Gold Open Access: This paper is published under the terms of the CC-BY license.

Staring out across the lush and unforgiving landscape of China’s Yunnan and southwestern Sichuan Provinces with a geologic map in hand, one is immediately humbled by the skill of the Chinese geologists who were tasked with a seemingly impossible job. The mapping was, by and large, accurate; however, the Cenozoic chronology was based entirely on biostratigrapic and lithostratigrapic correlation (e.g., Ge and Li, 1999) with little, if any, absolute age control. With limited exposure and elusive datable horizons except for a few dikes cross-cutting stratigraphy and interbedded lava flows (e.g., Studnicki-Gizbert et al., 2008), there were few constraints on absolute time in the Cenozoic. Because the area is important for understanding the evolution of the Tibetan Plateau as a whole, the community working in the area pressed on, leaning heavily on the regional stratigraphy in building a Miocene history of how rivers re-organized and incised the Tibetan Plateau’s southeast margin (e.g., Clark et al., 2004; Schoenbohm et al., 2006) in response to passive surface uplift resulting from flow of the lower crust (Clark and Royden, 2000) (Fig. 1A). This narrative was reinforced by subsequent low-temperature thermochronology that yielded a middle Miocene timing of river incision (Clark et al., 2005; Ouimet et al., 2010). Using largely the same geologic age constraints from the sedimentary basins, paleo-altimetry studies determined that the Plateau margin was already high in the Eocene (Hoke et al., 2014; Li et al., 2015; Tang et al., 2017), consistent with high elevations in southern and central Tibet (Rowley and Currie, 2006).

Over the past few years, sedimentary basins of the southeastern margin of the Tibetan Plateau have finally begun to yield absolute age constraints, including two new ages from volcanic ashes in the Lühe Basin reported in this issue of Geology (p. 3–6) by Linnemann et al. (2017). This spate of geochronology (Hoke et al. 2014; Gourbet et al. 2017; Linnemann et al., 2017) squarely reassigns the age of at least two basins from middle and late Miocene to the Paleogene (Fig. 1A). This reassignment of time, at a minimum, should force us to step back and assess how our interpretations of the southeast margin might change. Given their revised age for the Lühe Basin, Linnemann et al. point out that the much of the modern biodiversity was established in the early Oligocene, 20 m.y. earlier than previously thought. The past, apparently, looked very much like the present, and likely included the establishment of high elevation prior to the late Paleogene. Perhaps as important, time is much more compressed than initially envisioned. Taking the results from Gourbet et al. (2017) with those of Linnemann et al., it would appear that deposition of the Shuanghe and Xiaolongtan Formations occurred over a relatively brief window of time between 37 Ma and 32.5 Ma. These shake-ups in time will likely continue: the youngest zircon U-Pb ages (n = 12) for a tuffaceous sandstone reported by Wissink et al. (2016), in what is mapped as the Paleocene Yunlong Formation, average 36.3 Ma, making them equivalent in time with the Shuanghe Formation of the Jianchuan Basin. Keeping in mind that lithostratigraphy and biostratigraphy were the glue that bound correlations in the area, what other surprises await?

It is certainly within the realm of possibility that what we once thought was Miocene throughout the southeast margin of the Tibetan Plateau now resides in the later part of the Paleogene. This opens up the interesting possibility that these units record late Eocene to earliest Oligocene terrestrial temperatures from the region. There are a large number of plant fossil-derived temperature estimates data throughout the region from basins formerly thought to be Miocene (Xu et al., 2008; Xia et al., 2009; Sun et al., 2011; Yao et al., 2010; Jacques et al., 2011; Jacques et al., 2014), yet with a mean temperature of 17.6 ± 2.3°C over the 10° of latitude there appears to be no meaningful trend from basins that share fossil assemblages with the flora of Lühe and Jianchuan Basins (Fig. 1B). Branched glycerol dialkyl glycerol tetraether (GDGT) temperatures derived from a survey of modern soils of southeast Tibet and the southeast margin of the Tibetan Plateau yield coherent lapse rates that match weather station data (Wang et al., 2017), suggesting a potential horizon for understanding Paleogene temperatures and potentially elevation histories.

Linneman et al. and Gourbet et al. (2017) suggest that the paleo-elevation history of the southeast margin of Tibet be revisited in light of their revised ages, as they impact some of the assumptions used to calculate paleo-elevations. The most noteworthy of these are the shallower slope of the isotope-elevation gradient (Molnar, 2010) and the temperature corrections applied to pedogenic carbonates by Hoke et al. (2014). Gourbet et al.’s (2017) reanalysis of existing paleo-altimetry data applied a large continentality effect correction by assuming a mid-latitude −2.0‰ per 1000 km decrease in δ18O and a total distance of 1500 km. Li et al. (2015) applied the same continentality correction over much shorter distances. The continental effect at tropical latitudes, where this part of Asia was located in the Paleocene, can be absent or much weaker than that of the mid-latitudes. In South America, for example, there is a much weaker continental effect, with little to no fractionation of Atlantic sourced moisture across thousands of kilometers until the foot of the Andes (Insel et al., 2013).

Here I attempt to draw some additional insight on the elevation history of the Paleogene basins on the southeast margin of Tibet by examining the δ18Oc and δ18Omw values of authigenic carbonates alongside a revised isotope elevation gradient for a warmer, late Paleogene world. This simple analysis is limited to the data reported by Hoke et al. (2014) and Li et al. (2015) (Fig. 1A). Today, these basins span 5.5° of latitude along a NNW-SSE trend that reaches the edge of Tibet. If these basins were at low elevation, isotopic values would be little changed over this distance given their present and paleo-latitudes (Bowen and Wilkinson, 2002), and even with a strong continental effect, only ∼1.2‰ of fraction would be observed from north to south. Instead, the δ18Oc data exhibit a northward decrease in δ18Oc of 7.7‰ (R2 = 0.65; Fig. 1B). The δ18Omw values, calculated using plant fossil-derived mean annual air temperature (MAAT) plus 5 °C corrections, increase the total south-to- north decrease in δ18Omw to 10.4‰. It is difficult to explain such a large offset without a northward increase in elevation across the transect. A Rayleigh fractionation curve (Fig. 1C) using Rowley’s (2007)s model, assuming a +5 °C offset from modern temperatures and 81% relative humidity, generates a much shallower isotope elevation gradient than standard modern conditions (22 °C and 81% relative humidity). Again, in the spirit of keeping it simple and not going through the educated guesses of initial isotopic values, moisture pathways, or temperature corrections, the −7.7‰ difference in δ18Oc yields an estimate of ∼4.5 ± 1 km of relief from north to south (Fig. 1C). A modern isotope elevation gradient would result in a more modest 3.4 km ± 1 km of relief (Fig. 1C). These estimates are consistent with the modern topographic relief between these basins today. The assigned ages of basins of the southeast margin may change in age, but they still, almost certainly, reflect a high southeast margin of Tibet since at least the Eocene. More paleo-altimetry studies, ideally using different approaches, and continued refinement of geologic time, can test the elevation history developed to date. As the geochronology of this areas evolves, get ready for more surprises, for surely the narrative will change again.