SEISMOTECTONIC MODELING OF NW HIMALAYA: A PERSPECTIVE ON FUTURE SEISMIC HAZARD

Abstract

The NW Himalaya and adjoining region of the part of Alpide-Himalayan belt is one of the most complex and peculiar tectonic regions in the world, which has experienced several large to great earthquakes in the past. The purposes of present investigations are to prepare a homogeneous and complete seismicity database for the study area bounded by 25 - 40 N and 65° - 85° E which include Himalayan frontal arc, Hindukush, Pamirs, Tien Shan, Kunlun, Karakoram, Afghanistan and Pakistan region; analysis of spatiotemporal distribution of seismic activity; long-term earthquake prediction; seismic hazard analysis and Coulomb stress modeling for moderate to large earthquakes. The whole study region has been divided into three distinct parts, namely, zone-I includes Hindukush-Pamir-Tien-Shan region, zone-II includes Pakistan region and zone-Ill includes Himalayan frontal arc along with Tibetan Plateau. A homogeneous and complete seismicity database has been prepared for the period 1853 to Oct. 10, 2005 with the help of all existing earthquake catalogues and lists pertaining to the region. The prepared seismicity database is homogenized for surface wave magnitude (Ms) with the help of various empirical relationships developed among different magnitude scales (Ms, mb and Mw). The cut-off magnitude (threshold magnitude or magnitude of completeness, Mc) for this seismicity database is estimated as Ms 4.0. However, the spatial distribution of Mc shows its variation from 4.0 to 4.8 in while region. The completeness periods for different magnitude ranges are also estimated and it is observed that earthquakes falling in the magnitude range of 4.0 - 4.4, 4.5 - 4.9, 5.0 - 5.4, 5.5 - 5.9, 6.0 - 6.4 and > 6.5 are complete since 1965, 1960, 1955, 1920, 1915 and 1882, respectively. Further, this seismicity database is used for various approaches of the seismological investigations: such as spatiotemporal distribution of earthquake occurrences, b-value mapping, focal depth distribution and delineation of possible seismogenic source zones. On the basis of vertical distribution of cumulative number of earthquakes, three major seismogenic layers are observed in zone-I at depth range 0 - 40, 41 - 200 and > 200 km, whereas only two n seismogenic layers are observed in zone-II and zone-ETI at depth range 0-40 and > 40 km, suggesting the similar tectonic features beneath these two zones. The frequency-magnitude distributions (G-R relationships) of earthquakes (total, shallow and intermediate) show that the estimated 'a' and b values are observed as 7.90±0.09 and 1.03±0.02, respectively in the study region. The depth distribution of b-value shows a continuous increase up to depth 40 km in zone-I suggesting less stress build up and increase and decrease patterns below 40 km suggest the crustal heterogeneity. The similar characteristics of b-value with depth are also observed in zone-II and zone-Ill which suggest the similar crustal properties in these two zones as evidence by vertical distribution of cumulative number of earthquakes. In the spatial mapping of b-value, high values are observed in the eastern and western part of Pamir, Afghanistan region and Kashmir Himalaya along with Kunlun belt, which suggest the low stress build up. Low b-value in Hindukush region, southeast of Hindukush, Caucasus, Sulaiman-Kirthar ranges in Pakistan and a large part of south east of Himalayan frontal arc region with part of Tibetan Plateau suggest high stress regime which causes frequent occurrence of earthquakes in these regions. High b-value in the subducting slab of Indian plate beneath Hindukush-Parnir region clearly indicates a phase transformation of material. On the basis of spatial distribution of earthquakes during the period of 1853 to 1962, a total 5 seismogenic source zones have been delineated whereas total 10 seismogenic source zones have been identified on the basis of earthquakes data during the period 1963 to Oct. 10, 2005. The spatial distribution of earthquakes Ms > 7.0 in zone-II shows the migration of these earthquakes in the direction of northeast to southwest before the occurrence of Quetta earthquake whereas the migration of such earthquakes in zone-Hi was in opposite direction from northwest to southeast along the Himalayan frontal arc after the occurrence of great Kangra earthquake. The focal depth distribution of seismic activity in zone-I indicates a 'V shape structure, which is responsible to generate intermediate depth earthquakes in this zone and maximum seismicity is located on either side of the axis of this structure. The shallow depth earthquakes are observed in zone-II and zone-Ill along with very few intermediate depth earthquakes. A total, 17 possible seismogenic source zones have been delineated on the iii basis of seismotectonics and geomorphological features of the region studied above and used for the assessment of seismic hazard and long - term earthquake prediction. A regional time and magnitude predictable model has been used for the quantitative assessment of probabilities of occurrences and magnitude of future large earthquakes (Ms > 5.5) in the 17 seismogenic source zones. It is observed that the seismogenic source zone 7 which includes Hindukush region exhibits most hazardous among the all seismogenic source zones with expected magnitude of 6.6 having occurrence probability 0.86 in next 20 years. The seismogenic source zone 5 which includes the part of Herat fault in Afghanistan shows a low probability in comparison to other zones suggest, it presently inactive (less seismically active) source zone. The seismogenic source zone 16 which includes Uttarkashi and Chamoli region shows the highest probability 0.81 in the Himalayan frontal arc having expected earthquake of magnitude 6.8 in next 20 years. The seismic hazard has been also studied in three major zones to check the potentiahty of these regions for the future earthquake occurrences. The results indicate that zone-I has low mean return periods for the same earthquakes magnitude in comparison to the zone-II and zone-Ill suggesting that zone-I is more hazardous than others two. The Coulomb stress modeling has been studied to perceive the earthquake interaction processes for moderate to large earthquakes (Mw > 6.5) with their subsequent smaller earthquakes. The regions of increased and decreased stress have been identified and it is observed that the distribution of regional seismicity (or aftershocks) fall in the increased Coulomb stress region suggesting the earthquake triggering. This study reveals that the occurrence of majority of aftershocks for the Himalayan frontal arc earthquakes (Uttarkashi, Chamoli) is due to the high Coulomb stress at the updip edge offault rupture. Thus, it may be possible to predict the regions of future earthquake occurrences and aftershocks area.