Neogene‐Recent Reactivation of Pre‐Existing Faults in South‐Central Vietnam, With Implications for the Extrusion of Indochina

Vietnam contains complex faults coupled with a diffuse igneous province that has been active since the mid‐Miocene. However, existing fault maps demonstrate little consensus over the location of Neogene basalt flows and relative ages of mapped faults, which complicates interpretations of tectonic model for the evolution of Indochina. This paper identifies discrete tectonic blocks within Vietnam and aims to define the Neogene‐Recent tectonic setting and kinematics of south‐central Vietnam by analyzing the orientation, kinematics, and relative ages of faults across each block. Fault ages and relative timing are estimated using cross‐cutting relationships with dated basalt flows and between slickenside sets. Remote sensing results show distinct fault trends within individual blocks that are locally related to the orientations of the basement‐involved block‐bounding faults. Faults observed in the field indicate an early phase of dip‐slip motion and a later phase of strike‐slip motion, recording the rotation of blocks within a stress field. Faulting after the change in motion of the Red River Fault Zone at ∼16 Ma is inferred, as faults cross‐cut basalt flows as young as ∼0.6 Ma. Strike‐slip motion on block‐bounding faults is consistent with rotation and continuous extrusion of each block within south‐central Vietnam. The rotation of the blocks is attributed to the “continuum rubble” behavior of small crustal blocks influenced by upper mantle flow after the collision between India and Eurasia. We infer a robust lithospheric‐asthenospheric coupling in the extrusion model, which holds implications for other regions experiencing extrusion even in the absence of a free surface.


Introduction
The plate tectonic theory is a fundamental paradigm in the earth sciences, which has undergone modifications since its inception in the early 1960s; to accurately apply the plate tectonic framework to all settings, however, scientists must evaluate the most complex edge cases, such as collisional-adjacent regions like Indochina, Alaska, and Anatolia (Finzel et al., 2011;Redfield et al., 2007;Ridgeway & Flesch, 2007;Tapponnier et al., 1986).In such collisional settings, "extrusion" or "escape" tectonics is the process by which the collision of two tectonic terranes leads to the lateral escape of material formerly located between those terranes.Experimental studies have demonstrated that in addition to the impetus from an initial collision, the tectonic extrusion process also relies on a "free surface" for the escaping material to move toward (Tapponnier et al., 1982(Tapponnier et al., , 1986)).For example, extrusion has been invoked to explain the tectonic motion and evolution of Alaska, where the free surface is the Bering Sea, and of the Anatolian block, where the free surface is the eastern Mediterranean, such that extrusion is accommodated by extension in the Greek islands and subduction south of Crete (Finzel et al., 2011;Redfield et al., 2007;Ridgeway & Flesch, 2007;Tapponnier et al., 1986).In detail, however, the mechanisms behind extrusion tectonics remain poorly constrained, including the mechanism for accommodating continuous extrusion if a free surface becomes unavailable by collisions (e.g., for Indochina, where the presence of Borneo marks the removal of a free surface; Figure 1) or by other tectonic processes (e.g., and also for Indochina, the presence of the southern reach of the East Vietnam Sea/South China Sea [EVS/SCS; Table 1]).The topic of post-extrusion processes has been the subject of ongoing debate and remains an area of significant interest and scientific inquiry (Chen et al., 2017;Jolivet et al., 2018;Morley, 2002Morley, , 2016; V. V. Nguyen & Luong, 2019;Pubellier & Morley, 2014;Tapponnier et al., 1982Tapponnier et al., , 1986;;Taylor & Hayes, 1980).In this study, we thus aim to enhance our understanding of collision-adjacent tectonic deformation and extrusion tectonics exploring the consequences for extrusion following the removal of a free surface in Indochina.
We have selected south-central Vietnam, a part of the Indochina block, as our study area because of its location within the core of the larger Sundaland block (Figure 1), which has previously been described as experiencing homogeneous deformation as a rigid block with GPS-measured velocities between ∼6 and ∼10 mm/year and an absolute motion to the E-SE (e.g., Avouac & Tapponnier, 1993;Cardwell & Isacks, 1978;Curray, 1989;Fitch, 1972;Hall & Nichols, 2002;Hamilton, 1979;McCaffrey, 1991;Michel et al., 2001;Peltzer & Saucier, 1996;Simons et al., 2007;Tapponnier et al., 1982;D. T. Tran et al., 2013).Alternatively, some authors have suggested that Sundaland has not experienced homogeneous deformation as a rigid block (Clift & Figure 1.Map showing location of Indochina following India-Eurasia collision ∼45 Ma (modified after Leloup et al., 1995;Morley, 2007;Morley et al., 2001;Peltzer & Tapponnier, 1988;Phach & Anh, 2018;Tapponnier et al., 1990; J. E. Zhang et al., 2011).This tectonic map of East and Southeast Asia shows red lines with sawtooth patterns indicating active subduction zones, while the thin red lines are extrusion-related sinistral strike-slip faults, including the Red River Fault Zone (RRFZ), the Wang Chao Fault Zone (WCFZ), and the Three Pagodas Shear Zone (TPSZ).Darker blue shades indicate the locations of marginal sea basins such as the East Vietnam/South China Sea (EVS/SCS).The Sundaland Block is outlined by the black dashed line and modified from Meltzner et al. (2017), and the study area is outlined by the yellow dashed line.Wilson, 2023;Hall & Morley, 2004).Regardless, the stress field across the Sundaland block is heterogeneous rather than subparallel to the absolute motion vector (V.V. Nguyen & Luong, 2019;Tingay et al., 2010), suggesting that the question of whether this region can best be described in terms of block tectonics (Calais et al., 2006) or a continuous deformation field (Jade et al., 2004) is unresolved.In the case of Indochina, the block tectonics hypothesis of Calais et al. (2006) is potentially compatible with a continued extrusion-driven origin for Neogene-Recent deformation in the region, while a continuous deformation field hypothesis, similar to deformation fields observed in regions like the Tibetan Plateau (Jade et al., 2004;P. Z. Zhang et al., 2004), is more consistent with regional stretching and thermal subsidence related to EVS/SCS rifting.
This paper contributes to our understanding of post-extrusion tectonics by more thoroughly defining the Neogene-Recent tectonic setting, kinematics and extrusion processes recorded in south-central Vietnam, using the orientations, slip senses and where possible, ages of faults.Together with regional fault maps (Dja & Khoang, 1998;Kasatkin et al., 2017;V. V. Nguyen & Luong, 2019), our new data are then used to identify discrete tectonic microblocks within Vietnam, which are bounded by documented lithospheric-scale strike-slip faults, and to demonstrate that there has been Cenozoic fault activity in the Quang Nam, Kon Tum, Dray Sap North and South, Da Lat, and Tuy Hoa sectors (Figure 2) of southern Vietnam that (a) post-dates volcanic activity in the diffuse igneous province; (b) potentially reactivates older faults; (c) is more consistent with an extrusion-based tectonic history than an extension-based tectonic history (Fyhn, Boldreel, & Nielsen, 2009;Fyhn et al., 2018;Phach & Anh, 2018;Vu et al., 2017) for the region; and (d) illustrates how extrusion may be accommodated once a free surface is no longer present.Our work expands upon previous work (e.g., Huchon et al., 1994;V.V. Nguyen & Luong, 2019;Rangin et al., 1995) as our lineament analysis covers a broader area, and our data sets, synthesis, and analysis allow us to reconcile conflicting information and models into a holistic tectonic model for south-central Vietnam.

