TECTONICS, VOL. 21, NO. 1, 10.1029/2000TC001246, 2002
7.1. Timing of Uplift in Eastern Tibet
[46] The timing and rates of denudation across the eastern margin of the Tibetan Plateau have important implications for the development of topography in this region. As discussed in section 2, regional topographic gradients between the plateau and the Sichuan Basin are among the highest anywhere on the continents. Given the structural position of the samples at the foot of this topographic escarpment, it is unlikely that the slow denudation rates inferred prior to the late Miocene/early Pliocene could have been maintained in the presence of such extreme topographic gradients. This scenario requires climatic conditions approaching those in present-day Antarctica [Brook et al., 1995], and is clearly not appropriate for central Asia during the Miocene. Indeed, present-day incision rates along rivers draining this margin range from 1 to 2 mm/yr [Kirby and Whipple, 2000], attesting to the erosive power of streams along the eastern front of the escarpment. Thus we interpret the pronounced late Miocene increase in cooling rates as marking the initial development of regionally significant topographic gradients at the present plateau margin, presumably in response to crustal thickening.
[47] Determining the Tertiary denudation history in the headwater regions of the river valleys west of the Longmen Shan is complicated by the incomplete resolution of our data. As noted in section 4.5, our data do not discriminate between a period of slightly enhanced denudation and cooling starting in the early Tertiary and a prolonged period of slow cooling followed by an rapid increase in the middle to late Miocene. We prefer the latter scenario for several reasons. First, Arne et al. [1997] collected an age-elevation transect of apatite fission track samples that is located midway between the plateau margin and the interior. Apatite fission track ages are invariant over ~1500 m and range from ~4–7 Ma (Figure 2), implying relatively rapid cooling during the late Miocene and early Pliocene. As there are no major structural discontinuities between these samples and our plateau samples, this site probably represents a similar structural level. Second, a regional study of apatite fission track ages along the Xianshuihe fault [Xu and Kamp, 2000], south and west of the Longmen Shan region, concluded that much of the region was cooling slowly between circa 130 Ma and circa 22 Ma. Finally, a recent study of apatite (U-Th)/He ages in the same region [Clark et al., 2000] is consistent with slow cooling (~1°C/m.y.) from 25-9 Ma. From a regional perspective it appears that much of eastern Tibet was characterized by relatively slow cooling during the early and mid-Tertiary. While denudation rates may have increased (~0.1–0.2 mm/yr) on the plateau west of the Longmen Shan sometime during early Tertiary, we believe that it is more likely that rates remained slow until the Miocene.
[48] Slow denudation during the early and mid-Tertiary, however, does not preclude the existence of substantial elevation in the region. The region west of the plateau margin was a source for sediment in the Sichuan Basin from Jurassic times [Burchfiel et al., 1995] and may have been a residual highland from the Mesozoic orogeny. Although we find it unlikely that such low denudation rates (<0.1 mm/yr) could be maintained on steep regional topographic gradients, it is possible that a gently sloping, low-relief upland existed in the region prior to Cenozoic tectonism. Thus we can conclude little at this time about the absolute magnitude of elevation changes as a result of the development of the Tibetan Plateau in this region.
7.2. Landscape Evolution Along the Eastern Margin
[49] The thermal histories presented here have important implications for the long-term morphologic evolution of the eastern Tibetan Plateau adjacent to the Sichuan Basin. In particular, our data place preliminary bounds on (1) the “age” of the present surface of the plateau and (2) the evolution of the modern relief along the escarpment. As described in section 2, large regions of the eastern plateau comprise a low-relief surface cut on vertically foliated Triassic flysch. Although this surface (and any “erosion surface”) is a dynamic landform, currently evolving as an erosional and depositional feature, one can define three useful temporal characteristics of such a surface: (1) the time at which the surface was reduced to low relief, (2) the duration of such relief, characterized by slow erosion rates, and (3) the time at which the surface began to be significantly dissected. While our data are far from comprehensive, the thermal histories of samples from the surface (97-6) and 2 km below it (97-4) suggest that the low relief atop the eastern plateau today may have developed in the Mesozoic and been sustained since then by relatively slow uniform denudation. Subsequent dissection of this surface in response to the development of regional topographic gradients between the plateau and the Sichuan Basin apparently initiated in the late Miocene or early Pliocene, and most of the local relief along the margin (which now exceeds 3 km) probably developed since that time.
