F. (1857a). An account of the ancient and ruined city of Brahminabad, in Sind. Journal of the Bombay Branch of the Royal Asiatic Society 5,413 -425.
____. (1857b). Further observations of the ruined city of Brahminabad, in Sind. Journal of the Bombay Branch of the Royal Asiatic Society 5,467 -477.
Bendick, R., R. Bilham, E. Fielding, V. Gaur, S. E. Hough, G. Kier, M. Kulkarni, S. Martin, and M. Mukol (2001). TheJanuary 26, 2001 Bhuj, India earthquake. Seismological Research Letters 72,328 -335.[GeoRef]
Bilham, R. (1998). Slip parameters for the Rann of Kachchh, 16 June 1819 earthquake, quantified from historical accounts. Geological Society of London Special Publication 146,295 -319.
Burnes, A. (1833). A memoir on the Eastern Branch of the River Indus giving an account of the alterations produced to it by an earthquake in 1819, also a theory of the Runn, and some conjectures on the Route of Alexander the Great, drawn up in the years 1827-28. Journal of the Royal Asiatic Society of Bengal, 3, 550-588.
Burnes, A. (1834). Travels into Bokhara. Vol. 3.
Burrard, S. G. (1890). In Account of the Operations of the Great Trigonometrical Survey of India, ed. W. H. Cole, V, Cutch Coast Series, ii-xii, 14, Principal Triangulation S. W. Quadrilateral. Dehra Dun: Office of the Trigonometrical Survey of India.
Byrne, D. E., L. R. Sykes, and D. M. Davis (1992). Great thrust earthquakes and aseismic slip along the plate boundary of the Makran subduction zone. Journal of Geophysical Research 97,449 -478.[GeoRef]
Cousens, H. (1905). Conservation of ancient monuments in the Bombay Presidency. Journal of the Bombay Branch of the Royal Asiatic Society, extra number, centenary memorial volume,149 -162.
____. (1929) The Antiquities of Sind. Archaeological Survey of India, 46, Imperial Series, Oxford University Press, Karachi, reprint 1975. 184 pp.
Cunningham, A. (1871). The Ancient Geography of India. London: Trübner & Co.,589 pp.
Delhoste, E. (1832). Observations on Sind. Institute of Sindhology publication reprint 129 (1987),118 pp.
Elliot, H. M. (1857). The History of India as Told by Its Own Historians: The Muhammadan Period. 8 vols., 1867-1877, ed. by J. Dowson. 3rd ed. London: Trübner. reprinted AMS Press, New York, 1966.
Griesbach, C. L. (1893). Notes on the earthquake of Baluchistan on the 20th December 1892. Records of the Geological Survey of India 26, 57-64.
Giardini, D., G. Grunthal, K. Shedlock, and P. Zheng (1999). The GSHAP Global Seismic Hazard Map. Annali di Geofisica 42,1,225 -1,230.[GeoRef]
Gole, S. (1988). Maps of Mughal India, Drawn by Colonel Jean-Baptiste Gentil, Agent for the French Government to the Court of Shuja ud-Daula at Faizabad, in 1770. New Delhi: Manohar.
Grant, C. W. (1837). Memoir to illustrate a Geological Map of Cutch, Transactions of the Geological Society of London Series II, 5 25,289 -389.
Grindlay, R. M. (1808). Diary reproduced in Travels into Bokhara, ed. A. Burnes, appendix to vol. 3, 1834.
Habib, I. (1982). An Atlas of the Mughal Empire: Political and Economic Maps with Detailed Notes, Bibliography, and Index. New York: Oxford Univ. Press, 105 pp.
Haig, M. R. (1874). On the sites of Brahmanabad and Mansurah in Sindh; with others of less note in their vicinity. Journal of the Royal Asiatic Society of Bengal, 16,281 -294.
____. (1894) The Indus Delta Country; A Memoir, Chiefly on Its Ancient Geography and History. London: Kegan Paul, Trench & Trübner, 148 pp.
Heron, A. (1911). The Baluchistan earthquake of the21 st October 1909. Records of the Geological Society of India 41,22 -35.
