The Neogene Elevation History of Southern Tibet
R.A. Spicer, N.B.W. Harris, M. Widdowson, T. Argles, A.B. Herman
Introduction
Determining the elevation history of Tibet is not only crucial for understanding the nature of crustal deformation during continental collisions but it is also essential for understanding regional anf global climate change associated with the development of the Asian monsoon system. During NERC-funded reconnaisance work in 1995 a plant fossil locality was found in the Namling Basin south central Tibet . Logistical constraints at that time prevented large enough collections being made for palaeoaltitude studies. Return field work during September 1998 followed an unusually strong monsoon. Although conditions prevented sampling of the Namling basin, both the Pabai Zong dyke swarm and the Daggyai Tso rift were successfully sampled and geothermal samples were collected for He analysis. During the return to Nepal further geothermal samples were collected, together with river waters for Sr isotope analysis (NH). A large collection of fossil plant material was made at Karmaya, south Nepal (RS). During a third trip in May 2000 over 400 samples were collected from the Namling localities and the section logged and sampled for 40Ar-39Ar dating. All plant material was photographed in Lhasa (RS) and then retained by Prof Guo Shuang-Xing for curating in Nanjing.
The Namling Basin
The western side of the Namling basin is underlain by a series of steeply-dipping predominantly lacustrine sediments rich in volcanic ash extending from the valley floor (4300m) to a capping sequence of lava flows (>4800m). Occasional lignite-rich horizons and immature palaeosols occur between the lacustrine units. Two sites rich in fossil leaves were found which, due to stratigraphical dip, are separated by 300m of vertical distance, with the lower one (locality A) lying at 4300m. At both sites abundant leaves exhibited minimal mechanical fragmentation and no size sorting. Leaves were well preserved as impressions in water-lain grey tuffs. Sedimentary features suggest quiet water deposition with no long-distance water transport of the leaves, whilst a rapid deposition is indicated by excellent fossil preservation. We infer that the leaves were derived from trees growing proximal to the lake margins rather than elevated sources on the distal edges of the basin. Click here for more detailed and illustrated leaf data.
Both localities yielded assemblages with similar floral composition. Taxa that normally rot quickly (Alnus, Salix) showed no sign of biodegradation which suggests the overall species composition is not biased due to preferential species decay1. The age of flora from locality A was constrained at 15 Ma by 40Ar-39Ar dating of a systematic log of the Namling section (Fig. 1)
Collected specimens were categorised into 35 morphotypes using leaf architecture and venation characteristics, scored for 31 character states and evaluated by the Climate Leaf Analysis Multivariate Program (CLAMP) using appropriate datasets for northern hemisphere temperate vegetation and climate including Asian sites. Click here for the Namling CLAMP scoresheet.
Palaeoaltitudes can be calculated in three ways:
(i) Floristic composition and the nearest living relative (NLR) method
This approach relies on identifying correctly the nearest living relatives of the fossils and determining under what climate, and therefore at what altitude, such taxa live today. This method, although widely used, is only useful for Quaternary material because species rarely exist for more than a million years or so. For 15Ma floras the constituent species are all extinct and the specimens can only be compared to their presumed modern descendants at the genus level at best. Unfortunately the environmental range of species within a given genus is often large (for example the genus Salix contains some species that are tropical in distribution while others are predominantly Arctic) so the NLR methodology for assemblages that can only be compared at the genus level necessarily involves large unquantifiable uncertainties. For this reason alternative methods need to be employed to determine the palaeoaltitude of the Namling flora.
(ii) Mean Annual Temperatures and lapse rates
Change in vegetation with altitude is related to lapse rates. CLAMP values for the Namling fossils yield a mean annual temperature (MAT) of 6.9±1.7°C for the dataset containing the so called "alpine nest" samples (here designated the PHYSG3ar set) and an MAT of 8.1±1.2°C for the dataset without the alpine nest (the PHYSG3br set)2. Miocene AGCM results suggest an appropriate sea level MAT for the Namling site of 22°C, assuming sea surface temperatures derived from oxygen isotopes3. Using a global average lapse rate of 5.5°C/km this equates to a range of altitudes for the PHYSG3a set of 2420-3050m (mean = 2920m), while the PHYSFG3b set gives 2300-2740m (mean = 2530m). Unfortunately lapse rates are known to be extremely variable in mountainous regions and highly dependant on local topography. We therefore regard altitudes derived using lapse rates as highly unreliable.
