news > kashmir

Locating the Kashmir fault

COMET scientists have used satellite data to determine the exact location of the fault that caused Kashmir's recent earthquake. As well as contributing to our understanding of the earthquake cycle, the increasing speed with which a fault can be located may help authorities plan their emergency response in the wake of future earthquakes.

The powerful earthquake that struck on 8th October 2005 is located along the south of the Himalayan mountains, and caused major damage in the Pakistan-administered region of Kashmir (see figure 1). An earthquake in this area was not entirely unexpected. The region has a long history of large earthquakes, as it is where the two tectonic plates that have created the Himalayas collide (see figure 2).

The rigid Indian plate moves northeastward at about 55mm each year relative to undeforming parts of Asia far to the north. About 20mm per year of this motion is concentrated along the Himalayas. The fault that caused the recent earthquake is located along a line where the heights of the mountains rapidly increase (see figure 3). This is where the mountains are pushed higher as their southwest movement is blocked by the northwards movement of the Indian plate.

The exact location of the earthquake fault has been determined by COMET scientists using radar data from the European Space Agency's environmental satellite ENVISAT. The scientists have attempted to construct a radar interferogram to locate the fault, but for various technical reasons have not been successful to date. However, a by-product of these attempts is a map of displacements of the ground made using radar amplitude measurements. Although these estimates of the ground movements are not as accurate as those from a radar interferogram, they have enabled scientists to locate the fault to within a few hundred meters (see figure 4).

Overlaying the fault location over satellite images of the area shows that it occurred where the mountains start to increase rapidly in height (see figure 3). There is also evidence of the effects of previous earthquakes on the fault that are visible in the image, which was acquired before the October earthquake (see figure 5). Muzaffarabad, which suffered considerable damage, is located close to the north-western end of the fault.

One of the curious observations about previous large Himalayan earthquakes is that there does not appear to be any evidence of slip on faults breaking all the way to the surface (with the exception of secondary surface faulting in the great 1897 Assam earthquake). These earthquakes all occurred before the substantial developments in our knowledge of what happens during earthquakes that took place in the second of the twentieth century. So it is possible that this observation simply results from no-one looking in the right places at the time (the rapid erosion in these areas would have soon removed traces of any surface rupture). The information about the fault location obtained using the radar imagery will help the scientists who are currently in the area looking for evidence of ground ruptures.

The analysis of the satellite radar data was done within three weeks of the earthquake. The main delay of two weeks was waiting for ENVISAT to pass over the area again using the same SAR imaging modes that were used to image the area before the earthquake (35 days cycle). In future, when there are more radar satellites with shorter repeat cycles, it should be possible to produce fault locations within a day or two. In cases like the Kashmir earthquake, where land communications are disrupted, this would provide valuable information about where the damage is likely to have been greatest.

Figure 1: Satellite map of southern Asia (click an image for a larger version)

Figure 2: Locations and dates of historical earthquakes

Figure 3: 3D perspective of the Kashmir earthquake region

Figure 4: Range offsets for an ENVISAT pass

Figure 5: LANDSAT 7 image of the Muzaffarabad region

news > kashmir