7.3. Timing of Normal Faulting in the Longmen Shan
[50] Our results from the Pengguan Massif, adjacent to the Sichuan Basin, place additional constraints on the timing of normal faults developed within the Longmen Shan thrust belt. The massif is an elongate NE trending body of Precambrian gneisses and granitoids and is bounded on the NW by a series of low-grade mylonite zones with top to the NW sense of displacement [Burchfiel et al., 1995]. Although it is possible that displacement on these fault systems is responsible for some of the rapid cooling observed since ~11 Ma, we see no evidence for a 25–30 Ma event as suggested by Hames and Burchfiel [1993]. Indeed, throughout the Longmen Shan, our data suggest that the early and mid-Miocene was characterized by regionally uniform slow cooling. Thus, if the faults have Cenozoic displacement, it appears to be restricted to the late Miocene-Pliocene. We note that Cenozoic displacement on these faults provides a mechanism for exhumation without large amounts of net shortening across the margin. However, the exact timing of displacement on these normal fault systems remains an important problem in this region.
7.4. Processes of Plateau Formation
[51] The apparent absence of large magnitude Cenozoic shortening of the upper crust in eastern Tibet [Burchfiel et al., 1995; Wang et al., 1998; Kirby et al., 2000] suggests that crustal thickening may have been accomplished by ductile flow in a weak lower crust [Royden, 1996; Royden et al., 1997]. A recent analysis of the regional topography in eastern Tibet [Clark and Royden, 2000] suggests that the morphology of the plateau margins may be a consequence of the strength of the foreland region; adjacent to the Sichuan Basin the margin is steep because weak lower crust ponds against a strong foreland region, while in southeastern Tibet a weak foreland allows the development of a diffuse topographic transition between the plateau and the foreland. Our results bear on two aspects of this model: (1) the nature and degree of upper crustal deformation and (2) the lateral growth of the plateau.
[52] The amount and timing of Cenozoic shortening within this region of eastern Tibet have been the subject of some debate. Although most workers in the region acknowledge that there has been Cenozoic displacement on a number of the faults in the region [Chen et al., 1994b; Dirks et al., 1994; Burchfiel et al., 1995; Arne et al., 1997; Kirby et al., 2000], the magnitude and rates are uncertain [cf. Avouac and Tapponnier, 1993; Burchfiel et al., 1995]. The sharp contrast in cooling histories between rocks along the plateau margin and those in the foreland implies that there has been ~8–10 km of slip since the late Miocene along the frontal fault system adjacent to the Sichuan Basin. This displacement could be easily accommodated over the past 5–10 m.y. by mean slip rates of ~1–2 mm/yr. Such rates are consistent with the observed geodetic measurements that restrict modern shortening to <2–3 mm/yr [e.g., Chen et al., 2000] and with the observation of limited Cenozoic shortening in the southern Longmen Shan [Burchfiel et al., 1995].
[53] A similar contrast in cooling histories across the eastern front of the Min Shan implies that a similar amount of exhumation must be accommodated along the eastern range front. Our results, in conjunction with Cretaceous apatite fission track ages obtained by Arne et al. [1997], indicate that substantial late Cenozoic denudation in the Min Shan region was confined to a narrow (<40 km) corridor along the eastern flank of the range (see Figure 2). Differential exhumation was probably facilitated by displacement on the Huya fault and associated structures (Figure 2).
[54] A number of geodynamic models of plateau development [England and Houseman, 1986], including those invoking lower-crustal flow [Royden et al., 1997], suggest that the plateau grew laterally with time. If the plateau has been growing east and northeast by the evacuation of lower crust from beneath the plateau, our results suggest that it did not begin to abut the Sichuan Basin until at least ~11–12 Ma (perhaps as young as 5–6 Ma) and contradict the recent suggestion that the present configuration of the plateau margins was established by circa 20 Ma [Xu and Kamp, 2000].
[55] The slight increase in Tertiary cooling rates observed in the samples from the interior of the plateau could reflect the development of a migrating wedge of topography built as lower crust was evacuated from beneath central Tibet [e.g., Clark and Royden, 2000]. It is important to note, however, that there is no indication in our data of a steep former topographic margin, subject to deep erosion. Rather, the relatively slow cooling rates of samples now on the plateau (2°–4°C/m.y.; Table 2) imply that it would have had a relatively gentle regional slope, perhaps analogous to the present situation in southeastern Tibet. A definitive test of this hypothesis, however, awaits further sampling.
[56] Finally, the distribution of Cenozoic denudation across the eastern margin of the plateau has important implications for the coupling between surficial and tectonic processes along the margin. Our results suggest that since the late Miocene/early Pliocene, denudation has been heterogeneous across the plateau margin. In both the Longmen Shan and Min Shan regions the zone of deepest exhumation corresponds to the present-day topographic front of the plateau and to a zone of high gradients along modern rivers [Kirby and Whipple, 2000; Kirby, 2001]. This spatial correlation of long-term exhumation of midcrustal rocks and steep river gradients may reflect the erosional enhancement of mass flux along this margin [e.g., Beaumont et al., 1992; Willett, 1999], and it raises the possibility that the steep topography adjacent to the Sichuan Basin is, in part, a response to high erosion rates along the margin.
Citation: