The 2005 catastrophic Mw7.6 earthquake in Kashmir (death toll 85,000) was one of several earthquakes that were anticipated by seismologists because of the known seismic hazard of the Himalaya where similar damaging earthquakes have occurred throughout the past 1000 years. The earthquakes there are caused by the northward movement of India towards Asia at a rate of 2 inches/year. For the same reasons earthquakes also occur along the western and eastern edges of the Indian plate. The largest in the east was the Mw9.1 Sumatra/Andaman earthquake of 2004 (death toll 300,000). No great earthquakes have occurred recently on the western edge of the Indian plate in Pakistan, yet there are good reasons to suppose that they should. An earthquake in SW Pakistan could cripple Karachi, Pakistan's largest city with >12 million people, just as Quetta was destroyed in 1935 by an Mw7.7 earthquake (death toll 35,000). The destruction of Karachi would be catastrophic to the economy of Pakistan. Like Los Angeles, Karachi is a city that was a fishing village 250 years ago and has only in the past 50 years achieved megacity status. Unlike Los Angeles no earthquake resistance is in place, and no history of damaging earthquakes has survived. One aim of the project is to find out whether the historical absence of earthquakes is because no earthquakes occur, or whether, as we suspect, it is because they recur at intervals too long to be recorded by its 200 year colonial history.
The edge of the Indian plate, which passes from the coast west of Karachi in the south, to the mountains that separate Afghanistan from Pakistan east of Kabul in the north, could be locked, preparing for a future earthquake, it could be sliding aseismically (like parts of central California, or it could be a wide region of rotating geological blocks, that act like ball-bearings permitting the plates to slide past each other. We plan a series of Global Positioning System (GPS) measurements that will uniquely distinguish between these three possibilities. The measurements will tell us the relative velocities at the edge of the plate, and the width of the zone of deformation, and the depth at which the two plates are locked - information that will be used to quantify the seismic productivity of the region, and identify the most probable future location of future damaging earthquakes. This information is of fundamental importance to earthquake engineers in the region, and will dictate building policies for the next 30 years. The measurements will also provide important constraints on the physics and nature of the stresses that permit continental transform plate boundaries, like the San Andreas fault, to slip and evolve with time. The field investigations are being undertaken in close collaboration with scientists and engineers from three Pakistan, and two Indian Universities, and as such forge international ties that are of great benefit to the image of the United States in these developing nations. Because of the international nature of the project, this work is jointly funded by the Geophysics Program and the Office of International Science and Engineering.
In this five year study we measured and remeasured more than 80 GPS in Pakistan and India to determine the velocity and locus of deformation associated with the Indian plate plunging into southern Asia. The velocity of slip is determined to be 3 cm/yr , and deformation is distributed over a wide region that varies with latitude. The details of this distributed deformation are of great importance for seismic hazard studies and the characterization of scenarios for future earthquakes in the region. Strain fields and rates associated with earthquakes are summarized in the form of arrows in the figure. In the NW corner of this collision process we discovered that convergence with Tibet in the Kashmir Himalaya is relatively slow (12 mm/yr) although future earthquakes here range in magnitude from Mw=7.5 to Mw=9.0. Of the 33 possible earthquake scenarios we examined in the Indian segment of Kashmir, the most probable are assessed to approximate Mw=8.0 and to occur at intervals of >500 years. We examined the consequences of an earthquake reported in chronicles to have occurred in 883 AD, that flooded the Kashmir valley and caused damage to many of its temples. At latitudes near Quetta we examined two classes of deformation. A sequence of earthquakes in 2008 near Pishin NE of Quetta occurred on a zone that GPS studies reveal to be deforming in a shear sense between Pishin and Sibi at approximately 11mm/yr. Instead of a simple fault releasing this slip we discovered that the region is deforming on a series of short (20 km long) bookshelf faults which have the serendipitous advantage that the maximum earthquakes in this region cannot much exceed Mw=6. To the southwest of Kashmir we determined the Potwar plateau to be creeping at 3 mm/yr above a layer of salt (hitherto it was believed to moving at 11 mm/yr). In places the Kohat salt layer is so thin that the plateau has grounded, and it moves intermittently in rare Mw=6.0 earthquakes. To the south of Quetta we find shear strain is focused and released by faults between the villages of Chaman and Kalat, and compression in the mountains SE of Quetta is focused on thrust faults bordering the edge of the Indus River plain in Sindh province. We quantified the processes in a sequence of Mw>7 earthquakes that occurred between 1930 and 1935. Understanding the tectonic processes at the southern end of the great transform fault that permits India’s indentation into Asia is of particular importance because it lies close to Pakistan’s largest city, Karachi (pop. 15 M). NE of Karachi, near Hyderabad we discovered evidence for two historically damaging earthquakes in 995AD and 1668AD that we infer to be associated with our measured eastward compression of a north-south Khirtar at 2-3 mm/yr. We suspect earthquake here may range in magnitude from Mw=6 to Mw=7 at several hundred year intervals. Our GPS measurements revealed that deformation in Balochistan and Makran to the west and NW of Karachi occurs at rates of up to 2.5 cm year. In September 2013 a Mw=7.7 earthquake here revealed that SE Makran is probably rotating counterclockwise and that this rotation may explain the curious offset of the Chaman transform system eastward toward Karachi. The rotation of the SE Makran reduces the slip rate on the southernmost fault here (the Ornach Nal fault) thereby reducing seismic hazards near the city of Karachi. The results of our findings are described in detail in some two dozen articles in academic publications. They were undertaken collaboratively with local scientists who are now undertaking more detailed studies in the region characterizing seismic risks to critical facilities and infrastructure in Pakistan. Two graduate students completed their PhD’s in the course of the work, and we gave two training programs to educate local scientists and students in the methods we used. The 20 GPS receivers used in the project continue to be used by our collaborators.
