New insights into the structural heterogeneity and geodynamics of the Indo-Burma subduction zone from ambient noise tomography

Highlights

• We propose a new 3D Vs model for the crust and uppermost mantle in the Myanmar region.

• The forearc mantle of Indo-Burma subduction zone is at least 19% serpentinized.

• We identify a northward reduction of water input into the Indo-Burma mantle wedge.

• Localized accretion is observed beneath the northern Indo-Burman Ranges.

• Basaltic intrusions are found in the middle/lower crust beneath the Sagaing Fault.

Abstract

The absence of fine lithospheric-scale velocity models beneath Myanmar makes it difficult to understand the neotectonics and geodynamics along the Indo-Burma subduction zone. In this study, we present a high-resolution crustal and uppermost mantle 3-D shear-wave velocity (Vs) model of Myanmar to fill this knowledge gap, using ambient noise data from newly deployed seismic arrays. In the upper crust, our model reveals two thick (>10 km) N-S-elongated basins between the Indo-Burman Ranges and the Central Myanmar Basin. At middle to lower crustal depths, low velocities dominate the Indo-Burman Ranges, especially in its northern part where Vs is observed to be as low as ∼3.2 km/s in the lower crust. This feature is interpreted as sediments that were deposited west of the ranges and have since been subducted northeastward and accreted onto the overriding plate. Furthermore, our model reveals an N-S trending high-velocity anomaly beneath the Sagaing Fault, which could be explained by solidified basaltic magma that intruded upwards from the mantle where a low-velocity anomaly is imaged. In the upper mantle, the subducting Indian Plate is clearly imaged beneath Myanmar as an east-dipping high-velocity zone, overlain by a prominent wedge-shaped low-velocity body (Vs < 4.3 km/s). We interpret this low-velocity anomaly to represent partial serpentinization (19–38%) in the forearc mantle. The size and amplitude of this anomaly decrease towards the north, suggesting a northward reduction in serpentinization level within the forearc mantle, possibly related to a northward reduction of water in the subduction zone. This could be associated with lower water content in the subducting plate, as the thick sediments deposited in the north may have driven water out of the lowermost section, while the upper sedimentary section, which could still have been carrying water, would have been scraped off of the downgoing plate and accreted onto the overriding plate, forming part of the Indo-Burman Range.

Introduction

The Indo-Burma subduction zone connects the Eastern Himalayan Syntaxis in the north and the Andaman Sea in the south (Fig. 1a). This oblique margin represents the eastern section of the collision between the Indian and Eurasian Plates. The NNE motion of the Indian Plate relative to the Sunda Plate, with a modern velocity of ∼35 mm yr⁻¹ or larger (Mallick et al., 2019; Steckler et al., 2016), has resulted in a series of N-S elongate structures in the overriding plate, including the Indo-Burman Ranges (IBR) and associated folds, together with the central Myanmar basin (CMB). This region is bounded by the Shan plateau (SP) to the east (Fig. 1a). The 1200-km-long N-S oriented dextral strike-slip Sagaing Fault (SF) cuts through the eastern side of the CMB (Fig. 1b).

The present-day tectonics and geodynamics along the Indo-Burman convergent margin are mainly controlled by the oblique subduction of the Indian Plate. Global Positioning System (GPS) observations show slip partitioning in the overlying Burma Plate, with about half of the northwards motion absorbed by the SF (∼20 mm yr⁻¹), and the rest either accommodated by other right-lateral faults and deformation within the IBR and CMB, and/or taken up by the Rakhine-Bangladesh megathrust (e.g., Mallick et al., 2019; Steckler et al., 2016). The CMB is sub-divided by the Wuntho-Popa volcanic arc (WPA), which consists of a line of sparsely distributed volcanoes (Morley et al., 2020) (Fig. 1a). Sediments within the CMB, mostly Cenozoic, are up to ∼18 km thick, as observed by oil prospecting seismic reflection profiles (Pivnik et al., 1998). To the west of the CMB, the IBR represents an accretionary prism comprising thick Mesozoic and Cenozoic flysch deposits with some embedded Mesozoic ophiolites (Morley et al., 2020). This broad prism (over 300 km in E-W direction at the surface) exhibits a low surface slope (∼0.1° to 0.5°) in its westernmost extent (Steckler et al., 2016), suggesting that the décollement underlying the wedge in this region, which sits within the 20+ km thick sedimentary stack, must be weak (Betka et al., 2018). The present-day tectonic mosaic of this region records a complex evolutionary history comprising several stages of Tethyan subduction and a subsequent transition into an oblique subduction environment (Westerweel et al., 2019). Since many features associated with this complex tectonic system are not exposed, imaging the crust and upper mantle velocity structure is vital for unveiling the composition and geometry of geologic units in this tectonic system, and subsequently improving our understanding of the geologic and geodynamic evolution of this margin.

Study of the Indo-Burma subduction zone began decades ago, with the east-dipping subducted Indian Plate clearly delineated beneath Myanmar by earthquake hypocenters (discovery of the Wadati-Benioff zone) and gravity surveys (e.g., Mukhopadhyay and Dasgupta, 1988). Although there is still a dispute as to whether this subduction remains active, more recent studies tend to support ongoing eastward subduction (Belousov et al., 2018; Steckler et al., 2016), with a GPS-inferred E-W (arc-perpendicular) convergence rate ranging from 12 to 24 mm yr⁻¹ (Mallick et al., 2019). Nevertheless, fundamental questions remain regarding the subduction dynamics and associated tectonic processes, such as slab/sediment dehydration and volcanic processes (Lee et al., 2016), the detailed geometry of the subducting (or subducted) slab (Li et al., 2008; Yao et al., 2021; Zheng et al., 2020), the influence of the thick sediment input to the subduction system (Betka et al., 2018), and how the subduction fits into Indo-Eurasian collision (Westerweel et al., 2019). The absence of a high-resolution seismic model for the Myanmar region is one of the biggest obstacles to solve these puzzles.

Early seismic tomographic studies in Myanmar were mainly based on sparsely distributed data with highest resolution of ∼200 km (e.g., Li et al., 2008). Structures at shallower depths, including the crust and uppermost mantle, were beyond their resolution limit. Recently, Wang et al. (2019) made the first attempt to build a 3-D shear-wave velocity (Vs) model of Myanmar by jointly inverting receiver functions, teleseismic Rayleigh wave phase velocities and H/V ratios, based on the data recorded at recently installed seismic stations. Although the resulting model reveals much more details than earlier studies, there remains room for improvement. For instance, the inversion strategy used in that study first resolves local Vs beneath each station, and then interpolates the station-based 1-D velocity profiles to produce a 3-D regional model. This strategy generally performs well beneath dense seismic arrays with nearly uniformly distributed stations but can only generate limited-resolution and sometimes biased seismic images when the station distribution is highly uneven, such as the case in Myanmar. In addition, Wang et al. (2019) only used teleseismic data at relative long period (Rayleigh wave phase velocity > 30 s and H/V ratio > 25 s), which have limited sensitivities to shallow (crustal) velocity structures. In this study, we construct an updated 3-D Vs model using 6–50 s surface wave phase velocities extracted from ambient noise cross-correlation, which are most sensitive to the crustal and uppermost mantle structures and can generate a smoothed 3-D regional Vs model. This newly generated 3-D Vs model provides vital constraints on the tectonics and geodynamics of the Indo-Burma subduction zone.