Figure 2. CLAMP mean annual temperature calibration curve
(iii) Enthalpy
Because enthalpy is a function of temperature
and moisture, two parameters critical to plant growth, enthalpy
can be decoded from foliar physiognomy using CLAMP4,5. Enthalpy
is a thermodynamically conserved parameter of the atmosphere,
related to both temperature and humidity, that quantifies the
total energy content of a parcel of air excluding the negligible
kinetic energy 23, 25. The equation used is:
h = cpT + Lvq + gZ = H + gZ
where h = moist static energy, cp is the specific heat capacity
of moist air at a constant pressure, T is temperature (in Kelvin),
Lv is the latent heat of vapourization, q is the specific humidity,
g is gravitational acceleration and Z is the altitude. Enthalpy,
H, =cpT + Lvq. Thus the difference between two estimates of enthalpy
for a given location should yield an estimate of their difference
in potential energy and hence height separation.
To see how this approach was applied to the floras of the Namling Basin click here to access a .pdf file of our recent Nature paper.
Figure 3 CLAMP enthalpy calibration curve
East-West extension in southern Tibet
East-west extension in southern Tibet has been accredited both to the dissipation of potential energy following plateau uplift6 and to the local accommodation of plate boundary forces7. In order to advance the debate this project has established the relationship between the onset of plate extension, plateau uplift and the state of the lithosphere. The timing of E-W extension in southern Tibet has been investigated through geochronological studies of both a north-trending graben, and a dyke swarm.
(i) The Daggyai Tso graben
Lava flows on the graben floor and granites cross cut by the normal faults bounding the graben were sampled in order trace the relationship between graben development and post-collisional magmatism and to constrain the timing of normal faulting. Attempts to date normal faulting directly were unsuccessful; a slickenslide fabric from the fault surface yielded no argon and an 40Ar-39Ar age of 34.5±1.3 Ma from a granite in the hanging wall of one of the major normal faults reflects the timing of cooling of the granite following crystallisation, and does not constrain the timing of normal faulting itself. The earliest lava flow from the graben is a quartz porphyry, dated at 53.5±0.3 Ma (biotite). These lavas are overlain locally by a reddish quartz-feldspar porphyry, which is itself intruded by north trending calc-alkaline dykes <3m thick. Dyke ages of 19.3±1.5 and 17.3±1.9 Ma (Williams et al., 2001) are within error of crystallisation of the flow (18.8±0.5 Ma). Although it was not possible to determine directly the relationship of the Miocene dykes and lavas to normal faulting, the orientation of the dykes indicates that their emplacement may be related to the stress field responsible for graben development, in which case the dyke ages suggest that the graben will have developed at, or after, 19Ma.
(ii) The Pabbai Zong dyke swarm
The Pabbai Zong dyke swarm comprises a series of sub-parallel N-S trending dykes, 2-5 km thick and has an east-west extent of <5km yielding a b factor of 1.003. Such modest extension is insufficient for the dykes to result from decompression melting of mantle with normal potential temperatures.
The enriched isotopic and trace-element signatures of the dykes are indistinguishable from those of shoshonitic lavas with a well-constrained source in the sub-continental mantle lithosphere8. Hence regional E-W extension and SCLM-derived magmatism are clearly linked, consistent with plateau uplift being a response to convective removal of part of the SCLM. Four shoshonitic dykes were dated from Pabbai Zong yielding ages between 18.3±2.7 Ma and 13.3±0.8 Ma (Williams et al., 2001). We conclude that the crust was undergoing extension by 18.3±2.7 Ma and we infer that by this time the plateau of southern Tibet had reached a sufficiently high elevation to have attained excess potential energy. Since these dykes were emplaced ~8 million years after the earliest dated potassic lava flows in southern Tibet9 it is probable that that this phase of plateau uplift was initiated even earlier, by 25 Ma (Fig. 5). The inference that the plateau of southern Tibet has been maintained at close to its present elevation (~5000 m) since at least the mid Miocene is supported by d18O studies of Tibetan carbonates10.
 |
Fig. 4 Time elevation map for southern Tibetan Plateau. Data indicate that 15 Ma ago the basin floor in Namling lay at an altitude of ~4600m, implying that there was only small topographic relief between plateau surface (>5000m) and intramontane basins during the mid Miocene. The elevation differential between the Namling basin and the plateau surface has therefore hardly changed between 15 Ma ago, and today. Pre-Eocene elevation is inferred from structural constraints in southern Tibet11 |
He isotope study
During field work in 1998, He samples collected from southern Tibet (LH) greatly expanded the sample set collected by SL in 1995. The resulting isotope analyses clearly defined two principle domains. In southernmost central Tibet helium isotope ratios (RC/RA<0.05) are typical of radiogenic helium production in the crust. This domain includes both the Daggyai Tso rift, and the Pabai Zong dyke swarm. Further north there is a resolvable 3He anomaly (RC/RA=0.1-0.4) consistent with a mantle contribution (Hoke et al., 2000). The boundary between these two domains lies 50-100 km north of the Indus-Zangpo suture zone and represents the southern limit of recent mantle melting and melt extraction beneath the Tibetan Plateau. This finding is significant because mantle-derived eruptions of Quaternary age are only known from northern Tibet8, yet the He data suggests that the zone of partially molten upper mantle beneath Tibet extends considerably further south.