Stress Field Models
Extrusional and extensional tectonic models have previously both been invoked to characterize the complex stress field of Vietnam and Indochina (e.g., Simons et al., 2007;Tingay et al., 2010;D. T. Tran et al., 2013).One proposed model is that of "extrusion" or "escape" tectonics, the process by which the collision of two tectonic terranes leads to escape of material formerly located between those terranes, after Tapponnier et al. (1982Tapponnier et al. ( , 1986)).The proposed extrusion model for Indochina (e.g., Chamot-Rooke & Le Pichon, 1999;Chi & Dorobek, 2004;Chi & Geissman, 2013;Flower et al., 1998;Hoang & Flower, 1998;Michel et al., 2001;Morley, 2007;Tingay et al., 2010;Yan et al., 2006) posits that (a) strong coupling between the asthenosphere and lithosphere and a significant mantle drag torque has translated the Southern Indochina microplate, in response to extrusion of asthenosphere by the closure of the Tethys Sea and Himalayan collision; and (b) the extruded lithospheric block is characterized by a combination of giant strike-slip faults, smaller scale strike-slip faults and pull-apart basins, and minor normal faulting.Alternatively, an extensional model has been suggested based on seismic interpretation from two basins offshore from southern Vietnam, which exhibit a phase of rifting coeval with the propagation of the EVS/SCS rift zone, ascribing the presence of more recent faulting and diffuse continental volcanic activity purely to the westward propagation of this rift and associated thermal subsidence (Figure 1) (Fyhn, Boldreel, & Nielsen, 2009;Fyhn, Nielsen, et al., 2009).EVS/SCS spreading ceased around 16 Ma according to Li et al. (2015), with evidence for continued extension beyond this period noted by Clift and Wilson (2023).These basinal data indicate that normal faulting off-shore predates the voluminous, subaerial volcanism, and that the subsequent volcanic flows erupted into existing rift or pull-apart basins (Huchon et al., 1994).

Extrusion Tectonics
The Red River Fault Zone (RRFZ) in northern Vietnam, along with the Wang Chao Fault Zone (WCFZ) and Three Pagoda Shear Zone (TPSZ), have been described as sinistral shear zones related to the extrusion of Indochina during the Cenozoic (Figure 1) (Jolivet et al., 1999;Lacassin et al., 1997;Rangin et al., 1995).As noted above, here "extrusion" of Indochina refers to the modification of structures and geodynamics of Indochina following the India-Asia hard collision, which resulted in the lateral migration and clockwise rotation of Indochina (Hall, 2002;Hu et al., 2015;Michel et al., 2001;Richter & Fuller, 1996;Simons et al., 2007;Tapponnier et al., 1982;Zhao et al., 2016).The hard collision between India and Asia led to the thickening of the continental crust of the Tibetan Plateau, causing asthenosphere flow to migrate toward the thinned Southeast Asian lithosphere to the southeast (Jolivet et al., 2018).The extrusion model for Indochina assumes a component of mantle flow roughly parallel to the strike of the major strike-slip faults (e.g., Flower et al., 1998;Hoang & Flower, 1998;Yan et al., 2006), which is corroborated by anisotropy recorded in shear-wave splitting data for the upper mantle beneath the northern part of the Indochina-Shan Tai complex (Bai et al., 2009).One major challenge to the extrusion model, however, is that sinistral motion along the RRFZ ceased at ∼17 Ma and became dextral by ∼5.5 Ma (e.g., Fyhn & Phach, 2015;Leloup et al., 1995Leloup et al., , 2001;;Zhu et al., 2009).Cessation of sinistral movement has been considered to mark the end of extrusion of the Indochina block (Leloup et al., 1995(Leloup et al., , 2001;;Zhu et al., 2009) but may instead mark a change in regional or local kinematics.The cause of this regional change in plate kinematics is variously ascribed to the ∼23.6 Ma ridge jump in the EVS/SCS (Li et al., 2015), a change in Indian indentor motion (i.e., coupling of the Indian and Burmese blocks (Fyhn, Boldreel, & Nielsen, 2009;Fyhn, Nielsen, et al., 2009)), or an additional plate tectonic reconfiguration in the region such as the collision of Australian fragments to the SE of Sundaland (Pubellier & Morley, 2014).