Hodivala, S. H. (1939). Studies in Indo-Muslim History: A Critical Commentary on Elliot and Dowson's "History of India as Told by Its Own Historians." Vol. 1. Bombay: Kokil and Co.,712 pp.
Hough, S. E., S. Martin, R. Bilham, and G. M. Atkinson
(2002). The 26 January 2001 M
7.6 Bhuj, India, Earthquake: Observed and Predicted Ground Motions.
Bulletin of the Seismological Society of America
922,061
-2,079.
Jade, S. D., and 12 others (2003). Pre-seismic, co-seismic and post seismic displacements associated with the Bhuj earthquake derived from recent and historic geodetic data. Journal of the Indian Institute of Science 112 (3), 331-346.
Jarrett, H. S. (1891). The A'in-i Akbari by Abu'l-Fazl Allami. Trans. Col. H. S. Jarrett, Asiatic Soc. Bengal. Vol. 2. Calcutta: Baptist Mission Press.
Kukowski, N., T. Schillhorn, E. R. Flueh, and K. Huhn
(2000). Newly identified strike-slip plate boundary in the
northeastern Arabian Sea. Geology
28, 355-358
Lambrick, H. T. (1964). Sind: A General Introduction. Hyderabad: Sindhi Ababi Board, 274 pps.
Lawrence, R. D., S. H. Khan, and T. Nakata (1992). Chaman fault, Pakistan-Afghanistan. In Major Active Faults of the World—Results of IGCP Project 206. Special issue supplement to Annales Tectonicae 6, ed. R. C. Bucknam and P. L. Hancock, 196-223.
McMurdo, J. (1823). Papers related to the earthquake which occurred in India in 1819. Transactions of the Literary Society of Bombay 3,90 -116.
____. (1834a). Dissertation on the river Indus, By the late Captain James McMurdo, of the Honorouble East India Company's Military Service on the Bombay Estabishment, Communicated by Lt. Col. Edward Frederick, 20-44, Journal of Bombay Asiatic Society,223 -257
____. (1834b). An account of the country of Sindh, with remarks on the state of Society, the Government, manners and customs of the people, by the late Captain James McMurdo, of the Bombay military establishment, communicated by James Bird, Journal of Bombay Asiatic Society, 223-257
Mallet, R. (1852). Report of the twenty-first meeting of the British Association for the Advancement of Science, Ipswich, 1851. Second report on the facts of earthquake phenomena,272 -320.
Merewether, W. (1852). A report of the disastrous consequences of the severe earthquake felt in the provinces of Upper Scinde on the 24th January 1852. Journal of the Bombay Geographical Society 10,284 -286
Montessus de Ballore, F. Comte de (1924). La Geologie Sismologique, Lib. A. Colin, Paris. p.488 .
Nakata, T., H. Tsutsumi, S. H. Khan, and R. D. Lawrence (1991). Active Faults of Pakistan; Map sheets and inventories, Sp. Pub. 21, Res. Center for Regional Geography, Hiroshima University, pp. 141.
Oldham, R. D. (1898). A note on the Allah Bund in the north west of the Runn of Cuch. Memoirs of the Geological Survey of India 28,27 -30.
____. (1926). The Cutch (Kachh) Earthquake of 16th June 1819 with a revision of the great earthquake of 12thJune 1897. Memoirs of the Geological Survey of India 46,71 -146.
Oldham, T. and R. D. Oldham (1883). Catalogue of Indian earthquakes. Memoirs of the Geological Survey of India 19,163 -215, Calcutta.
Pendse, C. G. (1948). The Makran earthquake of the28th of November, 1945. Science Notes 10,141 -145, Indian Meteorological Dept.
Quittmeyer, R. C., and K. H. Jacob (1979). Historical
and modern seismicity of Pakistan, Afghanistan, northwestern India, and
southeastern Iran. Bulletin of the Seismological Society of
America 69 (3),773
-823.
Rajendran, C. P., and K. Rajendran (2001).
Characteristics of deformation and past seismicity associated with the 1819
Kutch earthquake, northwestern India. Bulletin of the Seismological
Society of America
91(3),407
-426.
Raverty, H. G. (1892). The Mihran of Sind, and its tributaries. Journal of the Asiatic Society of Bengal 61,156 -508.
Reid, H. F. (1911). The elastic rebound theory of earthquakes. Bulletin of the Department of Geology (1910-1911, Univ. of California, Berkeley), ed. A. C. Lawson and J. C. Merriam, 413-444.
Sarkar, J. (1947). Maasir-i-Alamgiri—An English Translation of Saqi Must'ad Khan's History of the Emperor Aurangzib. Lahore: Suhail Academy reprint, 1981, 350 pp.
____. (1978). The A'in-i Akbari by Abu'l-Fazl Allami, Translated from the original Persian by Col. H. S. Jarrett, corrected and further annotated by Jadunnath Sarkar, Second edition. Oriental Books Reprint Corp. New Delhi, 1978.
Sawar, G. (2004). Earthquakes and the neo-tectonic framework of the Kutch-Hyderabad-Karachi triple junction area, Indo-Pakistan. Pakistan Journal of Hydrocarbon Research 14, 35-40.
Schelling, D. D. (1999). Frontal structural geometries and detachment tectonics of the northeastern Karachi arc, southern Kirthar range, Pakistan. Geological Society of America Bulletin 328,287 -302.
Stein, S., G. Sella, and E. A. Okal (2002). TheJanuary 26, 2001 Bhuj earthquake and the diffuse boundary of the Indian plate. American Geophysical Union Geodynamics Monographs 30,243 -254.
Szeliga, W. M., M. Furuya, S. Satyabala, and R. Bilham (2006). Surface creep along the Chaman fault on the Pakistan-Afghanistan border imaged by SAR interferometry (abstract). EOS, Transactions, American Geophysical Union 87, p. 52.
Tuttle, M. P., J. Hengesh, K. B. Tucker, W. Lettis, S. L. Deaton, and J. D. Frost (2002). Observations and comparisons of liquefaction features and related effects induced by the Bhuj earthquake. Earthquake Spectra 18,supplement A , 79-100.
Wallace, K., R. Bilham, F. Blume, V. K. Gaur, and V. Gahalaut (2006). Geodetic constraints on the Bhuj 2001 earthquake and surface deformation in the Kachchh rift basin. Geophysical Research Letters 33,10.1029/2006GL025775 .
Walton, H. I. (1864). Correspondence, 27August 1864. Proceedings of the Geographical Society of Bombay 17,cxxv .
West, W. D. (1934). The Baluchistan earthquakes of25th and 27th August 1931. Memoirs of the Geological Survey of India 67,1 -82.
____. (1936). Preliminary geological report on the Baluchistan (Quetta) earthquake of May 31st, 1935. Records of the Geological Survey of India 69,204 -241.
Wells, D. L., and K. J. Coppersmith (1994). New
empirical relationships among magnitude, rupture length, rupture width,
rupture area, and surface displacement. Bulletin of the
Seismological Society of America
84,974
-1,002.
Wynne, A. B. (1872). Memoir on the geology of Kutch. Memoirs of the Geological Survey of India 9, 1-294.
Yeats, R. S., R. D. Lawrence, S. Jamil-Ud-Din, and S. H. Khan (1979). Surface effects of the 16 March 1978 earthquake, Pakistan-Afghanistan border. In Geodynamics of Pakistan, ed. A. Farah and K. A. De Jong, 359-361. Quetta, Pakistan: Geological Survey of Pakistan.
Yule, H., and A. C. Burnell [1903] 1995. Hobson-Jobson, A Glossary of Colloquial Anglo-Indian Words and Phrases, and of Kindred Terms, Etymological, Historical, Geographical and Discursive. Ed. W. Crooke. Madras: Asian Educational Services (reprint 1995), 1,021 pp.
Yule, H, and R. Maclagan (1882). Memoir of General Sir William Erskine Baker, KCB, Royal Engineers (Bengal) compiled by two old friends, brother officers and pupils. London: Printed for private circulation, 67 pps.
Cooperative Institute for Research in Environmental Sciences and the Department of Geological Sciences
University of Colorado at Boulder
Boulder, Colorado 80309-0399
USA
roger.bilham{at}colorado.edu
(R.B.)
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| JOURNAL HOME | HELP | CONTACT PUBLISHER | SUBSCRIBE | ARCHIVE | SEARCH | TABLE OF CONTENTS |