The boundary between He domains coincides with the junction between the subducting Indian Plate and Asian lithospheric mantle as imaged in seismic and gravity profiles14,15. This suggests that the He anomaly results from degassing of melts at the base of the Asian lithosphere. Although shoshonitic magmatism (sourced in the SCLM) of mid-Miocene age is found in both Daggyai Tso and the Pabai Zong, the hydrothermal source of these geothermal fields in these areas is distinctly crustal, suggesting a northwards migration of the southern limit of mantle melting over the past 10 Ma.
References 1 Spicer (1981)U.S. Geological Survey Professional Paper 1143, 77p 2 Meyer (1992) Palaeogeography, Palaeoclimatology, Palaeoecology, v.99, 71-99 3 Valdes et al. (1999) Palaeoclimates: Modelling Ancient Weather (CD ROM) 4 Forest et al. (1995) Nature v.374, 347-350 5 Wolfe et al. (1998) Geol. Soc. Amer. Bull. v.110, 664-678 6 Molnar et al. (1993) Rev. Geophys. v.31, 357-396 7 McCaffrey & Nabelek (1998) Geology, v.26, 691-694 8 Turner et al. (1996) Earth Planet. Sci. Lett., v.37, 45-71 9 Miller et al. (1999) J. Petrol. v.40, 1399-1424 10 Garzione et al. (2000) Earth Planet. Sci. Lett., v.183, 215-229 11 Murphy et al. (1997) Geology, v.25, 719-722 12 Pan et al. (1993) Int. J. Radiat. Applications and Instrumentation , v.21, 543-554 13 Arnaud et al. (1991) Compte. Rendu. Acad. Sci., v.312, 905-911 14 Kosarev et al. (1999) Science, v.283, 1306-1309 15 Yu & McNutt (1996) J. Geophys. Res., v.101, 11275-11290 Additional References arising from this work: Hoke, L., Lamb, S., Hilton, D.R. and Poreda, R.J., 2000. Southern limit of mantle-derived geothermal helium emissions in Tibet: impllications for lithospheric structure. Earth Planet. Sci. Lett., 180, 297-308. Williams, H.M., Turner, S., Kelley, S. and Harris, N., 2001. Age and composition of dykes in Southern Tibet: new constraints on the timing of east-west extension in relationship to post-collisional volcanism Geology (in press). Harris, N., Spicer, R., Williams, H., Kelley, S., and Widdowson, M., 2001. Timescales of uplift of the Tibet Plateau. Earth System Processes, Edinburgh, J. Conf. Abstr., in press. Hoke, L., Kim, K., Harris, N. and Williams, H. 2001. The Daggyai Tso graben and geothermal field in southern central Tibet. Abstract volume, 16th Himalaya-Karakoram-Tibet Workshop, Graz, Austria, in press. Spicer, R., Herman, A., Guo Shuangxing, Widdowson, M., Harris, N., Argles,, T., Wolfe, J., and Valdes, P., 2001. First Results on the Miocene Palaeoaltitude of Southern Tibet Using CLAMP. (EUG, Strasbourg), J. Conf. Abstr, 6, in press . Williams, H., Kelley, S., Turner, S. and Harris, N., 1999. Post-collisional volcanism in North and South Tibet: implications for tectonic processes and late orogenic extension. EOS Transactions, AGU 80, F1016. Williams, H.M., Kelley, S.P., Turner, S.P. and Harris, N.B.W., 1999. New isotopic and geochronological results for volcanics in southern Tibet. EUG Strasbourg, J. Conf. Abstr, 4, 48. Williams, H., Turner, S., Kelley, S., and Harris, N., 2000. Post-collisional volcanism in North and South Tibet: erosion of heterogeneous lithospheric mantle. EOS Transactions, AGU 81, F1144. Williams, H.M., Turner, S.P., Kelley, S.P., and Harris, N.B.W., 2000. Low-degree mantle melting beneath Tibet: signals of heterogeneous lithosphere erosion. 10th V.M. Goldschmidt Conference, J. Conf. Abstr, 5, 1095.