Extensional Tectonics
Two major back-arc basins existed in Vietnam during the Permian and Jurassic-Cretaceous periods (Ferrari et al., 2008;Hall, 1996Hall, , 2012;;Hara et al., 2018;Metcalfe, 2013Metcalfe, , 2017;;Morley, 2012;Waight et al., 2021).The E-NE directed subduction of the Paleotethys Ocean during the Permian led to the formation of the Sukhothai-Chanthaburi Volcanic Arc, located on the western edge of Indochina (Metcalfe, 2017;Waight et al., 2021).The Nan-Uttaradit and Sa-Kaeo sutures, situated in Southeast Asia, have been suggested to mark the boundary between Indochina and the Sukhothai Terrane, believed to be the site of a remnant Permian back-arc basin (Hara et al., 2018;Metcalfe, 2017;Sone & Metcalfe, 2008;Sone et al., 2012;Wang et al., 2018).Tri and Khuc (2011), through a remote sensing and field-based study, suggested that during the Early and Middle Jurassic, Southern Vietnam was situated in a passive margin setting along the eastern edge of the Indochina plate.This passive margin setting transitioned into a more dynamic back-arc fold-thrust belt, marked by a shift from passive to active tectonic setting, with subsequent deformation driven by changes in subduction angle and/or subduction obliquity during the Jurassic (Schmidt et al., 2021).Throughout the Late Jurassic and the Cretaceous, igneous activity along the coastlines of southern Vietnam and southeastern China was linked to an eastern subduction zone (Schmidt et al., 2021;Shellnutt et al., 2013;Thuy et al., 2004;Xu et al., 2016).During this time a NW-SE oriented back-arc formed in south-central Vietnam (Ferrari et al., 2008;Hall, 1996Hall, , 2012;;Hara et al., 2018;Metcalfe, 2013Metcalfe, , 2017;;Morley, 2012;Waight et al., 2021).
During the Late Cretaceous, rifting initiated within the proto-South China Sea basin, giving rise to an ENE-WSWoriented extensional fault system (Barckhausen et al., 2014;Chung et al., 1997;Ye et al., 2018;Zhou et al., 1995).Rifting of the Proto-South China Sea was followed by the opening of the NE-SW striking EVS/SCS at ∼32 Ma (Figure 1) (Briais et al., 1993;Carter et al., 2000;Chung et al., 1997;Clift et al., 2008).There is also evidence for the initiation of extension in the EVS/SCS at ∼52 Ma (Deng et al., 2020).Spreading ceased at ∼16 Ma in the southwest sub-basin of the EVS/SCS, closest to our study area (Li et al., 2015).After EVS/SCS spreading ceased, some studies have proposed that rifting may have propagated westward into continental Vietnam, while lingering upper mantle upwelling generated ongoing diffuse seamount activity within the EVS/SCS (Barckhausen et al., 2014;Cullen et al., 2010;Matthews et al., 1997;Yan et al., 2006).The 16 Ma age reported by Li et al. (2015) also approximately corresponds with the cessation of sinistral motion along the Red River Fault Zone (RRFZ; Figure 1), marking a simultaneous change in regional plate kinematics.Kasatkin et al. (2017) mapped a series of large-scale strike-slip faults that divide the study area into discrete lithospheric sectors with distinct basement lithologies (Figure 2).The northernmost sector, the Quang Nam sector (Figure 2), consists of a Precambrian to Paleozoic accretionary belt (Anh et al., 2021;Tung & Tri, 1992) bounded by the Tam-Ky Phuoc Son Shear Zone (TKPSSZ) to the south and the East Vietnam Transfer Zone (EVTZ) to the east.A major Mesozoic suture zone was likely present along what are presently the TKPSSZ and the Poko Shear Zone (PKSZ), inferred from (a) continuity of kinematic indicators and deformation, (b) allochthonous assemblages of ophiolitic mafic to ultramafic rocks; (c) highly deformed migmatites; and (d) arc-type volcanic sequences (Lepvrier et al., 1997(Lepvrier et al., , 2004(Lepvrier et al., , 2008;;T. H. Tran et al., 2014;Van et al., 2001).The EVTZ is a large-scale transform fault inferred to be a regional, extrusion-related fault and is linked to the RRFZ (Fyhn & Phach, 2015;Fyhn et al., 2018;Leloup et al., 2001;H. P. Nguyen et al., 2012;Phach & Anh, 2018;Tapponnier et al., 1986).Several models have proposed that the EVTZ serves as the eastern boundary of the Sundaland block, and the transtensional sinistral slip along the EVTZ has been identified by some authors as the primary driving force for the opening of the EVS/SCS (Hall, 2002;Leloup et al., 1995Leloup et al., , 2001;;Tapponnier et al., 1986).
Mesozoic and Paleozoic accretionary terranes compose the basement of the two southernmost lithosphere sectors, Da Lat and Dray Sap (Figure 2) (Anh et al., 2021;Tung & Tri, 1992).The Tuy Hoa-Cu Chi Fault (THCCF) separates the two southern lithosphere sectors (V.V. Nguyen & Luong, 2019) and is a prominent fault system believed to have originated during the Eocene-Miocene, following the Cenozoic extrusion of the Indochina block (Phach & Anh, 2018).

