Figure 3. Biotite 40Ar/39Ar release spectra for samples from granitoid plutons in the region. Arrows show the range of steps used in the weighted mean age calculation.
TECTONICS, VOL. 21, NO. 1, 10.1029/2000TC001246, 2002
[23] To better resolve the Cenozoic thermal history of this region of the Tibetan Plateau, we utilized a variety of thermochronometers including 40Ar/39Ar in biotite and potassium feldspar, and (U-Th)/He in zircon and apatite. Details of the analytical procedures are presented in Appendix A. Nominal closure temperatures in these systems range from ~300°–350°C (biotite) [Grove and Harrison, 1996] to ~70°C (apatite) [Farley, 2000], and multidiffusion domain modeling of K-feldspar 40Ar/39Ar spectra [Lovera et al., 1989] permits exploration of much of the temperature interval in between. Furthermore, the closure temperature for He diffusion in zircon ranges from ~160°–210°C [Reiners et al., 2002] and allows for an independent check on the reliability of the feldspar thermal models. Our sampling strategy was designed to test for possible spatial variations in the Cenozoic thermal history across a large geographic region. We chose to examine a suite of thermochronometers within each sample rather than utilize the age-elevation relationship of a single system because little information is available in eastern Tibet to effectively guide the latter approach. We collected three samples from the margin of the plateau. Two of these are from Precambrian massifs (Pengguan and Baoxing massifs; see Figure 2) along the margin adjacent to the Sichuan Basin, while one is from a Mesozoic pluton along the eastern foot of the Min Shan (Figure 2). We also collected four samples from the interior of the plateau, in the headwater reaches of one of the primary rivers in this region, the Hei Shui He (Black Water River). These four samples are from Mesozoic plutons within the Songpan-Garze terrane west of the thrust belt (Figure 2). We collected samples from the valley floors east of the drainage divide and from the low relief surface west of the divide, spanning ~2 km of relief.
4.1. 40Ar/39Ar Results—Biotite
|
|
[24] We obtained biotite separates from three granitoid plutons within the Songpan-Garze terrane. Samples EK 97-4 and 97-6 were collected from plutons in the headwaters of the Hei Shui He, while sample EK 97-14 was collected from a small (~3 km2) stock at the eastern foot of the Min Shan (Figure 2). Incremental heating of these samples yielded relatively straightforward release spectra (Figure 3), but no statistically defined plateaux. Moreover, the highly radiogenic nature of the 40Ar in the samples prevented their evaluation for possible excess 40Ar contamination [Roddick et al., 1980]. Our best estimates of the 40Ar/39Ar closure ages of these biotites are the 39Ar weighted means of the dates for increments defining relatively flat portions of the spectra: ~194 Ma for 97-4, ~208 Ma for 97-6, and ~171 Ma for 97-14. Although provisional, these dates are consistent with emplacement ages determined for similar plutons in the Songpan-Garze terrane to the south [Roger et al., 1995b], and probably record rapid cooling of the plutons following emplacement.
4.2. 40Ar/39Ar Results—K-Feldspar
[25] To resolve the thermal history of the Longmen Shan region between the Jurassic ages recorded in biotites and Miocene-Pliocene ages recorded in apatites (in section 4.3), we analyzed potassium feldspars from five samples distributed across the plateau and modeled the results following the multiple diffusion domain (MDD) theory [Lovera et al., 1989, 1991]. Whether or not this model accurately represents radiogenic 40Ar diffusion in natural feldspars over geologic timescales is controversial [Parsons et al., 1988; Villa, 1994; Parsons et al., 1999]. Persuasive as the mineralogical arguments against the MDD model may be, the fact remains that numerous applications of the method to samples from a variety of geologic settings have yielded sensible time-temperature paths that seem consistent with independent thermochronologic constraints [Arnaud et al., 1993; Leloup et al., 1993; Krol et al., 1996; Warnock and Zeitler, 1998]. Our approach has been to proceed with the modeling exercise and ask, whenever possible, whether the modeled temperature-time path is consistent with the higher- and lower-temperature constraints provided by the 40Ar/39Ar and (U-Th)/He data.
[26] The interpretation of the 40Ar/39Ar results in terms of the MDD model requires some assumption regarding the general form of the temperature-time path experienced by the samples [Lovera et al., 1989]. We restrict our models to the simplest possible solution, monotonic cooling. Transient reheating could significantly alter the interpretation of the thermal history. However, we are confident that we can discount the possibility of Cenozoic reheating for a variety of reasons. First, there is no evidence for Cenozoic magmatism in this region of eastern Tibet; all dated plutons within the study area are Mesozoic [Roger et al., 1995b]. The only recognized Cenozoic pluton is exposed in the Gongga Shan massif [Roger et al., 1995a], well south and west of the Sichuan Basin. Second, geothermal activity is very limited in this region of the plateau. Finally, preliminary analyses of fission track length distributions in samples from the northern portion of the margin (Min Shan) suggest slow cooling during the late Mesozoic and early Cenozoic [Arne et al., 1997]. Thus we feel that monotonic cooling is a reasonable first-order interpretation of the feldspar 40Ar/39Ar results. As discussed in section 4.2.3, it appears to work in all cases but one.
[27] Feldspars from the Longmen Shan region show differences in age and thermal history that appear to correspond with the location of the samples relative to the plateau margin. We first discuss samples collected from the margin adjacent to and north of the Sichuan Basin (93-4, 97-14), one from the southern Longmen Shan (93-1), and then turn to samples from the interior of the plateau (97-4, 97-6).
|
|
|
|
[28] Measured age spectra for samples 93-4 and 97-14 are presented in Figures 4a and 5a, respectively. Both samples are characterized by saddle-shaped release spectra that suggest the presence of excess argon in the low-temperature release steps. Duplicate isothermal heating increments [Harrison et al., 1994] permitted the isolation of late Miocene components of gas released early in the experiments. The ages of subsequent steps increase monotonically to maxima of ~147 Ma (93-4) and ~61 Ma (97-14). However, the spectra of sample 93-4 is complicated by the apparent presence of excess argon between 15 and 25% of gas released (Figure 4a). It is possible that the low-temperature steps (5–15% of gas released) are also contaminated with excess argon, and these should be considered maximum ages (circa 11–12 Ma). Diffusion parameters (activation energy, Eaand frequency factor, Do / r 2 ) calculated from the release of 39Ar are presented in Figures 4b and 5b and are typical of alkali feldspars [Lovera et al., 1997]. The domain size distribution for each sample is shown in Figures 4c and 5c.
[29] Model temperature-time paths (inverted from the release spectra and kinetic parameters; see Appendix A) are shown in Figures 4d and 5d. Both thermal histories are remarkably similar and suggest that the samples underwent extremely slow cooling during the Cretaceous and early Tertiary. The modeled cooling curve for 97-14 is poorly constrained by data at temperatures higher than ~240°C, but a simple extrapolation of the nearly linear portion of the curve (between ~200° and 240°C) to older ages is consistent with closure for 40Ar in biotite [Grove and Harrison, 1996] having been achieved at ~171 Ma, the approximate age of biotite from this sample. Modeled cooling curves show a dramatic increase in cooling rates (~20°–50°C/m.y.) during the late Miocene beginning at 11–12 Ma for sample 93-4 and 6-7 Ma for sample 97-14.
|
|
|
|
[30] In contrast to samples from the topographic margin, feldspars collected from Mesozoic plutons in the headwaters of the Hei Shui He (Figure 2) yielded minimum ages of ~50–70 Ma (Figures 6a and 7a, samples 97-4 and 97-6, respectively). Minor excess 39Ar was apparent in the first 5–10% of the gas released. The measured ages increase monotonically from the minimum to ~195–210 Ma. The diffusion behavior of these samples was slightly more complex than that of samples from the topographic margin. For example, duplicate isothermal steps for sample 97-4 display a systematic offset resulting in two subparallel arrays on an Arrhenius diagram (Figure 6b). We attribute this behavior to a slight hystersis in the temperature cycling that appears to be a consequence of our heating schedule for this sample (see supporting data Table A1, which is available as an electronic supplement1). Because the furnace had more time to equilibrate during the second isothermal increment, we relied on these results to extract kinetic parameters.
[31] Modeled thermal histories for these samples are also characterized by slow cooling during the Mesozoic and early Cenozoic (Figures 6d and 7d). Because these K-feldspars were more retentive of argon than the plateau margin samples, their modeled temperature-time paths extend to higher temperatures and older ages, permitting a direct comparison of the results with the biotite 40Ar/39Ar data for 97-4 and 97-6 (Figures 6d and 7d, respectively). In both cases, the biotite data are consistent with the K-feldspar cooling models.
[32] Sample 93-1 was collected from the southern portion of the Baoxing massif, in the southern Longmen Shan region (Figure 2). Its release spectra show evidence for considerable excess 40Ar contamination of the low temperature steps and a remarkable monotonic increase in the apparent ages of higher temperature steps from ~65 Ma to ~550 Ma.
|
|
[33] MDD modeling of the thermal history of this spectrum suggests relatively slow cooling between ~350 Ma and ~100 Ma (Figure 8d). Beyond this range the model is poorly constrained. This result is surprising because independent evidence implies that the southern Longmen Shan had a complex tectonothermal history over much of the ~350–100 Ma interval. For example, the region was the locus of voluminous basaltic volcanism during the Permian (Emei flood basalt) and of severe shortening during the Mesozoic [Chen et al., 1994a; Burchfiel et al., 1995]. It may be that the assumption of monotonic cooling for the purpose of multidomain modeling is inappropriate over this time interval, and we unfortunately have no other geochronologic data that might corroborate the result. In any event, the absence of Cenozoic ages in the low-temperature portions of the release spectrum appear to indicate that the sample resided at relatively high levels in the crust prior to Cenozoic time.
