Late Cenozoic Cooling History of the Xigaze Fore-Arc Basin along the Yarlung–Zangbo Suture Zone (Southern Tibet): New Insights from Low-Temperature Thermochronology

The Tibetan Plateau is currently the widest and highest elevation orogenic plateau on Earth. It formed as a response to the Cenozoic and is still ongoing collision between the Indian and Eurasian plates. The Xigaze fore-arc basin distributed along the Indus–Yarlung suture zone in southern Tibet preserves important information related to the late Cenozoic tectonic and topographic evolution of the plateau. In this study, apatite fission track (AFT) thermochronology was carried out on twelve sandstone samples from the middle segment of the Xigaze basin and additionally on four sedimentary rocks from the neighboring Dazhuka (Kailas) and Liuqu Formations. Inverse thermal history modeling results reveal that the fore-arc basin rocks experienced episodic late Oligocene to Miocene enhanced cooling. Taking into account regional geological data, it is suggested that the late Oligocene-early Miocene (~27–18 Ma) cooling recognized in the northern part of the basin was promoted by fault activity along the Great Counter thrust, while mid-to-late Miocene-accelerated exhumation was facilitated by strong incision of the Yarlung and Buqu rivers, which probably resulted from enhanced East Asian summer monsoon precipitation. Sandstone and conglomerate samples from the Dazhuka and Liuqu Formations yielded comparable Miocene AFT apparent ages to those of the Xigaze basin sediments, indicative of (mid-to-late Miocene) exhumation soon after their early Miocene burial (> ~3–4 km). Additionally, our new and published low-temperature thermochronological data indicate that enhanced basement cooling during the Miocene prevailed in vast areas of central southern Tibet when regional exhumation was triggered by both tectonic and climatic contributing factors. This recent and widespread regional exhumation also led to the formation of the high-relief topography of the external drainage area in southern Tibet, including the Xigaze fore-arc basin.Orogenic belts are dominant topographic features on Earth and are characterized by high tectonic activity and high elevations. They provide the best natural laboratory to study the coupling between tectonics, erosion, and climate [1-4]. In these regions, negative feedback between fast denudation and high elevation causes enhanced erosion that, in turn, tends to reduce the topography. The collision between India and Asia led to the formation of the Tibetan Plateau, which stands ~4–5 km high over a region of ~3 million km2 (Figures 1(a) and (b)). The southern Tibetan Plateau (i.e., the southern Lhasa terrane—Tethyan Himalaya) is characterized by a high-elevation and low-relief landscape (i.e., “flat” highland) [5]. Structurally, several ~E–W trending large-scale thrust faults and a series of ~N–S striking normal faults are well developed in the southern Lhasa terrane [6-8]. Furthermore, the west-to-east flowing Yarlung river runs through the southern Tibetan Plateau, its source is high in western Tibet, and it cuts through the Namche Barwa syntaxis before flowing through the Siang river into the Brahmaputra river (Figure 1(c)). Topography of the southern Tibetan Plateau was thought to have been significantly influenced by the long-term evolution of the Yarlung river since the Oligocene [9, 10]. In the mid-Miocene, enhanced incision and efficient detritus transport were considered to be related to the intensification of the Asian monsoon [11-13]. Hence, the southern Tibetan Plateau is an ideal location to study the interactions and feedback loops between tectonic, erosion, and climate, particularly in terms of the temporal and spatial variability of surface uplift and exhumation of this large orogenic plateau.The Indus–Yarlung suture zone is located in southern Tibet, where it delineates the contact between the Asian and Indian continents [14-16]. The Xigaze fore-arc basin trends east–west along the north of this suture and contains well-preserved successions that generally underwent limited deformation [17, 18]. It preserved key information on the closure of the Neo-Tethys Ocean and the collision between India and Asia [19, 20]. Figuring out the (postdepositional) tectono-thermal history of this fore-arc basin is therefore crucial for understanding the complex topographic evolution and geodynamic processes that shape the southern Tibetan Plateau. In the past few years, few thermochronological studies have been carried out to reveal the burial and exhumation histories of the Xigaze fore-arc basin. A northward younging age trend (from the Late Cretaceous to the Miocene; based on apatite fission track [AFT] thermochronology) related to changes in India–Asia convergence rates is documented. From the northern and central Xigaze basin, published zircon (U-Th)/He (ZHe) mean ages generally fall within the range of ~35–25 Ma [21, 22]. However, clear along-strike variation of the exhumation history is observed along this sedimentary sequence [21, 23, 24]. On the other hand, the incision of the Yarlung river and its tributaries has significantly influenced the morphotectonic evolution of the drainage area, including the fore-arc basin since the late Oligocene-early Miocene [13, 25-27]. The interaction between coeval tectonic activity and fluvial erosion in shaping the basin’s landscape was not fully explored in previous studies. These two major issues hence spur this study to establish a clearer picture of postdepositional basin evolution.Here we present a low-temperature thermochronological study from a cross section (Micun—Jiding) in the middle part of the Xigaze fore-arc basin and the nearby Dazhuka and Liuqu Formations [28, 29], based on AFT dating. In the first place, a new set of AFT age data together with inverse thermal history modeling results is provided. Second, a more complete late Cenozoic thermotectonic history of the entire fore-arc basin was reconstructed. Finally, we combined the cooling history of the Gangdese batholith in the north to better discuss the exhumation process and tectono-thermal evolution of the area, taking into account the roles of tectonic, fluvial incision, and climatic events.The formation of the Tibetan Plateau’s tectonic units is related to the accretion of several microcontinents, accretionary belts, and island arcs onto the southern margin of the evolving Asian continent from the early Paleozoic until the earliest Cenozoic [30-33]. It is comprised of four major tectonic units, including, from north to south, the Songpan–Ganzi, Qiangtang, Lhasa terranes, and the Tethyan Himalayan orogen (Figure 1(b)). Several major ophiolite-bearing suture zones (e.g., the Jinshan, Bangong–Nujiang, and Indus–Yarlung suture zones) delineate these terranes (Figure 2), they all trace consumed oceanic basins (mainly from the Tethyan realm) that separated the original terranes and thus mark the diachronous collision accretions of these terranes to the developing southern Asian margin. The Lhasa terrane is situated in southern Tibet, and it extends ~E–W for ca. 2000 km with a width of ~100–300 km [30, 33, 34] (Figure 1(c)). Based on clear differences in basement nature and sedimentary covers, the Lhasa terrane can further be subdivided into northern, central, and southern subterranes [35, 36]. The southern Lhasa subterrane is intruded by the vast Gangdese batholith, which is a long magmatic belt extending over 1500 km from west to east (Figure 2). It is dominated by Meso–Cenozoic plutonic rocks [37-39]. The Mesozoic igneous rocks were formed as a magmatic arc resulting from the northward subduction of the Neo-Tethyan oceanic plate underneath the amalgamating southern Asian (Tibetan) margin [40-42], while the Cenozoic sections of the batholith are closely related to the collision between India and Asia [39].The Indus–Yarlung suture zone marks the southern boundary of the Lhasa terrane and is characterized by a narrow east–west striking ophiolite and serpentinite matrix mélange belt, which formed during the Late Jurassic to Early Cretaceous [43-45]. The Xigaze fore-arc basin is preserved as an east–west trending synclinorium along the north of this suture zone (Figure 2). The southern margin of the Xigaze fore-arc basin is bounded by a south-dipping splay of the Great Counter thrust (i.e., a main structural feature in southern Tibet), which places ophiolitic mélange and sedimentary rocks structurally above the Xigaze fore-arc deposits. The northern margin juxtaposed with the Kailas Formation is delineated by the northernmost splay of the Great Counter thrust [18, 20, 46, 47]. The sediments in the Xigaze fore-arc basin were mainly deposited during the Cretaceous to Eocene and can be divided into the Xigaze and Tso–Jiangding Groups. The Xigaze Group can be geologically divided into the Chongdui and Ngamring Formations and is mainly comprised of marine sandstone, shale, and interlayered limestone [16, 17, 48] (Figure 3). The younger Tso–Jiangding Group unconformably overlies the Xigaze Group and generally shows fining-upward conglomeratic strata at the bottom and sandstone and limestone with lesser conglomerate in the upper part. This group is thought to have been deposited in a residual fore-arc basin after the India–Asia collision [16]. These sediments are mainly sourced from the Lhasa terrane to the north [18, 19, 48-50].The Kailas Formation is the name given to a series of Oligocene–Miocene nonmarine sediments that are also deposited along the Indus–Yarlung suture zone. They are exposed along a ~1300 km stretch from west to east and include conglomerates, fluvial sandstones, and fine-grained lacustrine sedimentary rocks [51, 52] (Figure 2). Depending on locality, different local names are attributed to the Kailas Formation and its equivalents. In the Xigaze area, it is referred to as the Dazhuka Formation [28, 29], and zircon U-Pb ages from volcanic tuffs and flows interbedded constrain its deposition at ~23–18 Ma [29, 52, 53] (Table 1). The Liuqu Formation is locally distributed along the southern margin of the Xigaze basin and comprises mainly sandy conglomerates and sandstones intercalated with silty mudstones [54]. Previous studies suggested that the Liuqu conglomerates were deposited in a contractional setting in the suture zone (during or soon after the Indian slab break-off, with detrital inputs from both the Gangdese batholith and accretionary mélange) [28, 52, 55].Field investigations were carried out over two field seasons in 2019 and 2021 when three sets of samples were collected in and near the middle segment of the Xigaze basin (Figure 2). From north to south, the first set is from the Dazhuka (Kailas) Formation and comprises one graywacke (X26) and one sandstone (X27). The second set contains twelve sandstone rocks (X28–39) from the Xigaze Group submarine fan mega-sequences (the Xigaze fore-arc basin) (Figure 3). The Buqu river runs through this sampling profile. The third sample set is from the Liuqu Formation near the ophiolitic mélange and is composed of two conglomerate samples (WG07 and 08). Specific sample information including lithologies, stratigraphic ages (roughly constrained by minimum detrital zircon U-Pb age peaks), locations, and elevations is presented in Table 1. Representative field and hand specimen photographs can be found in online supplementary Figure S1.AFT analysis is a powerful and versatile low-temperature thermochronological technique based on the spontaneous fission decay of uranium-238, by which submicroscopic linear radiation damage tracks are produced in the crystal lattice that can become visible for optical microscopic analysis after chemical etching with nitric acid [56, 57]. Fission tracks in apatite are considered stable on geological time scales at temperatures lower than ~60°C, whereas the fission tracks will anneal rapidly if they are heated over ~120°C [58, 59]. The temperature window between ~60 and ~120°C (ca. 2–4 km crustal depth when considering geothermal gradients of 20–25 °C/km) is known as the apatite partial annealing zone (APAZ). In this temperature window, tracks can accumulate but their lengths are shortened, resulting in lower mean track length values and broader length-frequency distributions [60, 61]. This process also depends partly on the chemical composition of the apatite crystal [62]. The AFT age together with the track length distribution can provide information on the thermal history of the apatite-bearing rock [63].Samples for AFT thermochronometry discussed in this study were analyzed with the external detector method [64]. They were irradiated using two irradiation containers in the well-thermalized (epithermal flux/thermal neutron flux = 0.0088) channel X26 of the BR1 reactor (Belgian Centre for Nuclear Research, SCK, Mol, Belgium). Thermal neutron fluence was indirectly determined using four mica-covered uranium-doped glass dosimeters IRMM-540R [65] spatially distributed in the irradiation package. Spontaneous tracks in apatite were etched using a 5.5 M nitric acid solution for 20 seconds at 21 °C. Induced fission tracks were revealed in the uranium-free Goodfellow “clear ruby” muscovite mica (i.e., external detector) with 40 vol% HF for 40 minutes at 21 °C. Sample mounts and external detectors were pasted on a glass slide before analysis. Induced and spontaneous fission tracks of both standards and samples were counted manually by applying the Nikon-TRACKFlow software [66] and using a Nikon Eclipse Ni-E microscope with an attached DS-Ri2 camera at 1000 × magnification. The overall weighted mean zeta value for analyst WS is 338.0 ± 3.7 (1σ) a*cm2, based on multiple Fish Canyon Tuff and Durango apatite age standards [67, 68]. Ages were calculated and reported as both central and pooled ages (Table 2). Due to the samples’ very low spontaneous track densities and young AFT ages, a subset of second mounts (for samples that yielded adequate numbers of crystals) was subjected to 252Cf irradiation (at The University of Adelaide [69]) and/or heavy ion bombardment (at the GSI Helmholtz Centre, Germany [70]) to generate measurable confined tracks. The lengths of confined fission tracks were measured using 1000 × magnification and included a measurement of the angle to the c-axis. We aimed at measuring 50–100 confined tracks for each sample, but only limited measurements (all <50 per sample) have been obtained due to very low spontaneous track densities.Thermal history modeling was performed using the QTQt program [71] (version 5.8.0) for the samples in which more than 30 confined track lengths have been measured. This software uses Bayesian transdimensional Markov Chain Monte Carlo statistics to model the thermal history of a sample. Each sample was modeled individually using the multikinetic AFT annealing model from Ketcham et al. [72], while the c-axis projection of confined track lengths [73] was not performed. The etch pit diameter (Dpar) was used as a kinetic parameter. The present-day surface temperature constraint was set as 10 ± 10°C, which is realistic for recent temperatures in southern Tibet. Considering that most of the published ZHe mean ages from the northern and central Xigaze basin are within the ~35–25 Ma time span as mentioned in section 1, a rough higher-temperature constraint of 180 ± 20°C at 30 ± 5 Ma was added for each individual model. The prior age was then set to 20 ± 20 Ma, and the temperature prior was set as 100 ± 100°C that covers the higher temperature constraint in the model. Samples that yielded less than 30 confined fission tracks were also modeled but their results (“expected” models are indicated by dashed lines) can only be considered in combination with other models that are based on more length data. The results of the thermal modeling are shown for each sample in Figure 4(b). Confined track length distributions, and the predicted and observed values for the kinetic parameter, AFT age, and mean track length are shown in Figure 5. Regarding the inverse modeling, we also used the HeFTy software [74] to stimulate and verify the accuracy of the QTQt results.In total, we present new AFT ages for sixteen sedimentary rocks in this study. Both their central and pooled ages are reported (Figure 2; Table 2), and the central ages will be used in the following discussion. Where possible, 20 or more grains were analyzed per sample. Radial plots of each sample are presented in online supplementary Figure S2. Except for X34, all samples passed the chi-squared test at the 95% confidence level (P(χ²) >5%), indicating homogeneous age distributions [75, 76] and suggestive of postdepositional resetting of the AFT system. Twelve Cretaceous sandstone rocks (X28–39) from the Xigaze fore-arc basin show broad Miocene AFT ages that range from ~23 to ~7 Ma. Four early Miocene samples from the Dazhuka and Liuqu Formations, including one graywacke (X26), one sandstone (X27), and two conglomerate rocks (WG07 and 08), exhibit younger mid-to-late Miocene central ages of ~8, ~15, ~10, and ~15 Ma, respectively. AFT central ages of all the detrital samples from the Xigaze fore-arc basin are significantly lesser than their Cretaceous depositional ages (Tables 1 and 2), consistent with previous observations in other sampling profiles [21, 24]. Considering that the chi-squared probabilities P(χ²) of these samples mostly reach >10% with low degrees of dispersion, they must have reached temperatures that are higher than the AFT total annealing temperature (through deep burial or due to heating produced by magmatic events in the neighboring Gangdese batholith), totally resetting the AFT system. Based on the same criterion, samples from the adjacent Dazhuka and Liuqu Formations also experienced complete AFT thermal annealing before their exhumation.After 252Cf irradiation and/or heavy ion bombardment on a subset of additional mounts (for samples with sufficient numbers of grains), six samples (X28, 33, 35, 36, 38, and 39) yielded more than 30 measurable confined tracks. Due to the very low spontaneous track densities, only around 20 confined tracks were found in samples X30, 31, and 37. Regarding the track length, sample X27 displays a very long mean track length value of ~15.4 μm, but it is based on four confined track measurements only. Generally, comparable mean track length values between ~14.0 and ~13.0 μm (Table 2) are found.All the expected models [71] from the modeled samples are combined in Figure 4(b) (with the indication of 95% probability range interval) and detailed individual thermal models for each sample can be found in online supplementary Figure S3 (HeFTy modeling results are also provided there for verification). As mentioned above, we realize that all the models are based on less than 50 confined fission tracks that do not match a normal practice; as a result, the modeling is less strongly constrained and should be interpreted with necessary precaution. For a more quantitative description of the cooling history, we empirically define the cooling rates (in our study area’s continental collisional zone context) of ~5–10 °C/Ma and >10 °C/Ma as moderate and rapid cooling, respectively [13, 27, 77].Therein sample X28 entered the APAZ in the late Oligocene (~27 Ma) and cooled down to ~50 °C in the early Miocene (~18 Ma), with a cooling rate of ~7.8 °C/Ma (Figure 4(b)). Samples X33, 35, and 38 present curves that generally display cooling through the APAZ during ~20−9 Ma with moderate to fast cooling rates (Figure 4(b)). Samples X36 and 39, with younger AFT ages, display comparable cooling histories. They stayed in the APAZ for a relatively short time span in the late Miocene (~10−5 Ma), giving a rapid cooling rate of ~12 °C/Ma (Figure 4(b)). It is noted that the HeFTy modeling presented similar cooling paths (compared with the QTQt ones) for the 6 samples mentioned (online supplementary Figure S3). The additional three samples (X30, 31, and 37) display comparable moderate cooling during the Miocene (~21–8 Ma). Although the three models are based on even lower amounts of confined track measurements, that is, <30, and will not be used for in-depth discussion, their cooling processes within the APAZ show some similarities with those of the nearby samples (i.e., X33, 35, and 38) (Figure 4(b)).Recent low-temperature thermochronological studies in our region of interest indicate that rocks displaying pre-Oligocene ZHe and AFT ages and cooling phases only occur in a narrow zone along the southern margin of the Xigaze basin. In contrast, the central and northern parts of the basin all record late Oligocene–Miocene cooling (Figure 2) [21, 22, 24]. Li et al. [21] suggested that the central and northern fore-arc basin can be characterized by a more protracted burial history spanning the Late Cretaceous to ~35 Ma, in contrast to the southern part, where the burial history was primarily confined to the Late Cretaceous. In addition, their findings reveal a distinctive pattern of episodic exhumation propagating northward in tandem with the migration of the depocenter within the basin. Our new AFT data dominantly yield mid-to late Miocene AFT ages (<20 Ma), including samples X38 and X39 that are located in the southern part of the basin (Figure 2). They exhibit AFT central ages of ~13 and ~8 Ma, respectively. These are even younger than those of the rocks in the central and northern parts. This age distribution is not totally consistent with the spatial pattern of the cooling/exhumation history proposed by Li et al. [21]. In this regard, differential along-strike exhumation seems to have occurred along the Xigaze basin, with late Oligocene–Miocene AFT ages appearing to be regionally extensive, while pre-Oligocene cooling signals are rather scarce and only locally preserved in the southernmost force-arc basin.The Great Counter thrust is a major structure in southern Tibet, and it has also affected our study area (Figure 1(c)). The central parts of the Great Counter thrust (Xigaze area) splay into a set of three distinct fault segments (Figure 2). Among them, the southernmost splay (F1) juxtaposes the Tethyan Himalayan sequence against the ophiolitic rocks. The second splay (F2) separates the Xigaze fore-arc basin from the aforementioned ophiolitic belt. Finally, the northernmost splay (F3) separates the Xigaze basin from the Kailas Formation, which locally covers the Gangdese batholith [8, 78]. As evidenced by the crosscutting relationship and geochronological evidence, the Great Counter thrust experienced its peak activity between ~23 and ~17 Ma [8, 78-80]. Sample site X28 located in the northern fore-arc basin is in the vicinity of the northernmost splay of the Great Counter thrust (Figures 2 and 3). This segment cooled through the APAZ between ~27 and ~18 Ma (Figure 4(b)). Hence, based on a close spatio-temporal relationship, the activity of the Great Counter thrust could be a possible driving force for the late Oligocene to early Miocene exhumation of the fore-arc basin (near the splay F3 of the Great Counter thrust) (Figure 6(a)). The samples collected close to splay F2 are situated in the footwall of the thrust fault and were less likely to experience accelerated cooling during the fault activity (Figure 3).The climate in southern Tibet (i.e., the Yarlung river drainage) has been significantly influenced by the Asian summer monsoon. Global monsoon systems are usually thought to be related with seasonal changes in response to land-sea thermal contrast (induced by seasonal evolution of solar radiation) [81, 82]. However, large-scale orography and zonal asymmetric diabatic heating can also significantly affect its development. For example, the uplift of the Tibetan Plateau enhanced the coupling between the lower and upper tropospheres as well as between the subtropical and tropical monsoon circulations, which intensified the East Asian summer monsoon [83-85]. The middle Miocene climate optimum (~17–14 Ma ago) is characterized by a significantly warm climate and high atmospheric CO2 concentration [86-88]. It corresponds to a phase of enhanced East Asian summer monsoon precipitation with large amplitude variability as recorded by the magnetic susceptibility of loess deposits [89]. According to the growing season precipitation modeling results for the eastern Himalayan Siwalik fossil leaf assemblages (the climate-leaf analysis multivariate program [90]), and the variation of the southeastern Asian offshore sedimentary accumulation rate [91], the intensification of the Asian monsoon may have lasted until ~10–9 Ma ago. In addition to the mid-Miocene, the Asian monsoon also intensified intermittently at ~5–2 Ma [92]. The increased precipitation could have resulted in enhanced river incision and stronger chemical weathering of rocks, being an important contributor to regional erosion and exhumation [93-95]. Hence with respect to southern Tibet, the influence of enhanced river incision due to climatic change could have been persistent in the late Cenozoic. The Buqu river finds its source in the Himalayas and is an important branch of the Yarlung river. The analyzed samples for our study were all collected from its valley (except for two Liuqu conglomerates; Figure 2) and have therefore been markedly influenced by the fluvial incision of this river. Considering the temporal and spatial coincidence, we hence propose that the enhanced Miocene cooling recognized in our studied profile was probably linked with the increasing Asian summer monsoon precipitation (Figure 6(b)). This may also account for the occurrence of relatively young cooling ages in the southernmost part of the Xigaze basin (Figure 2), which was previously considered to have only experienced pre-Oligocene exhumation [21].A series of ~N–S-trending normal faults were developed in southern Tibet and accommodated orogen-parallel extension [96-98]. Although their mid-to-late Miocene activities were almost simultaneous with the accelerated cooling phases recorded by the analyzed Xigaze basin samples, these faults are largely situated in the central and southern Lhasa subterranes and their extensional activity was more likely to have facilitated the basement exhumation of the Gangdese batholith to the north (Figure 1(c)). For example, the Pum Qu-Xainza rift is adjacent to our study area and the major phase of rifting took place ~12–8 Ma ago [99], but no field evidence reflects its southward extension to the fore-arc basin (Figure 2). Therefore, a tectonic triggering factor (e.g., rifting activity) for the mid-to-late Miocene denudation of the studied profile is excluded, and the climatic issues indicated above were likely the main driving force in controlling the river valley incision and resultant exhumation.It is worth noting that samples X26 and 27 from the Dazhuka Formation (i.e., the central section of the Kailas Formation) yielded similar mid-to-late Miocene AFT ages (~8 and ~15 Ma, respectively) to those from the Xigaze basin. Based on palynological data and detrital zircon U-Pb geochronology, it has been suggested that the Dazhuka Formation was deposited in the early Miocene (~23–18 Ma) in braided river systems [29, 52, 53]. The obtained AFT ages are only several million years younger than the depositional age of this formation, indicative of rapid burial and subsequent exhumation (> ~3–4 km). This observation is generally consistent with previous thermochronological studies from other sections of the narrow and elongated Kailas Formation on the southern margin of the Gangdese batholith [11, 100], which report mid-Miocene (~17–15 Ma) rapid exhumation of the Kailas conglomerates. We also stress the contemporaneous incision of the Yarlung river and its tributaries that could have facilitated the exhumation of the conglomerates. To the west of the main sampling profile, two conglomerates from the syn-convergent Liuqu Formation yielded AFT central ages of ~15 and ~10 Ma, respectively (Table 2). They are younger than the roughly constrained depositional age of this formation (~20–19 Ma [55, 101]). Based on detrital AFT, and apatite and zircon (U-Th)/He dating, Li et al. [101] suggested that exhumation of the Liuqu conglomerates commenced at ~12–10 Ma due to incision of the paleo-Yarlung river in the Indus–Yarlung suture zone. In Li et al. [101], the detrital AFT age peak (for grains that have been largely or totally annealed) is at ~11 Ma, and all AHe single grain ages are within the range of ~8–6 Ma. Our results are in good agreement with these data and also support a mid-Miocene exhumation of the Liuqu Formation.In this study, the analyzed sedimentary rocks from the middle part of the Xigaze fore-arc basin and the neighboring Dazhuka (Kailas) and Liuqu Formations display various late Oligocene to Miocene accelerated cooling events (Figure 4(b)). The late Cenozoic (e.g., late Oligocene to late Miocene) basement cooling is also widely documented in the Gangdese batholith to the north, particularly along its southern edge [102]. During this time span, the southern Lhasa terrane experienced a period of active tectonics and magmatism. On the one hand, regional faults such as the Gangdese thrust and Great Counter thrust became active during ~27–17 and ~23–17 Ma, respectively [78, 103, 104]. The west–east extension in the southern Tibetan Plateau also started in the latest Oligocene to early Miocene, causing the formation of a series of north-striking normal faults [6, 105, 106]. Particularly, to the north of Xigaze city, the Gangdese thrust is not exposed but overprinted by the south-dipping reverse fault system of the Great Counter thrust (Figure 2). These two thrust systems overlap, constituting a hinterland-dipping duplex controlled by subduction dynamics beneath the Gangdese batholith (magmatic arc) [8]. The development of this duplex has been suggested to be a possible driving force for the late Oligocene to mid-Miocene accelerated exhumation of the central Gangdese batholith [13]. On the other hand, Oligocene to mid-Miocene adakitic and ultrapotassic magmatism occurred in the Lhasa terrane and is thought to have resulted from a delamination event caused by the Indian slab roll-back and subsequent break-off [107-109]. Convective removal of the thicker lithosphere and resulting asthenosphere upwelling uplifted the overlying lithosphere over a wide area, rapidly raising the surface elevation [110-113] and potentially invoking a strong erosional response.Mid-to-late Miocene (~16–5 Ma) rock cooling is a very common feature observed in southern Tibet. In the central and eastern segments of the Gangdese batholith, a multitude of published thermochronological data indicates accelerated cooling during this period. It has been documented that southern Tibet underwent widespread syn-convergent extension in the mid-Miocene, and a number of ~N–S-trending normal faults were formed [114-116]. In addition, valleys and drainage areas of rivers such as the Yarlung, Lhasa, and Nimu rivers experienced rapid incision and hence facilitated exhumation of the underlying bedrock at ~15–9 Ma [12, 117], ~16–12 Ma [26], and ~12–8 Ma [118], respectively. This was probably due to the increased precipitation caused by the intensification of the Asia monsoon at ~15–9 Ma [12, 119, 120]. It is therefore suggested that the mid-to-late Miocene basement cooling and exhumation of the adjacent Gangdese batholith resulted from an interaction of both climatic and tectonic factors [13, 27, 77, 102], although it is difficult to determine which process was the main driving force based on low-temperature thermochronology alone.Our new AFT results, along with previously published low-temperature thermochronological data from southern Tibet, reveal that large parts of the Xigaze fore-arc basin (except for some outcrops in its southern part [21]) (Figures 4(b) and 4(c)), the neighboring central Gangdese batholith (Figure 4(d)), and the intermittent rocks along the middle reach of the Yarlung river (Figures 4(b) and 4(c)) exhibit highly comparable late Cenozoic thermo-tectonic histories. Episodic enhanced cooling phases prevailed between the late Oligocene and the present. As mentioned above, although the Tibetan Plateau has a prominent high-elevation low-relief topography [5, 121], the low relief of the Tibetan Plateau is not uniform. The Tibetan Plateau can be subdivided into internal and external drainage area (Figure 1(c)). The latter includes the southeastern part of the Lhasa terrane and adjacent ranges of the Himalayas and sources several major Asian rivers that flow out of Tibet such as the Yarlung river. The Xigaze fore-arc basin and the contiguous Indus–Yarlung suture zone also are part of the external drainage area and show relatively high relief. In the Lhasa terrane and the adjacent areas, the distribution of available low-temperature thermochronological ages is generally older than those in the external drainage area. Thermochronological age information from the internal drainage area is mainly between the Late Cretaceous and early Eocene, while most of the data from the external drainage area are younger than ~25 Ma as mentioned [102, 122-124]. Our new results from the fore-arc basin thus confirm the widespread nature of late Cenozoic (e.g., <25 Ma) exhumation of the external drainage area. The episodic late Oligocene to late Miocene cooling events in southern Tibet were jointly induced by deep crustal processes (e.g., rollback and break-off of the Indian slab and the Lhasa lithospheric root delamination), fault activity (e.g., movement of the Gangdese thrust and Great Counter thrust and development of normal fault systems), and enhanced river incision (e.g., the Yarlung river and its several tributaries) [12, 13, 25, 27, 77]. It is therefore further verified that more recent and widespread regional basement cooling (connected to both the tectonic and climatic aspects) facilitated the formation of the high relief of the external drainage area in southern Tibet.In order to further constrain the late Cenozoic thermo-tectonic evolution of the Xigaze fore-arc basin situated along the Indus–Yarlung suture zone in southern Tibet, we carried out AFT thermochronology on sedimentary rocks (mostly sandstone or conglomerates) collected from the middle segment of the Xigaze basin and the adjacent Dazhuka (Kailas) and Liuqu Formations. Our new AFT results reveal that all the analyzed sediments from the Xigaze Group and the Dazhuka and Liuqu Formations have been buried deeply enough to fully reset their AFT systems before their recent exhumation. Inverse thermal history modeling results derived from nine representative samples show that the northern part of the Xigaze basin experienced late Oligocene to early Miocene (~27–18 Ma) moderate cooling, and this event was probably associated with the activity of the Great Counter thrust. While mid-to-late Miocene accelerated rock cooling was more widely recorded along the studied profile (Xietongmen—Jiding), we propose that the coeval incision of the Buqu and Yarlung rivers contributed to the exhumation of the Xigaze basin and the adjacent Dazhuka and Liuqu conglomerates. Combined with published thermochronological data, it is shown that the Gangdese batholith, the Yarlung river bedrock, and the fore-arc basin in the central southern Tibet display a comparable late Cenozoic thermo-tectonic history, under the influence of both tectonic and climatic factors.The sampling information and compiled apatite fission track data used to support the findings in this study are included within the article, but detailed counting data and measurements will be available from the corresponding author upon reasonable request.The authors declare that they have no conflicts of interest.Shida Song analyzed the thermochronological data, completed the majority of visualization tasks, co-wrote the original draft, and prepared the supplementary materials; Zhiyuan He designed this work, took part in the field investigation, measured the confined track lengths, guided the analysis and figure drawing, and wrote the original draft; Wenbo Su participated in the fieldwork, carried out sample preparation, and analyzed and calculated the fission-track ages; Linglin Zhong took part in the field investigation and reviewed and edited the draft; Kanghui Zhong funded and organized the fieldwork; Stijn Glorie conducted the 252Cf irradiation and made corrections on the text; Yifan Song performed the literature study and provided suggestions on the discussion; Johan De Grave was responsible for text editing, supervision, and funding acquisition for lab equipment, consumables, and neutron irradiation.Hongjie Zhang, Zhao Yan, Jie Peng, and Yilong Long are thanked for their help and assistance during the fieldwork in 2019 and 2021. We thank Dr. Simon Nachtergaele, Ann-Eline Debeer, and Jan Jurceka for their assistance in sample preparation in the laboratory of Ghent. We are also grateful for the help provided by Dr. Bart Van Houdt and Dr. Guido Vittiglio with irradiation at the Belgian Nuclear Research Centre in Mol (SCK-CEN, BR1 facility). Dr. Christina Trautmann and Dr. Maria Eugenia Toimil-Molares helped to conduct the heavy ion bombardment for our samples at the GSI Helmholtz Centre, Germany. We thank Dr. Alexandre Kounov (University of Basel) and an anonymous reviewer for their very constructive comments on an earlier version of this work, and Dr. Andreas Wofler for editorial handling. This study is partly sponsored by the Deep Resources Exploration and Mining - National Key R&D Program of China (2022YFC2905001), and the Opening Foundation of the Ministry of Natural Resources Key Laboratory for Mineral Deposits Research, Chengdu University of Technology (grant number: gzck202104). The support provided by the China Scholarship Council (201908320260) is appreciated for financing the research stay of WS in Belgium. S.G. is supported by an Australian Research Council Future Fellowship (FT210100906).