Methods
This is a combined remote-sensing and field-based study.We interpreted lineaments across the field area using Landsat Enhanced Thematic Mapper Plus (ETM+) data and Digital Elevation Model (DEM) data.Landsat ETM+ and DEM data sets were sourced from the Global Land Cover Facility and the Open Development Mekong website, respectively.Landsat ETM+ data were downloaded as separate bands and combined into a false-color composite; specifically, bands 531 were combined as RGB in ArcGIS, and this raster was stretched using the histogram equalize operation.Landsat ETM+ data were also combined as a true color composite using bands 321 as RGB in ArcGIS.Data sets were already referenced to a WGS84 reference frame, so no conversions were necessary.The remote data sets were compiled into an ArcGIS project, and other information was incorporated by database upload or by georeferencing JPG files.Other information comprises: (a) our own field locations; (b) lineament maps from Huchon et al. (1994), V. V. Nguyen and Luong (2019), and Rangin et al. (1995) (Figure 3), and a series of Geological and Mineral Resources Maps (Dja & Khoang, 1998); and (c) locations with dated basalt samples from An et al. ( 2017), Hoang et al. (2019), and Lee et al. (1998).
Following the methods of Drury ( 2004), lineaments were picked from the remote data sets based on textural changes in the images and DEM (Table 2).Linear changes in the texture of the land surface often indicate a faultcontrolled change, although care must be taken to avoid regions where human activity has altered the land surface; such regions can be identified by the typical regular checkerboard pattern of cultivated fields and field boundaries and the proximity to dwellings.However, at the scale used (1:300,000, selected for ease of comparison to recent maps from V. V. Nguyen and Luong (2019)) the checkerboard pattern of human cultivation cannot be discerned in the DEM images.We were typically unable to ground-truth the lineament mapping in the field because of the variation in scale between mapped lineaments (typically 10s of km long) and field-scale observations.We undertook three field trips to the region in 2016, 2018, and 2020 (Figure 2).The first was a reconnaissance expedition in 2016 to the southern part of the Central Highlands, near Ho Chi Minh City and Buon Ma Thuot.The second was a more extended expedition aimed at observing the structural geology in the Central Highlands.For the final field season, we focused on fault slip sense and targeted key locations not previously visited in the Quang Nam, Kon Tum, and Da Lat sectors.We inspected rock exposures at every stop, and where relevant, we made measurements of bedding attitude, fault attitude, slickenside pitch within the fault plane, and where possible, fault kinematics or apparent offset, where true slip sense could not be determined.At each site, we documented any cross-cutting relationships between faults and host rocks, considering whether the fault terminated against specific rock units or cut all observable exposed rocks.In some locations, the relationships between different generations of slickensides were observed and noted.On our return to the lab, these data were synthesized using GIS, Stereonet 10™, and FaultKin™ (Allmendinger et al., 2012;Marrett & Allmendinger, 1990) to determine relationships to the lineament map, similarity in fault orientations, and analysis of stress regimes, respectively.
To bracket the age of faulting, the age of an alkali basalt flow (field sample number 2016-CH-10) (Richard et al., 2024), retrieved from location 11 (10.5076°N,107.2729°E, and ∼70 m. elevation) was determined using 40 Ar/ 39 Ar methods at the Oregon State University Argon Geochronology Lab in Corvallis, Oregon.We have dated only one alkali basalt flow that was cross-cut by a fault, because unaltered basalt flows that are cut by faults are relatively rare in the study area, and because we primarily aimed to supplement existing ages from the literature (e.g., Lee et al., 1998).This sample was thus used in conjunction with existing lava flow and core dates from basalt plateaus from An et al. ( 2017), Hoang et al. (2019), and Lee et al. (1998).The sample was crushed, sieved to ∼300 μm grain size, rinsed in distilled water, dried at low temperature in an oven at ∼80°C, then mildly leached to remove impurities.The procedure for leaching was a 20-min soak in 5% HNO 3 in an ultrasonic bath, followed by three rinses in distilled water, then a 20-min soak in distilled water in the ultrasonic bath and three more distilled water rinses; the sample was then again dried at 80°C or lower in the oven.The sample was then irradiated in the TRIGA experimental reactor at the OSU Radiation Center at 1 MW power.The neutron flux during irradiation was monitored using the Fish Canyon Tuff-NM standard, with an adopted age of 28.20 ± 0.02 Ma (after Kuiper et al., 2008), 40 Ar/ 39 Ar = 9.733 ± 0.008, and J-value of 0.001615 ± 0.000001.For mass spectrometry, the sample was analyzed by incremental heating using a bulk CO 2 laser heating method on the ARGUS-VI-D instrument at OSU. Ages were determined using a decay constant of 5.53 ± 0.05 × 10 10 a 1 (Steiger & Jäger, 1977) and age correction methods after Min et al. (2000).Heating plateau dates were determined using an error-weighted mean of plateau steps.Additional standard and procedural blank results are available in Richard et al. (2024).

Results From Remote Sensing Data
The DEM data set (Figure 4), is overlain on a hillshade data set for ease of visualization.The figure also includes field locations, notable strike-slip faults, and the identified lineaments from this study.Figures 5a-5f shows rose  Despite the discrepancy in the number of lineaments in each rose diagram, some patterns emerge with lineaments having orientations consistent with adjacent large-scale strike-slip faults (Figures 4 and 5, Table 2).Lineaments from our study (Figures 4 and 5a-5f) overall exhibit orientations consistent with the adjacent large-scale strike-slip faults mapped by Kasatkin et al. (2017) that bound each sector.Lineaments from other published studies (Huchon et al., 1994; V. V. Nguyen & Luong, 2019;Rangin, et al., 1995) (Figures 3 and 4, Table 2) overall exhibit three prominent orientations, N-S, NW-SE and NE-SW.The results of these three studies are also consistent with our data and the adjacent large-scale strike-slip faults.

Fault Orientations From Field Data
The fault plane orientations measured in the field (Figure 6), are grouped by lithospheric sector (after Figure 2, Table 2), and Table S1 in Supporting Information S1 presents geographic coordinates, sector names, brief outcrop descriptions, pertinent measurement information, and prominent fault orientations for all reported field sites.We measured a total of 349 fault attitudes at 26 field localities.Using this data set, we then inferred an overall, dominant fault trend for each locality, illustrated by red lines in Figure 6a.
We observe that the orientations of the dominant faults for each lithospheric sector are distinct from one another.In all sectors, fault orientations are similar to adjacent large-scale strike-slip faults and other regional fault systems, as outlined in Table 2.In the Tuy Hoa sector, fault orientations are similar to those of the adjacent SBF and THCCF.Similarly, faults in the Da Lat sector are subparallel to the adjacent strike-slip faults mapped by Kasatkin et al. (2017;VTCNF, THCCF, and SBF), along with the EVTZ.The dominant fault attitudes in the Quang Nam sector mimic the EVTZ, with some resembling the TKPSSZ.The Kon Tum and Dray Sap sectors exhibit fault patterns similar to the previously mentioned sectors in that they likewise mirror the adjacent strike-slip faults (Kasatkin et al., 2017).

Analysis of Faulting Regimes
The results from moment tensor solutions in FaultKin™, using slickenside data from the observed fault planes illustrates the orientations of tensional axes calculated from FaultKin™, all for moment tensor solutions categorized as younger, older, and undefined based on observed field relationships (Figures 7 and 8)."Oldest" and "youngest" sets of slickensides were defined where we observed clearly visible cross-cutting relationships in outcrop.We categorized all other slickenside data without visible cross-cutting relationships, or where we could not bracket fault movement age independently, as "undefined."Overall, older orientations produced solutions with a dominantly normal (to slightly oblique) sense of motion, while younger faults have an oblique (to strikeslip) sense of motion.
We developed moment tensor solutions for the Da Lat, Tuy Hoa, and Dray Sap sectors from seven locations (Figure 7) (locations 1, 4, 6, 14, 15, 23, and 26).Based on cross-cutting slickensides observed in the field (such as the examples shown in Figure 9), the moment tensor solutions (Figure 7), and the trends and plunges of tensional stress axes (Figure 8), there appears to be a heterogeneous stress field that has changed with time.In the Da Lat and Tuy Hoa sectors, older solutions have a normal to oblique-slip sense, while younger solutions have a strike- slip to oblique-slip-sense, regardless of fault orientation.In both Dray Sap sectors, we also have "undefined" strike-slip-like moment tensor solutions at locations 23 and 26 (Figure 7).
For the Kon Tum and Quang Nam sectors, our moment tensor solutions indicate a heterogeneous stress field similar to that of the Tuy Hoa, Da Lat and Dray Sap sectors (Figure 7).In the Kon Tum and Quang Nam sectors we have five moment tensor solutions (locations 16, 18, 19, 21 and 22).We were able to define younger and older sets of slickensides at locations 18 and 21, while slickensides at locations 16, 21, and 22 are undefined (Figure 7).From our solutions combined with pitch angles measured in the field (Figure 9), we observe that similar to Tuy Hoa, Da Lat and Dray Sap, the younger fault motion set in Kon Tum and Quang Nam has a strike-slip to obliqueslip sense, while the older motion has a dip-slip to oblique-slip sense.At locations 16, 19, and 22, our additional slickenside measurements, whose relative ages are "undefined," exhibit moment tensor solutions with a strikeslip to slightly oblique-slip sense.

Numerical Ages
The field photograph in Figure 10, for map location 14, shows the presence of closely spaced NE-SW striking faults that intersect the Pliocene Soc Lu Formation, indicating a fault age younger than Pliocene.To achieve more quantitative age constraints in this study's field area, we measured an additional 40 Ar/ 39 Ar date of 0.6 ± 0.004 Ma in basalt sample 2016-CH-10 from location 11 (Figure 11, see Supporting Information S1).For additional regional age constraints, Figure 11 presents a more detailed geologic map depicting part of the Da Lat sector with sample sites of interest.From the cross-cutting lineaments on the map intersecting multiple dated basalt flows, we infer that the fault ages are younger than ∼0.6 Ma, based on our 40 Ar/ 39 Ar age data and those reported by Lee et al. (1998).

Tectonic Microblocks of South-Central Vietnam
Given that fault orientations are distinct in each sector (Figure 6a), we interpret the varying lineament and fault trends within each fault population to indicate that the Da Lat, Dray Sap, Kon Tum, Quang Nam and Tuy Hoa sectors are tectonically discrete microblocks.From here on, we will use the word "microblock" to refer to the geographic sectors described above.We further infer that the lithospheric-scale strike-slip faults mapped by Kasatkin et al. (2017) bound the five discrete tectonic microblocks (Figure 2).Lastly, based on sharp changes in fault and lineament orientations in the Dray Sap microblock (Figure 4), namely a change from dominantly E-W lineaments in the south to dominantly N-S lineaments in the north, we further postulate the presence of a previously unmapped fault (shown as a dashed line in Figures 2,  4, and 7), which divides the Dray Sap microblock into northern and southern sections.As further justification for our microblock model, we note that each lithospheric sector or microblock contains a distinct basement lithology and has a distinct geologic history (Sections 2.2 and 2.3), and each is bounded by deep-seated oblique-slip faults or Mesozoic suture zones.The stress regime solutions (Figure 7) are likewise distinct in each microblock.

Evolution of Vietnam Microblocks
As described above, cross-cutting slickensides in several field locations, along with the trend and plunge of tensional axes (Figure 8), provide additional insights into changes in stress fields over time within each microblock.Figures 7-9 imply that there was an early phase of dip-slip on many faults, followed by oblique-to strike-slip motion.We infer from cross-cutting slickensides that the Kon Tum, Tuy Hoa and Da Lat microblocks initially hosted dip-slip, likely normal faults (Figure 9).Unfortunately, we were unable to confirm a similar history for the Quang Nam and Dray Sap microblocks based on our data set.
We lack timing constraints on the ages of the fault sets defined as "older" in this study, which can be found to cut Jurassic sedimentary rocks (e.g., Location 15) and are, therefore, only constrained as definitively younger than Jurassic age.Furthermore, although the moment tensor solution calculated for Location 15 aligns with the strike orientation delineated by Rangin et al. (1995) (Figure 3c), it does not match the prevailing fault trend observed in the region (Figures 6 and 7).We suggest this discrepancy implies that the principal stress indicated by the older moment tensor solution is likely related to a significantly earlier tectonic event, and that this fault likely persists in the subsurface.Overall, we postulate that the "older," dip-slip episode is related to back-arc rifting during the Jurassic-Cretaceous.Support for this interpretation comes from our cross-cutting age relationships and the co-alignment of fault and moment tensor solution orientations to local Jurassic-Cretaceous extensional tectonics (Figure 8) (Hall, 1996;Morley, 2012;Nam, 1995).
The information in Figures 7-10 together implies that there was a recent phase of oblique-to strike-slip motion on many faults in our study area.Based on published basalt dates (Hoang et al., 2013(Hoang et al., , 2019;;Lee et al., 1998), our   supplementary 40 Ar/ 39 Ar date, and the cross-cutting slickensides observed in the field , we postulate that the more recent oblique to strikeslip motion was active until at least ∼0.6 Ma.The more recent activity associated with oblique-to strike-slip motion, in contrast to the older dip-slip type of faulting, exhibits significant heterogeneity in the moment tensor solutions across the field area (Figures 7 and 8).For example, based on relationships in Figures 7 and 8, the differences in young moment tensor solutions between map locations 1 in the Tuy Hoa microblocks, locations 4, 6, and 15 in the Da Lat microblock, and locations 1, 18, and 21 in the Kon Tum microblock suggest that these microblocks are experiencing deformation independent of one another.This young, heterogeneous stress field, characterized by oblique-to strike-slip moment tensor solutions, is indicative of an extrusion tectonic regime rather than an extensional one, which would exhibit more homogeneity and predominantly dip-slip faults in the moment tensor solutions.

Continuum Rubble Tectonic Model for Indochina
As described by Dewey et al. (2008) for the Coso region of Southern California, we posit that some of the microblocks of south-central Vietnam are undergoing rotation about vertical axes, as well as potential deformation in a broadly transtensional regime.Dewey et al. (2008) described this working model as the "continuum rubble" behavior of small blocks, and we find this term valuable to describe the rotation and "jostling" of the microblocks of Indochina between the RRFZ, EVTZ, and the MPFZ (Figure 1).Numerical models for the Coso region, for example, by Eckert and Connolly (2007) and Pearce and Dewey (2008) show that both dip-slip and wrench faulting components characterizes such a "continuum rubble" regime, together with the development of significant breakup of a large lithospheric block into what Eckert and Connolly (2007) call "second-order" fractures.These models further support our suggestion that such continuum rubble behavior can explain the polyphase deformation observed on the faults in south-central Vietnam, with both dip-slip and wrench-faulting components.

