Holocene shortening rates and seismic hazard assessment for the frontal Potwar Plateau, NW Himalaya of Pakistan: Insights from ¹⁰Be concentrations on fluvial terraces of the Mahesian Anticline

Abstract

We present the results of a structural neotectonic survey undertaken on the Mahesian Anticline in the frontal Himalaya of Pakistan. This anticline resulted from the folding of Precambrian to Tertiary layers that was controlled by a thrust and a backthrust, interacting in a complex way. Four generations of fluvial terraces formed by the Jhelum River and two tributaries have been distinguished on the SE flank of the anticline. Two of these terraces, T2 and T3, have been left hanging by fold development and have been dated by cosmogenic ¹⁰Be to 6.5 ± 0.2 ka and 3.3 ± 0.7 ka, respectively. From such ages an uplift rate of ∼10 mm/y was determined for the Holocene. That uplift is induced by a shortening rate of ∼10 mm/y. We highlight that the Mahesian Anticline and Frontal Salt Range Thrust, together with the Kalabagh western lateral ramp and the Jhelum eastern lateral ramp, delineate the active tectonic boundary of the Potwar Plateau. This thrust sheet moves above the salt level without out-of-sequence deformation. Moreover, the small but significant difference between the long-term deformation rates (8.4 mm/y) and the geodetic velocities (2–5 mm/y) detected for the Central Potwar Plateau seems to be linked to episodic spurts of accelerated creeping of the thick salt level, triggered by earthquakes located to the north on the deep (>15 km) part of the MHT. In addition, the large difference between the long-term deformation rates and the geodetic velocities (less than 2 mm/y) reported for the eastern Potwar Plateau seems to be linked to the accumulation of a slip deficit around asperities formed where the salt is missing. This deficit may be recovered during earthquakes potentially as great as Mw 7.

Introduction

The Himalaya of Pakistan is characterized by regional-scale plateaus that mainly slip aseismically over the Indian Plate, as suggested from the lack of historical high-magnitude earthquakes in the area (e.g. Seeber and Armbruster, 1981, Fig. 1a and b). The largest of these plateaus is the Potwar Plateau, which bounds the western flank of the Hazara-Kashmir Syntax (e.g. Grelaud et al., 2002, Fig. 1a and b). This plateau is currently moving southwards above a late Precambrian–Early Cambrian salt level (Crawford, 1974) that partially defines the main detachment in the area: The Main Himalayan Thrust (MHT) (Schelling and Arita, 1991). It is indeed this salt unit which mostly precludes the generation of earthquakes in the area, unlike what occurs in most of the Himalaya frontal zones (e.g. Seeber and Armbruster, 1981). By combining paleomagnetic data and geological observations, Jaumé and Lillie (1988) propose that the northern Potwar Plateau deformed as a steeply tapered thrust wedge (sensu Chapple (1978), with an angle of 3.5°–5.5°, until ca. 2 Ma. Between 2 Ma and the Present, the propagating to the south main detachment reached the evaporates level, which has pushed the Potwar Plateau without frictional deformation. Then, erosion has reduced the initial steep slope of the plateau to its present gentle slope. Cotton and Koyi (2000), based on scaled sandbox models, simulate the evolution of the Potwar Plateau and surrounding areas above adjacent frictional and ductile substrates. Considering variations in the original thickness of the ductile salt level and the influence of the prekinematic and synkinematic style of the wedge above the main detachment, they establish that forward-vergent thrusts producing steep wedges develop above a frictional substrate. On the other hand, they suggest that both forward and backward vergent thrusts conducting low taper wedges occur over ductile substrates. Further, above a ductile level, deformation propagates farther and faster than above a frictional level; this differential rates produce a lateral inflection zone sub-parallel to the shortening direction. Within the wedge above the salt horizon, a frontal inflection and general folding is promoted, generating diapiric structures. These features are like those dominating the Salt Range - Potwar Plateau (e.g. Leathers, 1987).

Geodetic velocities indicate that the central part of the Potwar Plateau is currently slipping to the S–SE at rates of 3–5 mm/y (e.g. Jouanne et al., 2014, Fig. 1b). These velocities are lower than those estimated from geological evidence for the Late Cenozoic, which are ∼8.4 mm/y for the last 12 to 2 My (e.g. Baker et al., 1988, Mcdougall and Khan, 1990, Fig. 2a, Table Ai-ii in supplementary materials). No shortening rates have been estimated for the Potwar Plateau on the thousands of years' time-scale and this lack of information is a major obstacle when discussing the significance of the geodetic velocities in terms of how strain is being accommodated along the salt detachment of the Main Himalayan Thrust (MHT). To fill this gap, overriding active structures, such as those located close to the syntax in the frontal Himalaya of Pakistan (Fig. 1a and b), are appropriate sites for estimating Holocene deformation rates. There, the structural pattern is represented by sub-parallel fault propagation folds related to blind thrusts that accommodate convergence-induced shortening (e.g. Leathers, 1987, Fig. 2b). The most clearly expressed active structure in this area is the Mahesian Anticline (e.g. Nakata et al., 1991, Fig. 3a–c), located in a zone where the salt is ∼0.2 km thick (Leathers, 1987, Fig. 1b) and where GPS velocities (India Fixed Reference Frame) are almost zero (<2 mm/y, PK34 station; Jouanne et al., 2014, Fig. 3a). This anticline is bounded to the east by the Jhelum River, which has carved fluvial terraces lying unconformably over Tertiary formations (Siwalik Group; Fig. 4a–e). Some of these terraces have been uplifted by blind thrusts and backthrusts (e.g. Yeats and Lillie, 1991) and are preserved on the E–SE flank of this fold (Fig. 5a–c).

In this contribution, we present the results of a structural and neotectonic survey devised to characterize the structure of the Mahesian Anticline and quantify the Holocene deformation of its SE flank. First, we collected structural data about the orientation of the Tertiary layers of the SE limb of this anticline. Comparing these data with seismic profiles (Leathers, 1987) yields a model of how the leading thrust and backthrust have controlled the structure of the Mahesian Anticline. Second, we mapped four generations of fluvial terraces on its SE flank. Quartz clasts were collected from two levels of terraces for ¹⁰Be analysis in order to obtain their exposure ages. We calculate uplift and shortening rates from these ages for a kink-like geometry fold (e.g. Suppe, 1983) which is what we interpret the Mahesian Anticline to be. We compare our thousands of years' time-scale shortening rates with those obtained from geodetic methods in the surrounding area (e.g. Jouanne et al., 2014) and their difference is discussed in terms of the thickness of the salt detachment and its mechanical (creeping and/or stick slip) behavior. We finally propose a conceptual model of how the Mahesian Anticline has grown during the Holocene, as a result of the processes occurring at depth along the MHT. Our results and interpretations seek to evaluate the seismic risk in a region where the ninth largest dam in the world (the Mangla Dam (Fig. 1a and b)) is being constructed.