A recent study by Hetzel et al. (2011) applied thermochronologic and cosmogenic methods to a “low-relief” landscape in the northern Lhasa terrane at elevations >5 km (Fig. 1A). Modeling by Hetzel et al. suggests rapid late Cretaceous–early Eocene (between ca. 80 and 50 Ma) exhumation at a rate of ∼300 m/m.y., followed by a phase of slow exhumation to the present day (∼10 m/m.y.), which is consistent with cosmogenic results. Hetzel et al. interpreted the rapid exhumation stage to be a planation process, which formed a pre-collision (pre-50 Ma) near-sea-level “low-relief” landscape, which they termed a peneplain. As acknowledged by Hetzel et al., this conclusion can only be reached if (1) coeval regional crustal shortening was subdued, and (2) the drainage systems in the northern Lhasa terrane were connected to the sea, so as to allow rivers to incise and erode the bedrock laterally over large distances.

However, neither of these two conditions is supported by geological evidence. With respect to (1) above, in the central Lhasa terrane, >55% (>230 km) of late Cretaceous–Paleogene (ca. 90–53 Ma) north-south crustal shortening occurred and formed the north-verging Gangdese retroarc thrust belt (e.g., Kapp et al., 2007a) (Fig. 1A). In the northern Lhasa terrane and the Bangong suture zone between the Lhasa and Qiangtang terranes, late Cretaceous–Palaeogene (ca. 100–50 Ma) shortening of >58 km (>47%) is documented by the development of a south-verging northern Lhasa terrane thrust belt (e.g., Kapp et al., 2007b) and an ∼50 m.y. (late Cretaceous–Eocene) depositional hiatus in the southern Nima Basin (DeCelles et al., 2007) (Fig. 1A). Rohrmann et al. (2012) reported a similar exhumation history to that of Hetzel et al. from nearby localities, but by taking the shortening phase into account, suggested that a pre-45 Ma high plateau existed in what is now the Lhasa terrane.

With respect to (2), a compilation of sedimentary studies of late Cretaceous–Eocene basins in the Lhasa terrane and adjacent areas suggests a very different paleogeography from that required by Hetzel et al. During late Cretaceous–Paleogene time, the Gangdese Arc acted as a drainage divide, separating fluvial systems in the northern Lhasa terrane from the Neotethys Ocean; and areas to the north of the Gangdese Arc were well above sea level, and internally drained (Fig. 1B). Specific details include the following.

(1) In the early-mid Eocene Liuqu conglomerate (<55 Ma), overlying the Indus-Yarlung suture zone, Lhasa terrane detritus (e.g., porphyritic andesitic volcanic or granitic clasts) is conspicuously absent (Aitchison et al., 2011, and references therein). The conglomerate was more likely sourced locally from uplifted Cretaceous strata of the Gangdese forearc (the lower portion of the Xigaze Formation) (Wang et al., 2010).

(2) The Gangdese forearc basin (Xigaze Formation), whose deposition continued to ca. 50 Ma, was mainly derived from the proximal Gangdese Arc (Ding et al., 2005; Wu et al., 2010; Aitchison et al., 2011).

(3) The fluvial Lhunzhub Member (>1500 m) of the Takena Formation (<113 to <52 Ma), was deposited in a retroarc terrestrial foreland basin setting with the sediment-source inferred to be from the Gangdese Arc (e.g., Kapp et al., 2007a).

(4) The study of DeCelles et al. (2007, p. 654) on the Cretaceous-Cenozoic Nima basin, stated that “…By Aptian time the Nima basin was above sea level and was strongly influenced by nearby volcanic activity and crustal shortening in the reactivated Bangong suture zone ….”

The information presented above strongly challenges the possibility of the existence of a pre-collision near-sea-level peneplain in the northern Lhasa terrane. The exhumation history of Hetzel et al. is therefore reinterpreted as follows: late Cretaceous–early Eocene rapid exhumation was triggered by coeval crustal shortening (Fig. 1B). Eroded material was deposited locally in terrestrial basins within the northern Lhasa terrane. Greatly reduced exhumation in post-Eocene time led to the development of a low-relief landscape at high elevation in an internally drained setting. The revised interpretation presented here supports the proposition that the Lhasa terrane was well above sea level, and that elevation gain in the Lhasa terrane commenced before collision (e.g., Kapp et al., 2007a).

We are grateful to Wenjiao Xiao and Guangwei Li for their discussions and to Paul Kapp for his constructive review.