Continuous Extrusion of Microblocks
Contrary to existing literature (e.g., Rangin et al., 1995;Searle et al., 2010;Zhu et al., 2009), which suggests that faulting associated with both extrusion and extensional tectonic regimes has ceased in the Indochina Peninsula, our results (e.g., Figure 11) demonstrate that faulting has been more recent than lava flows dated 4.3 ± 0.2 Ma (An et al., 2017), 0.6 ± 0.004 Ma (this study), and 0.24 ± 0.1 Ma (Lee et al., 1998).These age constraints indicate that some faulting is significantly younger than the last two documented major pulses of basaltic volcanism (5.4-1.75Ma and 0.7-0.57Ma; Tri et al., 2011), than the postulated end of tectonic extrusion based on the change in motion of the RRFZ (5.5 Ma; Leloup et al., 1995Leloup et al., , 2001;;Zhu et al., 2009) and, significantly, than the cessation of rifting in the EVS/SCS (∼16 Ma; Li et al., 2015).Thus, a tectonic regime more complex than thermal subsidence must be operating across Vietnam.Given our evidence for a highly heterogeneous stress field and our inferred fault ages, we propose that the relatively recent fault slip motions within the study area can instead be attributed to ongoing extrusion tectonics.
As mentioned, paleomagnetic and GPS data from the core of the Sundaland block show that the Kon Tum microblock is likely moving to the east and rotating clockwise within Sundaland (Chamot-Rooke & Le Pichon, 1999; Chi & Dorobek, 2004;Chi & Geissman, 2013;Michel et al., 2001;Morley, 2007;Tingay et al., 2010;D. T. Tran et al., 2013).This rotation is consistent with a regional model whereby Sundaland is composed of not a rigid core, but a continuum rubble of fragments that interact, internally deform, and in some cases rotate about vertical axes with respect to one another under regional stresses.The eastward motion of the Kon Tum microblock is thus consistent with the continued extrusion of material from the Himalayan collision to the east and southeast.The Da Lat, Tuy Hoa, and Quang Nam microblocks lack GPS data at a fine enough scale to resolve the proposed motion and merit further investigation (V.V. Nguyen & Luong, 2019, and references therein), but we posit that they are responding to similar regional tectonic processes.
The RRFZ ceased sinistral motion ∼17 Ma and initiated dextral motion ∼5.5 Ma (e.g., Leloup et al., 1995Leloup et al., , 2001;;Zhu et al., 2009;Zuchiewicz et al., 2013).As noted above, the 17 Ma change in slip sense was previously inferred to mark the end of extrusion in Indochina.However, instead, we suggest that this event marks a change in the kinematics of the extrusion process.The postulated end of tectonic extrusion does not account for the asthenospheric flow associated with the extrusion of Indochina, or for documented ongoing volcanism (e.g., the 1923 eruption of Île des Cendres), which instead suggest continued mantle flow.To accommodate the motion of mantle flow beneath the Shan Tai, Kon Tum, Da Lat, and other microblocks would require rotation of those blocks within the confines of the larger-scale shear zone defined by the Red-River-East Vietnam Transform and Mae Ping Fault Zones.This rotation is required because there is no free surface into which these blocks can be extruded, given their position in the core region of Sundaland and the presence of Borneo.

Conclusions
To characterize controls on present-day deformation in south-central Vietnam, we measured fault data throughout the region, which we compared to remote sensing data from this and previous studies (Huchon et al., 1994;V. V. Nguyen & Luong, 2019;Rangin, et al., 1995).We also compared the cross-cutting relationships between numerous observed and measured slickensides to infer the change in fault slip over time, and the relationships of faults with basalt flows of a given age to constrain the age of faulting throughout the study area.We conclude that: (a) cross-cutting relationships between faults and basalt flows support recent faulting within Vietnam; (b) fault and lineament orientations suggest a locally heterogeneous stress field related to the faults mapped by Kasatkin et al. (2017); (c) dip-to oblique-slip faults were reactivated during the extrusion of Indochina; (d) recent strike-to oblique-slip faults and stress fields suggest continuous extrusion of Indochina; and (e) each lithospheric sector (or microblock) has a different tectonic history and moves independently.Finally, the breakup of Indochina into these microblocks is likely to have occurred at the same time as the cessation of spreading in the EVS/SCS and the change in motion sense on the RRFZ at ∼17 Ma, which was previously suggested to indicate the end of local tectonic extrusion.Instead, we postulate that these changes indicate a change in the kinematics of the extrusion process, as extrusion continues in the absence of a free surface.We suggest that once the free surface is removed by other tectonic processes, block breakup and rotation are the inevitable consequence of ongoing mantle flow and that for extrusion to occur, a strong lithospheric-asthenospheric coupling is, in fact, necessary.2), and locations with age constraints (colored stars), all located within the Xuan Loc and Cu Chi Formations.The red box in the inset map shows the location of the detailed geologic map.Numbered dark purple stars mark field locations from this study (Figure 2); yellow stars mark locations with previously dated basalts (either from outcrop or cored sampling; Lee et al., 1998) that range in age from 2.42 ± 0.08 Ma in the south to 0.24 ± 0.06 Ma in the north part of the Xuan Loc Formation.Basalts at location 14 have also been previously dated by An et al. (2017) to be 4.3 ± 0.2 Ma.We measured a sample from location 11 and determined an age of 0.6 ± 0.004 Ma (see Supporting Information S1).The geologic base map was modified after the Geological and Mineral Resources Map of Vietnam, Gia Ray Region (Dja & Khoang, 1998).

Figure 2 .
Figure 2. A simplified map of Vietnam where numbers indicate our field locations from this study.Volcanic plateaus (gray) with names were modified from Hoang et al. (2013), while Kasatkin et al. (2017) mapped the major strike-slip faults (red lines), which separate distinct lithosphere sectors.The large text surrounding the map is the name of each sector and the age of basement lithology.The thick black line shows the study area presented in the subsequent figures.The names of the faults mapped by Kasatkin et al. (2017) are bolded and abbreviated next to the faults, which are VTCNF, Vung Tau Ca Na Fault; THCCF, Tuy Hoa-Cu Chi Fault; SBF, Song Ba Fault; EVSZ, East Vietnam Shear Zone; PKSZ, Poko Shear Zone; TKPSSZ, Tam Ky Phuoc Son Shear Zone.

Figure 3 .
Figure 3. Contrasting fault maps for south-central Vietnam, after (a) V. V. Nguyen and Luong (2019) showing their interpretation of remotely sensed faults in the region; (b) Huchon et al. (1994), with their interpretation of the Paleogene fault framework in the region; and (c) Rangin et al. (1995) with their contrasting interpretation of the dominant fault patterns in the area.The black box marks the location of the present study area.Gray shaded areas mark the boundaries of known basalt flows; the cross-cutting relationships between faults and basalt flows are unclear in all three interpretations.

Figure 4 .
Figure 4. Digital Elevation Model of Southern Vietnam overlain on a hillshade map to better highlight the Central Highlands region.Also shown are field locations for this study (stars, symbols as in Figure 2), mapped lineaments from this study (black lines), and major strike-slip faults (red lines) after Kasatkin et al. (2017) and this study.

Figure 5 .
Figure 5. Rose diagrams showing the orientations of lineaments (Figure 4) within the fault-bounded sectors (a-f), along with (g-i) a comparison with the orientations of mapped lineaments within the overall study area from previous work (indicated by the green border (Huchon et al., 1994; V. V. Nguyen & Luong, 2019; Rangin et al., 1995)).All rose diagrams are in 5°b ins, and the perimeter of the diagram is 5% of the data.(j) The key and red bars illustrate the orientations of the major blockbounding faults; EVTZ is the East Vietnam Transfer Zone, THCCF/VTCNF is the Tuy Hua Cu Chi and Vung Tau Ca Na faults respectively, which are broadly parallel to one another, SBF is the Song Ba fault, PKSZ is the Poko Shear Zone, and TKPSSZ is the Tam Ky Phuoc Son Shear Zone.

Figure 6 .
Figure 6.(a) Stereonets numbered 1-26 illustrate the orientations of measured fault planes for each field study location (Figure 2), with an example diagram shown as a legend in (b).The bold black lines in (a) group the local stereonets by lithospheric sector.Thick red lines in each stereonet indicate the main fault strands inferred from the average of our fault measurements for that field site, while blue and orange lines indicate sinistral and dextral faults, respectively.The grayshaded lines are additional faults with field orientation measurements, but which lacked slip indicators.

Figure 7 .
Figure 7. Map of Vietnam with major volcanic centers in gray and the field area outlined in black, after Figure 2, and showing FaultKin™ stress regime solutions for field locations in this study; field locations are indicated using small numbers that correspond to the numbered locations in Figure 2.Colored borders indicate the age category, where teal is "oldest," yellow is "youngest", and purple is "undefined", as described in the text.

Figure 8 .
Figure 8. Combined stereonet showing the calculated trend and plunge for the tensional axes derived from "oldest" and "youngest" cross-cutting slickensides, as well as "undefined" slickensides.Note the shift in trend and plunge values between the younger and older pairs of tensional axes, indicating a temporal change in principal stress directions.

Figure 9 .
Figure 9. Three locations showing cross-cutting strike-to oblique-slip and dip-slip lineations.(a, b) Are from location 6, (c, d) are from Location 15, and (e, f) are from location 21.In each location, one photograph (images a, c, and e) is shown unaltered, and one photograph (images b, d, and f) has been artificially lightened and annotated to enhance the visibility of outcrop textures.All three locations show that the strike-slip to oblique-slip lineations cross-cut and are therefore younger than the dip-slip lineations.

Figure 10 .
Figure 10.Fault surfaces in the Pliocene Soc Lu Formation at map location 14.The black dashed line shows the strike of the fault, and the red arrow shows the trend and plunge of lineations on one fault plane.Other, similar fault plane lineations are illustrated with black arrows.

Figure 11 .
Figure11.Map of a subregion of the Da Lat study area showing age relationships between our mapped lineaments (Figure2), and locations with age constraints (colored stars), all located within the Xuan Loc and Cu Chi Formations.The red box in the inset map shows the location of the detailed geologic map.Numbered dark purple stars mark field locations from this study (Figure2); yellow stars mark locations with previously dated basalts (either from outcrop or cored sampling;Lee et al., 1998) that range in age from 2.42 ± 0.08 Ma in the south to 0.24 ± 0.06 Ma in the north part of the Xuan Loc Formation.Basalts at location 14 have also been previously dated by An et al. (2017) to be 4.3 ± 0.2 Ma.We measured a sample from location 11 and determined an age of 0.6 ± 0.004 Ma (see Supporting Information S1).The geologic base map was modified after the Geological and Mineral Resources Map of Vietnam, Gia Ray Region(Dja & Khoang, 1998).

Table 1
Common Names and Their Acronyms

Table 2
Sector Name and Lineament Orientations