[34] (U-Th)/He thermochronology is a recently redeveloped technique that provides age information in the low-temperature range of thermochronologic systems [Zeitler et al., 1987; Lippolt et al., 1994; Wolf et al., 1996; House et al., 1997; Warnock et al., 1997; Wolf et al., 1997]. Helium diffusion in most fluorapatite appears to be a thermally activated process with a nominal closure temperature (for a cooling rate of 10°C/m.y.) of ~70°C [Farley, 2000]. Preliminary work on helium diffusion in zircon suggests that the closure temperature (dT/dtof 10°C/m.y.) is probably between 180° and 200°C [Reiners et al., 2002]. However, some aspects of zircon helium diffusion characteristics, possibly related to radiation damage, are not well understood at present; for the purposes of this work, we consider a possible closure temperature range of 160°–210°C. Details of the analytical procedures associated with age determination are presented in Appendix A. Data for apatite and zircon analyses from samples in this study are presented in Table 1; sample locations and associated ages are presented in Figure 2.
[35] Helium ages in both zircon and apatite cluster in two distinct populations that reflect the variations observed in the feldspar thermal models. Samples collected from the topographic margin of the plateau record systematically younger ages (late Miocene to Pliocene in both systems) than those from the plateau interior (Late Cretaceous/early Tertiary in zircon and mid-Miocene in apatite). Samples from the Pengguan massif (93-3 and 93-4, ~900 m elevation), along the margin of the plateau adjacent to the Sichuan Basin, yielded an age of 4.6 ± 0.3 Ma in apatite and 11.0 ± 0.9 Ma in zircon. A sample from the eastern foot of the Min Shan (97-14, ~1100 m elevation) yielded an apatite age of 3.2 ± 0.5 Ma and a zircon age of 4.7 ± 0.4 Ma. It is worth noting that sample 93-3 was collected in close proximity to two apatite fission track samples of Arne et al.[1997] and yields an age that is statistically indistinguishable from their results (see Figure 2).
[36] Samples collected from the plateau yielded apatite ages of 13.4 ± 0.8 Ma (97-2, ~1800 m elevation), 19.9 ± 1.2 Ma (97-4, ~ 2000 m elevation), 20.6 ± 1.2 Ma (97-5b, ~4500 m elevation), and 8.2 ± 0.5 Ma (97-6, ~4300 m elevation). Although these samples span ~2 km of relief, there is little correlation between age and elevation. Given the limited number of samples and their geographic distribution, however, we can say little about the geometry and position of the partial retention zone for He over the time interval of interest [Wolf et al., 1998]. An additional complication is introduced by the anomalously young age of one of the highest samples (97-6). At present, we cannot address whether there is substantial variation in apatite (U-Th)/He ages from the surface of the plateau [e.g., House et al., 1998]. Zircon from two of these samples (97-4 and 97-6) yielded ages of 86.8 ± 6.9 and 55.0 ± 4.4, respectively. Thus the variation in cooling ages between the topographic margin of the plateau and the interior appears to be a robust feature of both the apatite and zircon (U-Th)/He data. Similar variations in apatite fission track ages were observed in the Longmen Shan [Arne et al., 1997] and to the south, along the Xianshuihe fault [Xu and Kamp, 2000].
4.4. Comparison of Feldspar Thermal Models and (U-Th)/He Data
[37] One of the most striking results of this combined data set is the correspondence between the thermal histories derived from the feldspar MDD models and the (U-Th)/He data. In all samples where we have multiple systems whose ages overlap, the feldspar models are consistent with (U-Th)/He ages. This is particularly true of samples from the plateau margin (93-4 and 97-14), where the zircon and apatite ages tightly bracket the rapid increase in cooling rate inferred from the thermal models (Figures 4 and 5). Thus we are encouraged that feldspar MDD models retain reliable geologic information about the thermal histories experienced by these samples.
4.5. Summary: Cooling History of the Longmen Shan Region
|
|
[38] Thermal histories inferred from the combination of 40Ar/39Ar thermochronometry on biotite, (U-Th)/He thermochronometry on apatite and zircon, and alkali feldspar MDD modeling provide a relatively complete picture of the low-temperature cooling history of rocks in the Longmen Shan region following Mesozoic tectonism. Despite differences in age and in structural setting, all samples are characterized by remarkably slow cooling (<1°C/m.y.) between the early Jurassic and the mid-Tertiary (Figure 9 and Table 2). However, samples from the topographic front of the plateau margin (93-4 and 97-14) record systematically younger zircon and apatite (U-Th)/He ages than samples from the plateau interior (97-4 and 97-6). These differences are captured by feldspar thermal models which record a pronounced increase in cooling rate in the late Miocene and early Pliocene at the plateau margin. The Cenozoic thermal history of samples from the plateau interior is somewhat less well defined; cooling rates appear to have increased (~2°–4°C/m.y.) at some point between complete feldspar closure (~50–70 Ma) and apatite ages (~20–8 Ma). However, our data do not constrain when during this interval the increase took place. Plausible scenarios are discussed in section 7.1, but confident interpretation of thermal history of this region during the mid-Tertiary awaits additional data. Regardless, it is clear that samples from the plateau resided at or below the closure temperature for helium in apatite by the late Miocene/early Pliocene. In what follows, we explore the implications of these variations in the thermal history between the plateau margin and its interior for the degree and extent of late Cenozoic denudation.
Citation:
